Home

inForm - User Help

1. category In this case click the drop down next to Column Labels and de select any categories that are not of interest 8 In the PivotTable Field List click the check box next to Area pixel to add Area pixel to the Values box 9 You should now have a PivotTable that shows total pixel area for each tissue category of interest and a Grand Total The table has one row for each Sample Name or Slide ID depending on the selection for Row Labels 10 To display pixel area as a percent right click on Grand Total and select Value Field Settings Select the tab Show Values As In the drop down select of Row Total and click OK You should now have a pivot table that displays percent of total area for each tissue category The values of interest can be read directly from the table For example a tumor category displays the tumor load Compute a Ratio The PivotTable does not easily display a ratio of two categories it only shows each category as a fraction of the total Additional ratios can be added as new computed columns outside the pivot table Use explicit row column identifiers in the formula such as D3 Don t use the GETPIVOTDATA function that Excel provides if you click on target cells For example 1 Click on the cell in the first empty column in the first data row 2 Type the desired formula such as C5 B5 and press Enter 3 Select the entire range in the target c
2. i A f 1 2 Threshold Optical Density Cellular objects above this value are included in the 3 bin g 0 100 Score Results Percent Shows the percentage positivity of the cell nuclei membrane or cytoplasm within each bin 2 3 Threshold Optical Density H Score The calculated H Score of the image The H Score g 0150 6 is calculated using the percentages in each bin and ranges from 0 to 300 Score Results Percent W Show Score Colors Change Color Oto 1 1 to 2 2 to 3 3 a H score 14 Figure 38 0 3 Scoring 6s inForm User Help Ten bin Score IHC or IF Settings research use only This histogram bins the data into ten bins for more Tissue Category Category2 quantitative work The 10th Bin Nominal Level slider Scoring 10bin x specifies the ninth bin lower threshold The other eight Marker Settings thresholds are equally spaced between zero and the ninth Compartment Nuclei Cyto Membrane threshold Component ER DAB z The Histogram Results boxes display the percentages of Threshold Max 0 30 z pixels falling into the bins determined by the nominal level 10th Bin Nominal Level Optical Density slider position g 0m 0 0 03 Histogram Results Percent V Show Score Colors Bin 1 Bin 2 coe Bin 3 Bin 4 Bin 5 Bin 6 Bin 7 Bin 8 Figure 39 10 bin Scoring Scoring IHC or IF e Fifty bin Divides signals into 50 equal bins The 50th Bin N
3. Algorithms may also include threshold or colocalization settings depending on the project created to analyze the images Processing regions are user drawn regions of interest Use processing regions to only process specific regions on an image All areas outside of the marked processing regions are ignored A set of images that will be processed using the settings in the Manual Analysis tab Images in the processing set display a green square on the image thumbnail when viewing images in Single mode All images added to a project are included in the processing set by default To add or remove images from the processing set right click on the large image in the image display area and click Processing Set You can also click the Image List Editor button to open the Available Images window and then click the Load Images button inForm projects include the images that were added to the project and the settings for each step in the algorithm A new project that contains only the Prepare Images and Export Step is automatically created when opening inForm You can work in the new default project create a new project or open a saved project Tissue Categories identify structures such as tumor cells vs healthy cells that are present in the tissue samples Regions in the image can be marked manually using inForm Basic Analysis or higher Use inForm Tissue Finder to use the Tissue Segmenter to automatically classify tissue The Tissue Segmenter auto
4. Second Marker Settings Compartment Nuclei Cyto Membrane Component Her2 Fast Red M Threshold Max 0 20 Auto _ Positivity Optical Density g 0 100 0 0 0 2 Negativity Positivity 100 00 0 00 Histogram Results Percent V Show Score Colors Double positivity z Change Color Double Positive Single Positive 2 Figure 37 Double Positivity Scoring gle Positive 1 Double Negative 77 L Scoring IHC or IF value in the Positivity text box Positive regions are indicated by the Single Positive 2 color Negative regions are indicated by the Double Negative color 7 To view the double positivity colors select Double Positivity in the Histogram Results box 0 3 4 bin Score IHC or IF Settings research use only This option bins spectrally unmixed signals in nuclei Tissue Category Tumor membrane or cytoplasm into four bins This score type can Scoring 0 3 4bin z be used to calculate H scores with nuclear stains Marker Settings 0 1 Threshold Specifies the value below which to include eS ae Go O Maia cellular objects in the 0 to 1 bin Component ER DAB z Threshold Max 0 30 1 2 Threshold Cellular objects above the 0 1 threshold and 0 1 Threshold Optical Density below the 1 2 threshold are included in the 1 to 2 bin g 0 050 6 2 3 Threshold Cellular objects above the 1 2 threshold and 0 0 03 below the 2 3 threshold are included in the 2 to 3 bin
5. and many other data analysis programs 8 In Table Fields to Export select whether to export all available table data fields or only the fields that are visible in the table view To select which fields are visible in the table use the View Editor see Displaying the Extracted Data 73 9 Select File Name Options If files are already saved in the selected export directory select how to name the new files You can overwrite existing files if the new file names match or rename the new exported files to preserve any existing files in the export directory 10 If you want to replicate the directory structure of the source images to the export directory enter the number of directory levels to replicate in the Copy ___ level s text box For example by entering 2 in this field data in a CancerStudy Sample001 directory would be exported to the lt ExportDirectory gt CancerStudy Sample001 directory This helps keep the exported images organized 11 Click Export For Selected or Export for All to export the data The figure below shows a sample list of exported files File Edit View Favorites Tools Help Q Bak Q x oS Search 5 Folders Ei Address C Program Files inForm Data Al Name Size Type File and Folder Tasks 13036_8160_5100_cell_seg_data txt 699KB Text Document 13036_8160_5100_score_data txt 2KB Text Document al Reame fhis Re 13036_8160_5100_tissue_s tif 85KB TIF Im 8160_5100_
6. include in the selected Tissue Segmentation mask Figure 23 Editing Tissue Segmentation Masks 4 To continue processing the images using the edited Tissue Segmentation masks advance to the next step without re segmenting the images 5 Ifthe project contains a cell segmentation step and cell segmentation was performed before the mask was edited see Segmenting Cells 5 you should click the Segment Images button in the Cell Segmentation editor to re segment the cells in the edited tissue segmentations 52 inForm User Help Re Training the Tissue Segmenter If some areas of the image were not classified as the desired tissue type you can re draw the training regions around the areas that were incorrectly classified and re train the Tissue Segmenter To re train the tissue segmenter 1 Inthe Tissue Segmentation Training panel click the Draw radio button for the tissue category that you want to draw additional training regions for 2 Draw training regions around areas that were incorrectly classified 3 Re train the tissue segmenter and then re segment the image Verify that the desired regions are now correctly classified in the desired Tissue Segmentation mask 4 To see the tissue segmentation on all images segment all images Note that re segmenting an image after retraining the tissue segmenter deletes any manual edits to the Tissue Segmentation mask 10 6 Completing the Automated Tissue Segmentation Step A
7. inform ADVANCED IMAGE ANALYSIS SOFTWARE inForm V2 0 2 User s Manual inForm 2 0 Tissue Finder Advanced Image Analysis Software US Patents 7 555 155 7 953 264 8 280 140 and patents pending i Copyright 2008 2013 PerkinElmer Inc All Rights Reserved PerkinElmer Notice The information in this documentis subject to change without notice and should not be construed as a commitment by PerkinElmer Inc PerkinElmer assumes no responsibility for any errors that may appear in this document This manual is believed to be complete and accurate at the time of publication In no event shall PerkinElmer be liable for incidental or consequential damages in connection with or arising from the use of this manual This manual describes how to use inForm version 2 0 2 software For more information contact PerkinElmer Inc 68 Elm Street Hopkinton MA 01748 USA Phone 800 762 4000 or 1 203 925 4602 Fax 1 203 944 4904 Email global techsupport perkinelmer com Web site http Wwww perkinelmer com This software covered by US Patent 7 555 155 7 953 264 8 280 140 and patents pending Table of Contents Chapter 1 Welcome to InFOrM 2 522225222252222522022220222202222222222222222222222 522 x 6 Chapter 2 INtroductions sssssssssssssssssnsssssenserssnsenseessnseseessssnsnssessnnscorsenanseessssennoess 7 1 SAD OUR DEIO cit atin Gaveatendelde onic Ss cnc wdet bane chivanadtiec yaaa E N 7 2 Software Con
8. without scaling or other changes in value wavelength by wavelength from every pixel of the image This alters the spectral shape of the remaining signals and their overall intensity The magnitude of these changes depends on the intensity of the spectral signal being subtracted Typically when background spectra are small compared to the signals spectral library entries derived from unsubtracted data sets can still be used but this should be validated when quantitative results are required Ideally the remaining spectra in the library should also have the haze spectrum subtracted in Nuance 3 inForm User Help 8 3 Viewing a Spectral Library The Spectral Library window displays a chart for each spectral segment in a spectral library To view the spectral library 1 Open the desired images and spectral library 2 Select the spectra that will be used to unmix the images in the Spectra for Unmixing list box 3 Click the View button below the Spectra for Unmixing list box to open the Spectral Library Window 24 4 Select the desired Units to display the stains in the charts See Spectral Library Windowl 24 for complete descriptions 5 Select the desired Display options for the chart s 6 To save an image of the chart with the currently selected display options click the Save Chart button type the desired name for the image file select the desired image format and click the Save button 7 Click the Close X button
9. 0 020 00 a we Cho Ji Visbile More gt gt Figure 24 Colocalization Settings 6 To view the threshold for each component clear the Visible check box for all components Colocalization except the component that you are adjusting Clear the Visible check box for Colocalization Adjust the Threshold for the component until all of the desired areas are marked as positive 7 If there are small regions that are marked as positive that should not be included or if smaller regions should be marked as positive and are not click the More button and adjust the Minimum Connected Pixels value as desired 8 Under Colocalization click the Visible check box to show the colocalization map If desired click the Color button and select the color to use to indicate areas of colocalization 9 To view the Colocalization statistics see Viewing the Quant Data Tablel s and Viewing the Colocalization Data Tablel 84 10 To export the data see Exporting the Datal 89 ss inForm User Help 13 Segmenting Cells Cell segmentation locates individual cellular or subcellular objects within an image field or within a selected tissue category if tissue has been segmented By first identifying the location of all cell nuclei inForm is able to identify cells and their associated cytoplasm and membrane Once individual cells are located the set of pixels associated with each cell is identified so that marker signals can be extracted for analysis a
10. Classification step cannot be used in Batch processing Creating Tissue Categories Tissue categories divide the image into specific areas that correspond to specific types of tissue for example cancer necrosis stroma etc or occasionally other structures of interest The Manual Classification step available in inForm Basic Analysis or higher enables you to draw the appropriate tissue categories over areas of each tissue type on the images Cell segmentation scoring and object counting can all be restricted to a specific tissue category 1 Click the Manual Classification step to display the Tissue Segmentation panel 2 Click the New button under the Tissue Categories section to create a new tissue category Tissue Segmentation Tissue Categories Draw Merge Batch Analysis Manual Analysis 3 To change the tissue category name highlight the name and then type the desired name for the tissue category 4 To change the display color for the tissue category click the Color pull down and select the desired color 5 Repeat steps 2 through 4 to create all of the desired tissue categories 6 See Drawing Tissue Category Regions a Drawing Tissue Category Regions Draw tissue categories around the areas on the image that contain each defined tissue type A single area on the image cannot belong to more than one tissue category To draw tissue category regions on an image 1 Click the Draw option button next t
11. Red 0 000 2 000 E Secondary 0 000 2 000 Tertiary 0 000 2 000 Figure 32 Cytoplasm Pixel Validation At the top of the Cell Segmentation Settings panel select the Membrane check box in the Compartments to Segment box Note that you must segment the nuclei in addition to segmenting membrane inForm only searches for cytoplasm where nuclei have been found Click the Membrane tab Selecting Components for Membrane Segmentation 1 Select the Primary component signal for membrane segmentation 2 Select the Secondary component if desired 3 For each component set the Full Scale e Select an image that is typical of all of the images to be analyzed e Click the Auto button to find the brightest pixels in the image for each of the selected components The brightest value for each component displays in the Full Scale box for each component The units can be counts or OD Components for Membrane Segmentation Full Scale Primary oD ER DAB 0 012 Secondary 0 000 Auto Figure 33 Components for Nuclear Segmentation ea inForm User Help Specifying the Segmentation Priority The Segmentation Priority specifies which cell compartment Nucleus or Membrane has higher priority when When Membrane crosses Nucleus assign to segmenting a membrane that crosses a nucleus e Seiden Segmentation Priority Figure 34 Segmentation Priority Specifying a Maximum Cell Size The Distance to
12. Scoring box Descriptions of each method are given below Compartment Select whether you want to score the Nuclei the Cytoplasm or the Membrane cell compartment If desired you can score the nuclei save the data and then score the cytoplasm and or membrane separately The scoring data is a result of binning a histogram of all pixels within the selected compartment Component Select the component to use for calculating scoring data Batch Analysis Manual Analysis Renew Morgo Score IHC or IF Settings research use only Tissue Category Category z Scoring Positivity 2din z Marker Settings Compartmert Nucl Cyto Membrane Component Hemataxyin Threshold Max 0 20 Ato Postivty threshold Optcal Densty 0 032 0 0 0 2 Score Resuts Percert J Show Score Colors Change Color Negatvty Postiv Figure 36 Score IHC or IF Settings Threshold Max Click the Auto button to calculate the Threshold Max for the selected component The Threshold Max specifies the maximum value of the Threshold sliders If necessary adjust this value up or down as needed For OD converted images 3 0 is the maximum possible value Positivity Threshold Specifies the threshold values for the selected Scoring Type Drag the slider type the desired value in the text box or click the up and down arrow buttons to change the Positivity Threshold value See Scoring Types 6 for descriptions of each Threshold Show Score Color
13. Tissue Finder algorithm in the Algorithms folder of the VectraData directory Tissue Segmentation Data Saving it here makes it easier to find the algorithm from the Vectra software The file name should identify it as the Tissue Finder algorithm for the specific set of tissue slides Export all fields Use view settings P Overwrite existing files Rename new files Export for Export for Akeni JADE Figure 56 inForm Export Settings Create the HPF High Power Field Training algorithm 1 Create a new project selecting the Vectra HPF Finder option in the New Project Windowl 2 2 Open the first set of low power RGB color images from the LPF fullres Vectra directory 3 In the Prepare Images step specify the sample format brightfield or fluorescence Important Do not perform a conversion of the images to optical density Also never change the resolution of the images for Tissue Finder or HPF Finder projects The algorithm will not work with Vectra if the image resolution is changed 4 Inthe Segment Tissue step add the desired tissue categories draw training regions select the Segmentation Options and train the Tissue Segmenter Each of these steps is explained in detail in Trainable Tissue Segmentation 47 5 When you are satisfied with the HPF Finder algorithm export the images 6 Select File gt Save gt Algorithm to save the new HPF Finder algorithm in the Algorithms folder of the VectraData directory Saving i
14. U 100 whenever images are removed from or added to the project V Tissue White e Scale Views Based on Selected Images ot g 100 Select to show the brightness of components for each image all scaled relative to the selected images Bright components appear bright and dim oo options components appear dim The first time the option 7 Tissue Segmentation Map is selected all current images in the project are Z Nuclear Segmentation Map selected The scaling limits remain the same E Gopi Seon even if images are added to or removed from the en project Click the Reset button to open the RAR Scaling Image Selection Window 7 to change Wi a Ces the images selected to determine the scaling E Equalize Display Histogram limits Figure 44 View Editor Composite Image e Display Intensity Select True to display each component in the composite image at the original intensity For example if Component A is dimmer than Component B they display that way in the composite Select Adjustable to display each component in the composite image at the intensity selected on the slider e Display Color Select True Color to display each Displaying the Extracted Data component in the calculated color from the spectral shape of the component Select False Color to display each component in the selected false color Select the False Color using the color picker next to each component only visible when False
15. and membrane Exported data includes position shape and component signal statistics for each segmented cell See Segmenting Cells 59 for instructions This feature is available with inForm Cell Analysis or higher Score IHC or IF Use to quantify IHC or IF staining levels in segmented cells The tools provided are intended to automate visual assessment Positivity 0 3 10 bin 50 bin and Double Positivity and to provide histogram data of staining levels Exported IHC or IF data includes the score for each selected cell compartment See Scoring IHC or IF se for instructions This feature is available with inForm Cell Analysis or higher Objects B Manual Analysis Tab Count Objects Use to segment objects other than cells Exported data includes the position shape and component signal statistics for each segmented object See Counting Objects 71 for instructions This feature is available with inForm Cell Analysis or higher Export Use to select the export directory the images to export the data to export and the file name options See Exporting the Datal 89 for instructions The Manual Analysis tab on the left displays the settings for each step Use these panels to select the settings for the algorithm or project These settings can only be edited when Viewing a single image and not while viewing images in Gallery mode See also Image Display Areal 18 Understanding the inForm Work Area C Toolbar The t
16. and conditions of this License each Contributor hereby grants to You a perpetual w orldw ide non exclusive no charge royalty free irrevocable copyright license to reproduce prepare Derivative Works of publicly display publicly perform sublicense and distribute the Work and such Derivative Works in Source or Object form 3 Grant of Patent License Subject to the terms and conditions of this License each Contributor hereby grants to You a perpetual w orldw ide non exclusive no charge royalty free irrevocable except as stated in this section patent license to make have made use offer to sell sell import and otherw ise transfer the Work where such license applies only to those patent claims licensable by such Contributor that are necessarily infringed by their Contribution s alone or by Software EULA 103 combination of their Contribution s w ith the Work to w hich such Contribution s w as submitted If You institute patent litigation against any entity including a cross claim or counterclaim in a law suit alleging that the Work or a Contribution incorporated within the Work constitutes direct or contributory patent infringement then any patent licenses granted to You under this License for that Work shall terminate as of the date such litigation is filed 4 Redistribution You may reproduce and distribute copies of the Work or Derivative Works thereof in any medium w ith or without modifications and in Source or Objec
17. areas where components are simultaneously expressing on a pixel level inForm identifies cells and finds nuclei membrane and cytoplasm quantitates the component signal within those cell compartments and calculates consistent scoring judgments based on that quantitation inForm identifies non cellular objects and can provide quantitative statistics inForm can automatically identify tissue types based on training regions drawn by the user Features can be combined to perform more advanced analysis such as computing the double positivity of cells within a cancer region or computing the colocalization of a multi stained tissue sample 26 inForm User Help 5 1 Working with Projects and Algorithms An inForm project includes the images in the project the algorithm to process the images and any processing regions training regions or manual tissue regions The topics in this section describe how to create open and save projects how to change the project settings and how to open and save an algorithm An algorithm is used in batch mode to process images the same way each time A project includes all the images so that you can go back and modify algorithm settings including the trainable tissue classifier Creating Projects Each time inForm starts a new project is created automatically The new project contains only the Prepare Images step and the Export step To create a project with additional steps either Create a N
18. drag to identify regions of the image that contain the selected tissue category Only available during the Manual Classification step Q Select Regions to Unclassify Click and drag to outline the area on the image to unclassify Only available during the Manual Classification step Select Pixels to Unclassify Click and drag over the specific pixels to remove from the classification region Only available during the Manual Classification step White Picker Click on a white area on a Brightfield image to select the white background spectra prior to converting the image to OD optical density Available only N A 16 inForm User Help during the Prepare Images step Add to Thresholded Regions Click the button to select the desired unmixed Z component and then click and drag to outline the area on the image to be added to the selected threshold map Only available during the Threshold step Erase Thresholded Regions Click the button to select the desired unmixed Z component and then click and drag to outline the region to remove from the threshold map Only available during the Threshold step Edit the Tissue Segmentation Click the button to select the tissue 7 category and then click and drag to manually draw the selected tissue category over an area After segmenting the images inspect the images to Tumor make sure the tissue regions are segmented as you intended If areas are Other under the wrong masks the t
19. for component 2 Single lt Component2 gt The percent of cells that is positive for component 2 and negative for a2 inForm User Help component 1 e Double Positive The percent of cells that is positive for both markers e lt Componenti gt Threshold The threshold used for component 1 e lt Component2 gt Threshold The threshold used for component 2 The following additional columns are visible when 0 3 4 bin scoring is selected 0 The percentage of cells whose component strength is less than the first threshold 1 The percentage of cells whose component strength is greater than or equal to the 0 1 threshold and less than the 1 2 threshold 2 The percentage of cells whose component strength is greater than or equal to the 1 2 threshold and less than the 2 3 threshold 3 The percentage of cells whose component strength is greater than or equal to the 2 3 threshold H score The calculated H Score of the image The H Score is calculated using the percentages in each bin and ranges from 0 to 300 Threshold 0 14 Clicking the Auto button sets this value to 1 4 of the Threshold max This value can be changed Threshold 1 2 Clicking the Auto button sets this value to 1 2 of the Threshold max This value can be changed Threshold 2 3 Clicking the Auto button sets this value to 3 4 of the Threshold max This value can be changed The following additional columns are visible when 10 bin or 50 bin sco
20. for Exposure reports the number of counts in the image after dividing by a factor that accounts for all the acquisition settings in use gain binning bit depth and exposure time Using Counts Normalized for Exposure allows you to compare images taken using different camera settings or exposures with the assurance that signal levels indicate the actual brightness at the sample regardless of the instrument settings Counts Normalized for Exposure counts 2 it depth x exposure time x gain x binning area where bit depth is the bit depth of the imagery typically 8 bit 256 levels of gray or 12 bit exposure time is in seconds gain is the gain setting of the camera and binning area is 1 for 1x1 4 for 2x2 etc The values are small when using Mean weighting and it may be hard to develop an intuition about what is normal Selecting Total weighting default produces larger values For brightfield images that are converted to Optical Density inForm produces unmixed component signals in OD Optical Density units Optical Density indicates the amount of absorbing material present at a location Optical Density log10 pixel value white reference pixel value Weighting Specifies whether to use the Mean Peak or Total signal to calculate the displayed units Total The sum of all signals of the selected component across all wavelengths Available for all images except images converted to Optical Density Default for fluorescence ima
21. image display area in the Manual Analysis tab has two modes Single and Gallery Single mode displays one image in the large Image area and displays a row of thumbnails of all other open images below the large image Click any thumbnail to view the image in the display area A blue square on a thumbnail indicates that the image is included in the training set a green square indicates the image is in the processing set The Algorithm settings in the left hand panel can only be edited in Single mode Gallery mode displays all the project images in the large Image area Use the Image Size slider at the top of the workspace to adjust the number of open images shown in the image display area In Gallery mode all Algorithm settings are read only Gallery mode is for image and table review only you cannot edit the Algorithm settings in this mode Right Click Shortcut Menu The following actions are accessible by right clicking on an image in the Image Display area Save Saves a copy of the image as shown in the Image Display Area to a file in the selected format tif jpg png gif or bmp Default is the last selected image format Delete Region Deletes the region that you right clicked on Only available if you right clicked on a region Delete Regions Provides options to delete all training regions all processing regions or both from an image Only available if training regions and or processing regions have been creat
22. in the upper right corner to close the Spectral Library window lf a chart cannot be displayed because no spectra are selected or there are no compatible units a message displays in place of the spectral charts 8 4 Component Units The components produced by unmixing are calculated Select Reported Units Weighting using the settings selected in the Select Reported Units and Weighting drop down lists The Reported Units can 2 be Raw counts counts Normalized for Exposure or OD Raw optical density units The Weighting can be Total Mean Normalized for Exposure or Peak Brightfield versus Fluorescence Brightfield and Fluorescence images use different measuring schemes because of the inherent differences of the images e For Brightfield images darker pixels indicate a higher expression of the biomarker e For Fluorescence images brighter pixels indicate a higher expression of a biomarker or reporter protein Therefore we use optical density for brightfield images and fluorescence counts either raw or normalized for fluorescence images Preparing Images s Select Reported Units Monochrome black and white images and RGB color images produce signals in counts For multispectral images that are not converted to optical density component signal counts are either raw or normalized for exposure Raw Counts reports the measured signal associated with a given component such as DAPI FITC etc Normalized
23. inspection of the results so that cytoplasm segmentation approximately matches the average size of cells This distance can be set large enough to include membrane signals if desired In this case it is important to note that the signal extracted from the segmenter for cytoplasm is a mixture of membrane and cytoplasm signal gt Minimum Size The minimum cytoplasm sample size in pixels If cytoplasm segmentations for detected nuclei have fewer pixels than this number those cells are excluded from the analysis This is common in cases where cells are tightly packed together without much space between nuclei for cytoplasm areas to be segmented If the cytoplasm is clipped at the edge of a Process region or Tissue Category region only the cytoplasm inside the region is used to determine the size of the cytoplasm Segmenting Cells e Selecting Components to Include You can further restrict what you consider valid cytoplasm based on the underlying signal You can pick an individual component signal as the primary and then select secondary and tertiary components if desired and enter a signal range for each one Or select Every Component or Any component and specify a signal range Only pixels that fall within that signal range are counted as cytoplasm The signal range varies based on image type 13 3 Segmenting Membrane Cytoplasm Pixel Validation Component Valid Signal Range Min Max 4 Primary oD Her2 Fast
24. loading a project only the algorithm settings are used The images from the project are not loaded or processed 4 lf desired select the Create separate directories for each item image option to save the exported files in a separate folder for each of the images or slides in the data set This is recommended for large batch processes that result in many data files and images 5 Select the desired export settings to export images component images maps and data tables Export settings are explained in detail in Exporting the Datal 84 6 Click the Add Images button to select images to process or click the Add Slides button to select single slides or folders containing multiple slides A slide is a scan from the Vectra software If a slide has been reviewed with Vectra Review the batch run only includes the accepted fields If the slide has not been reviewed all fields are included in the batch run 7 If necessary use the Remove Selected button to remove selected images or slides from the batch or use the Remove All button to remove all of the images or slides from the batch 8 Click the Run button inForm processes the items one at a time and exports the specified data 9 When the image processing is complete click the Done button in the Batch Progress window o2 inForm User Help 19 Merging the Data Use the Review Merge tab to combine exported data files from multiple images or slides Each type of data file from a batch
25. patterns that are measured to support machine learning and image analysis If the selected pattern scale e g Maximum is too large for the drawn training regions a grey mask displays on training regions that are too small after training and those training regions are not used to train the tissue segmenter Trainable Tissue Segmentation e You should use as large a pattern scale as is compatible with the training regions that you draw If most of the training regions have gray masks over them after training reduce the pattern scale e g from Large to Medium and then retrain For example if analyzing tumors in tissue sections in a 20x magnification image the architecture is fundamentally on a large scale and you probably want to select Large For items with inherently fine scale structures select Small or Medium An understanding of the different scales can be learned by seeing what size training regions are masked with grey after training Train the Segmenter Click the Train Tissue Segmenter button to begin training the segmenter The training process may take some time so be patient The percent accuracy of the new tissue segmenter displays If for example the segmenter was 99 accurate this would mean that if the new segmenter was to be applied to all of the images used in training the segmenter it would classify the pixels in the training regions with 99 accuracy Fewer than 1 of the pixels in all the drawn regions
26. pixel in the component The slider range is set to 0 Threshold Max The threshold is set to an optimal value based on the component values The threshold max and threshold can be set manually without clicking the Auto button if desired 4 Set the desired Threshold for each unmixed component Move the slider type the desired value or click the up and down arrow buttons to select the threshold The image display shows the areas that are above the threshold value as the slider is adjusted 5 To change the color displayed for the regions click the Color button and select the desired color 6 To change the minimum size of the thresholded regions e Click the More button under the name of the Threshold map that you want to change e Set the desired value for Minimum Connected Pixels by either clicking the Up or Down arrow buttons or typing the value in the text box The Minimum Connected Pixels specifies a minimum region size based on the number of pixels Only regions that are larger than the Minimum Connected Pixels value are considered Regions with fewer than this number of pixels are ignored 7 lf there are multiple images in the project either select each image and apply the current settings to the image by clicking the Accept for Image button or click the Accept for All button to apply the Threshold settings to all of the images in the project Verify that the threshold settings are appropriate for all images in the project and ad
27. replacement of the media on which the Softw are was provided so long as that media has been returned to CRI under a CRlissued return authorization CRI shall have no responsibility to replace media damaged by accident abuse or misapplication 6 No Other Warranties EXCEPT FOR THE LIMITED WARRANTY STATED IMMEDIATELY ABOVE THE SOFTWARE IS PROVIDED AS IS WITHOUT WARRANTY OF ANY KIND AND CRI EXPRESSLY DISCLAIMS ANY AND ALL IMPLIED WARRANTIES INCLUDING WITHOUT LIMITATION ANY WARRANTY OF MERCHANTABILITY FITNESS FOR A PARTICULAR PURPOSE OR NON INFRINGEMENT OF THIRD PARTY RIGHTS CRI DOES NOT REPRESENT OR WARRANT THAT THE RESULTS OR THE USE OF THE SOFTWARE WILL BE CORRECT ACCURATE OR RELIABLE OR THAT THE SOFTWARE WILL OPERATE UNINTERRUPTED OR ERROR FREE OR THAT DEFECTS IN THE SOFTWARE WILL BE CORRECTED YOU ASSUME ALL RISK ASSOCIATED WITH THE USE RESULTS AND PERFORMANCE OF THE SOFTWARE 7 Limitation of Liability INNO EVENT SHALL CRI ITS AFFILIATES OR SUPPLIERS OR THEIR RESPECTIVE EMPLOYEES OFFICERS OR AGENTS BE LIABLE FOR ANY DAMAGES ARISING OUT OF THE USE OR INABILITY TO USE THE SOFTWARE INCLUDING WITHOUT LIMITATION INCIDENTAL SPECIAL CONSEQUENTIAL PUNITIVE EXEMPLARY INDIRECT OR DIRECT DAMAGES INCLUDING WITHOUT LIMITATION DAMAGES FOR LOSS OF PROFITS LOSS OF DATA 100 inForm User Help RE RUN TIME INACCURATE INPUT WORK DELAYS BUSINESS INTERRUPTION OR ANY OTHER COMMERCIAL DAMAGES OR LOSSES WHETHER IN AN ACTION IN CONTRACT T
28. select the desired parameters See Segmenting Nucleil 521 for descriptions of the parameters 4 lf Cytoplasm is selected click the Cytoplasm tab and select the desired parameters See Segmenting CytoplasmlI 63 for descriptions of the parameters 5 If Membrane is selected click the Membrane tab and select the desired parameters See Segmenting Membrane 64 for descriptions of the parameters 6 Click the Segment Image button to segment the current image using the selected parameters 7 f desired change the parameters to obtain the desired results 8 Click the Segment All button to segment all of the images 13 1 Segmenting Nuclei At the top of the Cell Segmentation Settings panel select Nuclei in the Compartments to Segment box You can also select Cytoplasm and Membrane if you intend to find those as well Click the Nuclei tab Segmenting Cells s Selecting a Tissue Category Select the Tissue Category in which to find the cell nuclei or select All Categories to find cell nuclei in all tissue categories Cells that are outside the chosen category are ignored This option only displays if tissue has been manually or automatically segmented Tissue categories of interest might be tumor regions in oncology samples and is et cells in pancreas samples Selecting an Approach e Select the Pixel Based Threshold approach when there is a reliable nuclear counterstain and nuclear pixels can be found by applying a s
29. tasks using realistic scenarios The PDF file can be read online or can be printed for reference inForm also includes complete documentation in an HTML based help system which includes all of the information in this Users Manual Online help provides three ways of locating information Use the Contents tab to navigate through the document the Index tab to find topics by keyword or the Search tab to look up topics that contain specific words or phrases Installing and Starting inForm Close all applications that are running before beginning the inForm installation Installation requires an administrator account Starting the Install from CD DVD or USB Drive Insert the inForm CD DVD or USB drive If AutoRun is turned on the installation wizard starts automatically If the installation does not start automatically in Windows Explorer navigate to the install media and double click inFormSetup exe Starting the Install from a Network or Downloading If you are installing inForm from a network location download and extract the inForm installation folder to a local drive Double click on inFormSetup exe to start the installation Installing the Software Follow the wizard prompts to install the software inForm requires the Microsoft NET 4 5 framework The installer will install Microsoft NET 4 5 if it is not already installed The installer asks if you will be using inForm to create Vectra procedures to analyze Vectra data Be s
30. the components The relative brightness of components cannot be compared between images Scale Views Equally for All Images in the Project Select to show the brightness of components for each image all scaled relative to each other Bright components appear bright and dim components appear dim All images are rescaled relative to all images in the project whenever images are removed from or added to the project a i View Editor all Data Displayed Component Her2 Fast Red x Rendering Options Show As Brightfield Fluorescence Component Color Black and White Color Scaling Scale Views for Each Image Individually Scale Views Equally for All Images in the Project Scale Views Based on Selected Images Adjust Brightness J Contrast p Reset to Default Image Options V Tissue Segmentation Map V Nuclear Segmentation Map 7 Cytoplasm Segmentation Map V Membrane Segmentation Map Show Score Colors Training Regions IS 8 Processing Regions Equalize Display Histogram g Figure 42 View Editor Color Image Scale Views Based on Selected Images Select to show the brightness of components for each image based on the brightness range of the selected images Bright components appear bright and dim components appear dim The first time the option is selected all current images in the project are selected The scaling limits remain the same even if im
31. the name of the Threshold map that you want to edit 2 To draw additional threshold regions click the Add button under the component name and draw the desired regions on the threshold map 3 To remove threshold regions click the Erase button and draw around the area to be removed Only the threshold areas for the selected component are erased 4 Toclear all manual edits from a threshold mask click the Clear button under the name of the component Only the manual edits on the threshold map for that component are cleared To clear all manual edits from all threshold maps 1 Click the down arrow next to the Erase Threshold Regions button and click Clear Edits om F Red Green X 2 Click Yes in the Clear Mask Edits window s inForm User Help 12 Colocalization Colocalization is a pixel based analysis used to find overlapping components in images based on thresholds set for each component Colocalization of multiple markers can be determined and displayed both visually and statistically It is designed for analyzing and quantitating molecular markers and can be used to determine the amounts of colocalization of multiple markers 1 Click the Colocalization button in the step bar D Colocalize Under Markers for Colocalization select the markers for which you want to see overlap colocalization and positivity percentages Under Denominator Counterstain select the marker s to use as th
32. the two axes of the minimum area bounding box enclosing the region Axis Ratio The ratio of the major axis minor axis of the minimum area bounding box enclosing the region inForm User Help 17 Exporting the Data After analyzing the images use the Export step to export images component images maps and data tables Images are exported in the format selected in the Export Settings step Data tables are exported to tab separated text files which can be opened in Microsoft Excel and many other data analysis programs 1 Click the Export button to display the Export Settings panel Export Click the Browse button to select or create the desired Export Directory Select the Image Output Format you want to export images to either JPEG or TIFF Generally TIFFs retain more resolution and are usually larger JPEGs are compressed slightly and are often better for emailing and publishing on the web From the Images to Export list select the images and maps to export Images can be exported with or without the maps Exporting the images with maps can be used with Review Merge to view the segmentation for all images in a batch Select the Component Images multi image TIFF option to save a multi image TIFF file of component data for analysis using third party analysis software The values in the Component images are in the units selected in the Prepare Images step Components are saved in the spectral library
33. your accepting any such warranty or additional liability 16 This software uses the bzip2 program and libbizip2 libraries http www bzip org This softw are makes use of the bzip2 program and associated libbzip2 libraries The program and libraries have not been modified by CRI and all rights are reserved by the copyright holder This softw are is provided as is without express or implied w arranty and with no claim as to its suitability for any purpose The libraries are covered by the follow ing license 104 inForm User Help The bzip2 license is reproduced below This program bzip2 the associated library libbzip2 and all documentation are copyright C 1996 2006 Julian R Seward All rights reserved Redistribution and use in source and binary forms with or without modification are permitted provided that the follow ing conditions are met 1 Redistributions of source code must retain the above copyright notice this list of conditions and the follow ing disclaimer 2 The origin of this softw are must not be misrepresented you must not claim that you w rote the original softw are If you use this softw are in a product an acknow ledgment in the product documentation w ould be appreciated but is not required 3 Altered source versions must be plainly marked as such and must not be misrepresented as being the original software 4 The name of the author may not be used to endorse or promote products derived fro
34. 1991 1997 Silicon Graphics Inc Permission to use copy modify distribute and sell this softw are and its documentation for any purpose is hereby granted without fee provided that i the above copyright notices and this permission notice appear in all copies of the softw are and related documentation and ii the names of Sam Leffler and Silicon Graphics may not be used in any advertising or publicity relating to the softw are without the specific prior written permission of Sam Leffler and Silicon Graphics THE SOFTWARE IS PROVIDED AS IS AND WITHOUT WARRANTY OF ANY KIND EXPRESS IMPLIED OR OTHERWISE INCLUDING WITHOUT LIMITATION ANY WARRANTY OF MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE IN NO EVENT SHALL SAM LEFFLER OR SILICON GRAPHICS BE LIABLE FOR ANY SPECIAL INCIDENTAL INDIRECT OR CONSEQUENTIAL DAMAGES OF ANY KIND OR ANY DAMAGES WHATSOEVER RESULTING FROM LOSS OF USE DATA OR PROFITS WHETHER OR NOT ADVISED OF THE POSSIBILITY OF DAMAGE AND ON ANY THEORY OF LIABILITY ARISING OUT OF OR IN CONNECTION WITH THE USE OR PERFORMANCE OF THIS SOFTWARE 11 This software uses the Vigra 1 6 0 library http hci iwr uni heidelberg de vigra This software includes machine executable object code generated by a source language processor from the Vigra libraries covered by the VIGRA license These libraries have not been modified by CRI and all rights are reserved by the copyright holder This software is provided as is without expr
35. AEE 73 2 Viewing the Component IMaAgE ssenarisinin saian a a a A R 74 3 Viewing the Composite IMag esssernsiieeieies inneni iE E E N N 77 4 Viewing the Tissue Segmentation Data Table ec ceeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeaaeeeeeeeaeeeeseaeeeeeeeaes 78 5 Viewing the Cell Segmentation Data Table cece eeccceeeeeeeaneeeeee nate eeeeaaeeeeeeaaeeeeeaaeeeeeaaeeeees 80 6 Viewing the Score Data Table 0 ccc eece cece eeee cena eeeeeeae eee sean eee eeaaaeeeeeaaeeeeeeaaeeeeeeaaeeeeeeaaeeeeees 82 7 Viewing the Colocalization Data Table ececcceeeceeeneeeeeeeaa eter ee aa eee sean eeeeaaeeeeeeaaeeeeeeaaeeeeees 84 8 Viewing the Count Data Table esien EAEE A AEA A AA 85 9 Viewing the Quant Data Table eneee E E EEEE EEEE EAEE E 86 10 IMAGES OPTIONS eeen aa E E E SE 86 11 Component SUAS seeder iE E EaR E RE S EEE EEEE a 87 12 TMA CUG IMiOcorairee iii TAE EE E EO EE E ET ERA 87 13 POSITION Stats ea EE E E EE EE S 87 T4 Shape Sals airceas n A AEE E EEEN N E E E A E 88 4 inForm User Help Chapter 18 Batch ProceSSing ssssssseeeccceeeeesseneeeeeeeessneneeeeeneensneneaaeseeennnseaaeaaeens 91 Chapter 19 Merging the Dattar sssssssesssssssscsserserssssesssssnsessssnnsnesesncnscorsnsansaessnseenoess 93 Chapter 20 Creating Algorithms for V Ctrad cccceceeeeeeeeeeeeetessnnnnnnnneeneeeseeeeeeeeeneenens 94 Chapter 21 Appendix A Calculating Fractional Tissue Area sseseeeeee
36. Color is selected Component check boxes Select to display each component If not selected the component does not display The Intensity Slider for each component only displays if the Display Intensity is Adjustable The Color Selector for each component only displays if the Display Color is False Color Reset to Default button Click to reset the Rendering Options back to the default settings Image Options See Image Options s 16 4 Viewing the Tissue Segmentation Data Table The Tissue Segmentation Data Table is available when an algorithm contains a manual tissue segmentation or trainable tissue segmentation step The table provides information on the manually or automatically segmented tissue regions Select Tissue Segmentation Data in the Data Displayed list then choose the desired options under Table Contents Each option you select adds one or more data columns or rows to the table During image analysis segmented tissue categories are assigned ID numbers These numbers are used in the data tables to identify the tissue category for stats such as area pixels area percent etc inForm User Help The following options are available for the Tissue Segmentation Data Table Components Select all or individual components for which to display component statistics Each selected component creates one column for each component stat selected Tissue Categories Select all or individual tissue categories for which t
37. Displayed list then choose the desired options under Table Contents Each option you select adds one or more data columns or rows to the table The following options are available for the Count Data Table OES e Data Displayed Components Select all or individual components for Count Data Z o which to display component statistics Each selected ENER component creates one column for each component stat components Tissue Categories selected E All Components E All Categories V Her2 Fast Red V Category Tissue Categories Select all or individual tissue a E Category2 categories for which to display data If selected a eS Tissue Category column displays in the Count Data table cee ae to specify the Tissue Category for each cell This om E X Postion selection does not display if there is only one Tissue 7 Mean E Y Postion Category defined or if there is no Segment Tissue step Ei Max 2 in the project E Std Dev L Distance from Process Region Edge E Total Category Region ID Component Stats Select the values to display for each ee ee component signal See Component Stats 8 for eS para descriptions of each option F Area percent TMA Core Info Select the options to include in the i cars data table See TMA Core Infol s for descriptions of E Major Axis each option F Axis Ratio Position Stats Select which values to display for each segmented object See Position Stat
38. Load Image button in the Available Images Windowl 27 Removing Images from a Project To remove images from the open project In Single mode click on the thumbnail of the image to be removed to show the image in the Image Display Area Right click on the image in the image display area and select Remove from Project In Gallery Mode right click on an image in the gallery and choose Remove from Project Images can also be removed from the project by right clicking on the image name in the Available Images Windowl 27 and selecting Remove Image 6 2 Viewing Images Images are viewed in the Image Display area in the inForm Window View images in Single Mode to zoom in on an image draw regions on an image or edit the algorithm settings View images in Gallery Mode to compare all the images in the project or to quickly locate specific images Gallery Mode To view a gallery of images click the Gallery button at the lower left of the display area Note that when in Gallery mode you cannot edit the Step settings E Adjust the display by moving the Image Size slider at the top of the Image Display Area Move the slider to the left to fit more images in the display area Move the slider to the right to fit fewer images in the display area 32 inForm User Help E inform 20 Manual Arabs NewPr File Edit Views Tools License Help Image Preparation Settings E g S 78 Image Fomat Mati
39. MPs JPGs etc 2 Navigate to the location of the image files 3 Select the desired image files The image files must all be of the same image type im4 im3 RGB or Monochrome Ctrl click to select multiple image files Shift click to select a range of files 4 Click the Open button to add the images to the project The images are automatically included in the Processing setl 101 5 To change the view of the images see Viewing Images 32 6 To process only part of an image see Drawing Processing Regions 34 7 To change the image resolution convert to optical density or unmix the images see Preparing Images Eg Opening Images Using the Available Images Window You can add images to the project using the Available Images window 1 Click the Edit the list of images button on the button bar The Available Images Window 21 opens 2 Click the Load Image button The Select Image window opens 3 Select the desired images and click the Open button You can also use the Available Images window to select which images are in the training and processing sets by selecting or clearing the check boxes next to each image name Opening Images 31 Adding Additional Images to a Project After images have been added to a project only files of the same type Multispectral TRIO RGB or Monochrome can be added to the project Add images to an open project either by selecting File gt Open gt Image or by clicking the
40. Membrane value specifies the maximum size of the cells defined here as the distance in pixels from the edge of the nucleus to the outer edge of the membrane Distance to Membrane pixels f2 A value of 12 for images taken at 20x magnification is usually k a good starting value If the algorithm is finding a membrane that is either too large or too small based on your knowledge of the nucleus cell size then try reducing or increasing the Distance to Membrane value This value limits how far from the nucleus the algorithm is permitted to go when segmenting the cell membrane Maximum Cell Size Figure 35 Maximum Cell Size Segmenting Cells e 14 Scoring IHC or IF After segmenting nuclei cytoplasm and membrane the detected cell compartments are quantified The options in the Score IHC or IF Settings panel provide functionality similar to that which is done manually or visually when pathologists assess chromogenic immunohistochemical stain levels The Score IHC or IF settings specify the component thresholds and score the images 1 Tissue Category Select the tissue category to use for scoring Cells that are outside the chosen tissue category are ignored This option only displays if tissue has been segmented Only tissue categories in which cells have been segmented are available Scoring Type There are five Histogram score types which are used to assess the intensity of each label Select a score type from the
41. ORT INCLUDING NEGLIGENCE AND STRICT LIABILITY OR OTHERWISE AND EVEN IF CRI HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH DAMAGES AND REGARDLESS OF WHETHER ANY REMEDY FAILS OF ITS ESSENTIAL PURPOSE CRs AGGREGATE LIABILITY ARISING OUT OF OR RELATING TO THIS AGREEMENT FOR ANY AND ALL DAMAGES THAT YOU MIGHT INCUR REGARDLESS OF THE FORM OF ACTION GIVING RISE TO SUCH LIABILITY WHETHER IN CONTRACT TORT OR OTHERWISE SHALL NOT EXCEED ONE HUNDRED DOLLARS 100 Some jurisdictions do not allow the exclusion or limitation of liability for consequential or incidental damages In such jurisdictions our liability is limited to the greatest extent permitted by law or the amount you paid for your purchase whichever is less 8 U S Government End Users The Software qualifies as commercial computer softw are for purposes of the Federal Acquisition Regulations FAR 52 227 19 and the Department of Defense Supplement to the FAR DFARS 52 227 7013 If the Software is acquired by a civilian government agency it is furnished with only the minimum Restricted Rights provided by FAR 52 227 19 If the Software is acquired by a military agency it is furnished with only the minimum Restricted Rights provided by DFARS 52 227 7013 c 1 ii 9 Miscellaneous This Agreement contains the entire agreement of the parties with respect to the subject matter hereof and supersedes any proposal or prior agreement written or oral and any other communications betw een the pa
42. Remove Figure 3 Available Images Window Understanding the inForm Work Area 3 3 New Project Window Use the New Project Window to select the steps to include in the project Certain steps are only available if a software option has been purchased and installed For available software configurations see Software Configurations 8 lela jew r New Project New Project Prepare Images Segment Tissue Find Features Score Export Prepare the images for analysis Segment the images into different tissue Find features within the imagery using Score cells based on instensity thresholds Save the table data and imagery categories thresholds or segmentation algorithms Manual Tissue Segmentation Cell Segmentation Scoring requires Cell Segmentation Trainable Tissue Segmentation Object Segmentation Skip This Step Skip This Step Colocalization Threshold Skip This Step All projects must contain an Export step All projects must contain a Prepare Images step Figure 4 New Project Window The New Project window contains the following options and buttons Project Type These options only display in inForm Tissue Finder if you chose to enable Vectra algorithms when inForm was installed If these options are not displayed all new projects are Custom Projects e Custom Project option Creates an inForm project
43. SOFTWARE BE LIABLE FOR ANY DAMAGES OR OTHER LIABILITY WHETHER IN CONTRACT TORT OR OTHERWISE ARISING FROM OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE 13 This software uses the Quan 1 0 0 library https lists sourceforge net lists listinfo quan matters This software includes machine executable object code generated by a source language processor from the Quan libraries covered by the Boost license These libraries have not been modified by CRI and all rights are reserved by the copyright holder This software is provided as is without express or implied warranty and with no claim as to its suitability for any purpose 14 This software uses the CommandLine 1 6 0 0 library http commandline codeplex com This softw are makes use of the CommandLine libraries These libraries have not been modified by CRI and all rights are reserved by the copyright holder This softw are is provided as is without express or implied w arranty and with no claim as to its suitability for any purpose The libraries are covered by the MIT License The MIT License is reproduced below The MIT License MIT Copyright c 2005 2012 Giacomo Stelluti Scala Permission is hereby granted free of charge to any person obtaining a copy of this software and associated documentation files the Softw are to deal in the Softw are without restriction including without limitation the rights to use copy modify merge pu
44. Step Skip This Step Colocalzation Threshold Skip This Step All projects must contain a Prepare Images All projects must contain an Export step step Create Cancel Figure 6 New Project Window Working with Projects and Algorithms 2 Type aname for the new project in the New Project Name text box 3 To segment tissue select the desired option under Segment Tissue a Select Manual Tissue Segmentation to manually mark tissue categories on the images b Select Trainable Tissue Segmentation to train the tissue segmenter to automatically identify areas as specific tissue categories c Select Skip This Step to skip tissue segmentation 4 To find specific features select the desired option under Find Features a Select Cell Segmentation to identify cellular features such as finding the nucleus cytoplasm and membrane for each cell b Select Object Segmentation to identify objects of interest c Select Colocalization to identify areas on the images where selected components are colocated d Select Threshold to identify the areas on the images where the strength of a component at each pixel exceeds a specific threshold e Select Skip This Step to skip finding features 5 If Cell Segmentation was selected select Scoring to calculate score statistics for the cells 6 Click the Create button to create a new project with the selected steps 7 See Opening Images 3 to add the desired images to t
45. This table is Show Score Colors available when an algorithm contains a manual Elion tess segmentation step The table provides Eea ea information on the manually segmented tissue hne regions See Viewing the Tissue Fi 2 View Edi i t Segmentation Data Tablel 7 for detailed gure ices information Understanding the inForm Work Area Quant Data This table is available when an algorithm contains a threshold or colocalization step The table provides information on the thresholded regions See Viewing the Quant Datal s for detailed information Colocalization Data These tables are available when an algorithm contains a colocalization step See Viewing the Colocalization Datal 84 for detailed information There are three tables e Colocalization Table Shows colocalization data of the selected components and positivity data for each selected component e In Channel Table Shows the positivity of one component within another for all selected colocalization components e Quant Table Shows stats for the whole image and colocalization region Views Available In Cell Analysis Includes all views in Basic Analysis plus Cell Segmentation Data This table is available when an algorithm contains a cell segmentation step The table provides information on the segmented cells See Viewing the Cell Segmentation Datal 80 for detailed information Object Segmentation Data This table is available when an algorithm con
46. You can now use inForm to create the Tissue Finder and HPF Finder algorithms Note While 3 to 10 images is usually sufficient you may need more images to get a good segmentation of the target class across all slides if the target has wide variation Creating Algorithms for Vectra Create the Tissue Finder algorithm 1 Choose File gt New gt Project Choose the Vectra Tissue Finder option and click Create o inForm User Help Prepare Images Prepare the images for analysis Segment f categorie Figure 55 Vectra Tissue Finder project 2 Choose File gt Open gt Image and select the first group of monochrome images 3 Click Prepare All to prepare the images Important Do not perform a conversion of the images to optical density Also never change the resolution of the images for Tissue Finder or HPF Finder projects The algorithm will not work with Vectra if the image resolution is changed 4 Either click the Advance button at the bottom of the window or click the Segment Tissue button 5 In the Segment Tissue step two tissue categories are automatically created Target Tissue and Other 6 Draw training regions around tissue and around non tissue regions on each of the images using the appropriate tissue categories see Drawing Training Regions 481 7 Set the Segmentation Options as necessary for the current images see Training the Tissue Segmenter 491 for instructions In most cases yo
47. acted data These options do not change the underlying images they only change how the images are displayed The options below are available in the Data Displayed list depending on which software configuration is installed and which steps are included in the project e Viewing the Color Imagel 73 e Viewing the Component Image 74 e Viewing the Composite Imagel 77 e Viewing the Tissue Segmentation Data Tablel 78 e Viewing the Cell Segmentation Data Tablel a e Viewing the Score Data Tablel s e Viewing the Colocalization Data Tablel s e Viewing the Count Data Tablel 82 e Viewing the Quant Data Tablel 8 Viewing the Color Image Select Color Image in the Data Displayed list then choose the desired options The following options are available for the Color Image Rendering Options Brightness Change the value to change the brightness of the selected image Lighten or darken the image using the Brightness percentage slider Contrast Change the value to change the contrast between light and dark portions of the selected image Increase or decrease the contrast using the Contrast percentage slider Use Enhanced Contrast Select to artificially increase the vividness of the image Use Equal RGB Color Weighting only for Fluorescence images Select to distribute the wavelength range equally in the red green and blue components If not selected the image displays as it looks to the human eye Reset to Defau
48. ages are added to or removed from the project Click the Reset button to open the Scaling Image Selection Window 7e to change the images selected to determine the scaling limits See Scaling based on Selected Imagesl 751 for the suggested workflow When this option is selected the scaling limits are saved with the algorithm and project If the algorithm or project is used in Batch Mode then all images processed in the batch job use the saved scaling limits Brightness Change the value to change the brightness of the selected image Lighten or darken the image using the Brightness percentage slider inForm User Help Contrast Change the value to change the contrast between light and dark portions of the selected image Increase or decrease the contrast using the Contrast percentage slider Reset to Default button Click to reset the Rendering Options back to the default settings Image Options See Image Options s Scaling Based on Selected Images Images with bright signals in each component such as positive controls or positively expressing images should be used to scale all images in the project When using Scaling Views based on Selected Images the workflow below is suggested to set the desired scaling limits 1 2 3 4 a Load only the images of the positive controls for the images Process the images as desired Open the View Editor and select any component in the Data Displayed drop down list Selec
49. al results 10 1 Adding Tissue Categories Tissue categories divide the image into specific areas that correspond to specific types of tissue for example cancer necrosis stroma etc or occasionally other structures of interest The Segment Tissue step only available in inForm Tissue Finder enables you to draw training regions around representative areas of tissue on images and then train the tissue segmenter to automatically locate similar regions of tissue in each image The Tissue Segmentation Training panel is used to define tissue categories train segmenters and segment images in the image processing set Creating Categories 1 Click the Segment Tissue step to display the Tissue Segmentation Training panel Tissue Segmentation Training Tissue Categories Draw Red Wilke 2 Decide how many tissue categories you want to segment the images into Tumor 3 Click the New button and type the desired name of Category2 Green ike the tissue category in the text box 4 Create the desired number of tissue categories for the images in the project 5 To change the display color for a tissue category click the Color pull down and select the desired color Figure 19 Creating Tissue Categories new Remove _Gear_ ew Merge Batch Analysis Manual Analysis Deleting Categories To remove a category click the category s radio button and click the Remove button To clear
50. all tissue categories and start over click the Clear button Trainable Tissue Segmentation 10 2 Drawing Training Regions Training regions teach the tissue segmenter which cells should be included in each tissue category It is important that you already know what to look for and are able to visually recognize structures characteristic of the tissue category For example if there are tumor cells present in any image that are negative i e not stained be sure to include those also It is important to train on the full range of appearances of the tissue category as in the example where negatively stained as well as the positively stained cells are included To draw the training regions 1 Inthe Tissue Segmentation Training panel click the Draw radio button for a tissue category 2 Click and drag to draw regions around groups of cells For example if the tissue category is named Tumor draw regions around groups of tumor cells Tissue Segmentation Trainin j a h fae Tissue Categories Draw Tumor Components for Training 7 blue hematox DO Review Merge l Batch Analysis Manual Analysis Figure 20 Drawing Training Regions 3 To delete a region either click the Delete Region button and click inside the region to be deleted or right click on the region and select Delete Region from the pop up menu This menu also lets you delete all training regions o
51. ary forms with or without modification are permitted provided that the follow ing conditions are met Redistributions of source code must retain the above copyright notice this list of conditions and the follow ing disclaimer Redistributions in binary form must reproduce the above copyright notice this list of conditions and the follow ing disclaimer in the documentation and or other materials provided w ith the distribution Neither the name of the libjpeg turbo Project nor the names of its contributors may be used to endorse or promote products derived from this softw are without specific prior w ritten permission THIS SOFTWARE IS PROVIDED BY THE COPY RIGHT HOLDERS AND CONTRIBUTORS AS IS AND ANY EXPRESS OR IMPLIED WARRANTIES INCLUDING BUT NOT LIMITED TO THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED INNO EVENT SHALL THE COPY RIGHT HOLDERS OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT INDIRECT INCIDENTAL SPECIAL EXEMPLARY OR CONSEQUENTIAL DAMAGES INCLUDING BUT NOT LIMITED TO PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES LOSS OF USE DATA OR PROFITS OR BUSINESS INTERRUPTION HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY WHETHER IN CONTRACT STRICT LIABILITY OR TORT INCLUDING NEGLIGENCE OR OTHERWISE ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE Software EULA 105 The BSD for libjpeg license is reproduced below The aut
52. au commercial time sharing or other computer services to third parties iv transmit the Softw are or provide its functionality in whole or in part over the Internet or other netw ork except as expressly permitted v use the Softw are for any purpose other than solely for your internal research except as may be otherwise agreed in writing by CRI vi use the Software to provide any analytics or diagnostics or otherwise for the benefit of any third party except as may be otherwise agreed in writing by CRI or vii sell distribute rent lease sublicense or otherwise transfer the Softw are to a third party 3 Ownership of Software CRI and or its suppliers own all right title and interest including all copyrights trademarks tools know how and processes in and to the Software The Software contains confidential information and trade secrets of CRI You i acknow ledge and agree not to contest CRIs rights in the Softw are and ii agree not to disclose any confidential information of CRI regarding the Software or that is otherwise disclosed to you in connection with this Agreement You recognize that the covenants contained in this License are reasonable and necessary to protect the legitimate interests of CRI that CRI w ould not have entered into this Agreement in the absence of such covenants and that your breach or threatened breach of such covenants shall cause CRI irreparable harm and significant injury the amount of which shall be e
53. blish distribute sublicense and or sell copies of the Software and to permit persons to whom the Softw are is furnished to do so subject to the follow ing conditions The above copyright notice and this permission notice shall be included in all copies or substantial portions of the Softw are 102 inForm User Help THE SOFTWARE IS PROVIDED AS IS WITHOUT WARRANTY OF ANY KIND EXPRESS OR IMPLIED INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM DAMAGES OR OTHER LIABILITY WHETHER IN AN ACTION OF CONTRACT TORT OR OTHERWISE ARISING FROM OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE 15 This software uses the log4net libraries http logging apache org This softw are makes use of the log4net libraries These libraries have not been modified by CRI and all rights are reserved by the copyright holder This softw are is provided as is without express or implied warranty and with no claim as to its suitability for any purpose The libraries are covered by the Apache License The Apache License is reproduced below Apache License Version 2 0 January 2004 http w w w apache org licenses TERMS AND CONDITIONS FOR USE REPRODUCTION AND DISTRIBUTION 1 Definitions License shall mean the terms and conditions for use reproduction and distrib
54. bnail image at the bottom of the screen see figure indicates images are included in the training set A green square Trainable Tissue Segmentation 5 i indicates that the image is in the processing set See gt E r s A p fr 13036_8160_5100 In training set In processing set Figure 22 Training Set and Processing Set Indicators 6 See Drawing Training Regions 48 to add training regions to the desired images in the training set 10 5 Editing Masks and Re Training After reviewing the segmented images you might discover that some structures were not segmented correctly or the segmenter may have selected some unwanted regions The Tissue Segmentation masks should accurately cover their respective tissue categories To correct the tissue masks you can either edit the mask manually or you can re train the tissue segmenter to detect the desired regions Manually Editing the Tissue Segmentation Mask If the experiment or study involves few enough images you can manually edit the Tissue Segmentation mask Manual editing does not change the tissue segmenter and if the images are re segmented the changes to the Tissue Segmentation masks are lost To manually add areas to a Tissue Segmentation mask 1 Click the Edit Tissue Segmentation button on the toolbar 2 From the drop down list select the tissue segmentation mask that you want to edit Clear Em 3 Draw around the areas on the image that you want to
55. ch process images e Export data tables Cell Analysis inForm Cell Analysis includes all of the features available in inForm Basic Analysis plus the following features e Use cell segmentation to identify cells and cellular compartments e Score segmented cells e Use object segmentation to identify objects 3 inForm User Help Tissue Finder inForm Tissue Finder includes all of the features available in inForm Cell Analysis plus the following features e Train the Tissue Segmenter to automatically identify tissue categories e Automatically segment the image into tissue categories e Merge data files from a set of images or slides into summary data files Steps Available in Each Software Configuration The table below specifies which algorithm steps are available in each software configuration Manual Tissue Segmentation Trainable Tissue eae Threshold o e Score only used with Cell Segmentation Batch Processing is available with inForm Basic Analysis or higher Review Merge is available with inForm Tissue Finder Introduction 9 2 3 Key Terms The following terms describe components of inForm You should understand the terms below when using the software Processing Region Processing Set Tissue Category Tissue Segmenter Algorithms include all steps in processing an image from image preparation tissue segmentation and or cell segmentation object counting or scoring to export
56. cts smaller than a specific number of pixels increase the value in the Minimum Size pixels text box To exclude objects larger than a specific number of pixels select the Maximum Size pixels check box and specify the largest object size in pixels To automatically fill holes in objects select Fill Holes To limit the size of the holes that are automatically filled select the Max Hole Size check box and specify the size in pixels of the Counting Objects largest holes that should be filled 11 To detect objects that touch other objects as individual objects instead of as one object select the Refine Splitting check box The Maximum Size pixels check box must be selected to enable this option 12 To exclude objects based on the roundness of the object select the Roundness check box and specify the desired Minimum Circularity Zero returns any shape Higher decimal values up to 1 restrict selections to more round objects only 13 To exclude any objects that are touching the edge of the image process region or tissue region select the Discard Object if Touching an Edge check box 14 After processing use the View Editor 73 to view the resulting data tables You can also export the Count Data table by advancing to the Export step Exporting the Datal sd inForm User Help 16 16 1 Displaying the Extracted Data After analyzing the images use the View Editorl 141 to customize the view of the images and extr
57. d Scale option the range for each component is from 0 to 2 for OD images 0 to 255 for 8 bit images and 0 to the Fixed Scale value for all other images Also whenever you switch between auto scale and fixed scale the values are reset to zero Only pixels above the minimum signal value are counted as nuclei Using secondary and tertiary components is primarily useful in brightfield IHC applications where nuclear IHC stains are dark or intense and substantially block the Hematoxylin counterstain In these cases performing nuclear segmentation on the logical intersection of the Hematoxylin and nuclear IHC signals is more effective This approach does not work if the dark IHC stains reside outside the nucleus Specifying a Nuclei Size Range Set the Minimum Size If too many small unwanted structures e g not nuclei are found during segmentation try increasing the minimum size value Any nuclei that have fewer pixels than this number are not segmented and no associated data is collected If a nucleus is clipped at the edge of a Process region or Tissue Category region the original size of the nucleus is used to determine if it meets the minimum size threshold the clipped size is reported in the tables For images taken at 20x magnification 100 is a good starting value If appropriate set the Maximum Size This is required if you intend to select the Refine Splitting clean up option discussed below This sets an upper
58. data set or click the No button to exclude the data set e f desired use the Merge hot keys listed on the bottom left of the window to review the exported images for each data set 7 When all data sets have been reviewed click the Merge button Type a file name prefix to identify the new files as the merged files The merged data files are saved as tab separated text files in the batch directory with the specified file name prefix followed by the name of the data table Data from any excluded data sets are included in the lt prefix gt _rejected_ files Merging the Data s 20 Creating Algorithms for Vectra Customers who have purchased PerkinElmer s Vectra Intelligent Slide Analysis System use inForm to create Tissue Finder and High Powered Field HPF Finder algorithms for use with Vectra slide scanning protocols The Tissue Finder algorithm differentiates between tissue and non tissue regions The HPF Finder algorithm finds specific targets such as tumor cells inflammation arterial plaque etc To fully utilize the imaging and tissue classification capabilities of the Vectra system both of these algorithms need to be created for each unique set of tissue slides and paired with their corresponding Vectra protocols Selecting and Scanning a Slide Set for Algorithm Training Before creating algorithms for a Vectra slide scanning protocol use the Vectra Slide Analysis System to scan some slides selected from the slide set Se
59. e counterstains To see colocalization of markers throughout the entire image and not just within a particular counterstain select All Image Pixels For each component selected in the Markers or Denominator a Threshold settings area displays below the Denominator check boxes To adjust the Threshold for each component a Click the Auto button to calculate a preliminary value for the Threshold Max and to choose a preliminary Threshold value The Threshold Max is set to the value of the brightest pixel in the component The slider range is set to 0 Threshold Max The threshold is set to an optimal value based on the component values The threshold max and threshold can be set manually without clicking the Auto button if desired b Set the desired Threshold for each unmixed component Move the slider type the desired value or click the up and down arrow buttons to set the threshold The image display shows the areas that are above the threshold value as the slider is adjusted c To change the color of a threshold mask click the Color button and select the desired color Colocalization Settings Colocalization Coe 7 Visible Markers for Colocalization J Her2 Fast Red ER DAB J Hematoxytn Denominator Courterstain Her Fast Red ER DAB Hematoxyin J Ab image Pueis Her2 Fast Red Threshold Max 0 25 Ato 4 Q 0 0100 00 0 25 ed Color Y Viste More gt gt Hematoxylin Threshold Max 0 20 s Ato
60. ected Cell segmentation may take considerably longer when using this function If you do not select Refine Splitting nuclei that touch are detected as a single nucleus e Select Roundness Minimum circularity to find only nuclei that are more round Zero returns any shape Higher decimal values up to 1 restrict selections to more round objects only 62 inForm User Help 13 2 Segmenting Cytoplasm At the top of the Cell Segmentation Settings panel select the Cytoplasm check box in the Compartments to Segment box Note that you must segment the nuclei in addition to segmenting cytoplasm inForm only searches for cytoplasm where nuclei have been found Click the Cytoplasm tab Selecting Cytoplasm Shape Parameters gt Inner distance to nucleus The distance in pixels from the edge of nuclear segmentations to the inner edge of the cytoplasm annular regions drawn around nuclei This Nuclei Cytoplasm Membrane Cytoplasm Shape buffer distance should be large enough to ensure that eer P imperfections in the nuclear segmentation do not lead to Outer distance to nucleus pixels 6 signal from nuclei inadvertently becoming part of Minimum Size pixels cytoplasm signals Figure 31 Cell Segmentation gt Outer distance to nucleus The distance in pixels from Settings for Cytoplasm the edge of the nuclear segmentation to the outer boundary of cytoplasm segmentation This distance should be adjusted based on visual
61. ed on the image Show Regions Provides options to show hide all training regions processing regions or both on an image Processing Set Adds or removes the image from the processing set If checked the image is included in the processing set If cleared the image is not included in the processing set Click on the option to toggle the check mark Training Set Adds or removes the image from the training set If checked the image is included in the training set If cleared the image is not included in the training set Click on the option to toggle the check mark Only displays if training images have been drawn on the image Draw training regions on the image or use the Available Images Windowl 24 to add images without training regions to the training set View Editor Opens the View Editorl 19 to select the display options Remove from Project Removes the image from the project te inForm User Help E Batch Analysis Tab After a suitable project or algorithm for processing images has been created validated and saved use the Batch Analysis tab to process large numbers of images Use the Batch Analysis Editor 9 to select a project or an algorithm for batch processing select export directory folder preferences and select the image table options This feature is available with inForm Basic Analysis or higher F Review Merge Tab After completing a batch run which often generates large numbers of output fil
62. eeeeeneees 97 Chapter 22 Software EULA s s ssssssssrununsenrnnnnnnnnnnnnnununnnnnnnnnnnnnnnnnnnnnnnnn nennu nenna 100 UN assests aces cee hat eee cee tet cere T 107 Contents 5 1 Welcome to inForm PerkinElmer s inForm software has been developed to address the biggest challenge facing microscopists once quantitative imagery has been acquired extracting data The challenge increases when dealing with multiplexed assays since signals can be weak and highly mixed inForm provides a powerful and fast solution with an intuitive easy to use software interface What might have been impossible or have taken days can now be done in minutes inForm is an advanced image analysis package that solves challenging image analysis problems by combining fast and easy learn by example automated image processing with object recognition and data analysis tools It is based on machine learning which means that you can train the program to create an effective solution by simply drawing around examples of what you want segmented within the images inForm then creates an algorithm that you can subsequently apply across the entire image or across as many images as you want PerkinElmer s multispectral images and industry standard RGB images including TIFF files generated using PerkinElmer s Slide Imaging systems can be read and analyzed by inForm inForm is for research use only A breakthrough in image understanding and quantitation that extracts
63. es inform 20 Mansal Analysis inFomeWerires i File Edit Views Tools License Help a OFr Rhen ib ct Specro for Urweang F Herd fan Pad aa coad i A e O l l Figure 1 inForm Work Area with Image Showing Segmented Tissue A Step Bar The Step bar displays an icon for each processing step included in the project that is open in the Manual Analysis tab The Step bar only displays when the Manual Analysis tab is selected A new project is automatically created when inForm opens By default the new project only includes the Prepare Images and Export steps To create a new project with additional steps see Creating Projects 2 To add remove or change the steps in the open project see Configure a Projecta Click the icons in the step bar at the top of the work area to display the settings for each step in the Manual Analysis tab For example click Prepare Images to display the Image Preparation Settings panel in the Manual Analysis tab click Export to display the Export Settings panel Configure button Opens the Configure Project window to change the steps in the current project See Creating Projects 2 for instructions Vectra Tissue Finder and Vectra HPF Finder projects cannot be configured Prepare Images Use to specify the image format and sample format change the image resolution load a spectral library convert images to optical density and unmix images See Preparing Images 35 fo
64. es use the Review Merge Editorl 93 to combine the data from multiple output files into a single file for detailed and quantitative analysis The Review Merge Editorl 931 has tools to filter out data from samples that are inadequately prepared stained and or segmented so that merged data sets are reliable This feature is only available with inForm Tissue Finder 3 1 View Editor Use the View Editor to select the display options for images or data tables after processing The images and data tables available are determined by the steps in the project The steps available are determined by the inForm license This section lists the views that are available for each software configuration For definitions of the options see Displaying the Extracted Datal 73 To open the View Editor click the View Editor button on the image toolbar lt 1 View Editor balak Views Available In Viewer panse Color image x Color Image Shows the original color image EREE plus the regions selected in the Image ict Options Brightness g 0 Component Image Shows an image of a E single unmixed component Contrast Y 0 Composite Image Shows a reconstructed image of all components ieee nies Reset to Defaut Image Options Views Available In Basic Analysis Z Tissue Segmentation Map V Nuclear Segmentation Map Includes all views in Viewer plus 7 Cytoplasm Segmentation Map V Membrane Segmentation Map Tissue Segmentation Data
65. es 48 3 Training the Tissue Segmon osinaren a ia aa RA EE AE RARA 49 4 Adding Images to the Training Selaecsssisererrriisemenici in E E EEN 51 5 Editing Masks and Re Training cececccseeeeceeeeeeeeeaaeeeeeeaaeeeeeeaaaeeeeeaaeaeeeeaaaaeeeeaaeeeeeaaeeeeees 52 6 Completing the Automated Tissue Segmentation Step cc ceeeececeeeceeeneeeeeeaeeeeeeeaaeeeeeeaaeeeeees 53 Chapter 11 ThreShold ccsssseececieeeeenneceeeeeeeenesseeeeeeeenssesneeeeseeeesensneaeeeeeennnsenaaeeaees 54 Chapter 12 ColocaliZation ccccsecesssenecceeeesseeneeensessenneeeeeeensensneaaeseeennnsssaneaeens 57 Chapter 13 Segmenting Cells sssssssssssssssssserssrssssrssssssesssnssssessnnseorensanserssssennoess 59 T SEGMENTING NUCO heerenean ed danesy eeptunstes danoasadebie toad daad df espeuniead denedsaaeeaneecibeae i eceeedoedan 59 2 Segmenting Cytoplasm cccceeeeeeeeeeeeeeeeeeeeeeeeaeeeeeeeaeeeeeeaeeeeeeeaaeeeeeeeeeeeeseaaeeeeeeeaeeeeeeeeeeeenees 63 3 SSOMENTING Memia reisais S EAE E issued duu E OEE AE OE 64 Chapter 14 Scoring IHC or IF ssssssssssssnunnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnn nnmnnn nnana 66 Chapter 15 Counting ObjectS sssssssssssesnnnununnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnn nnana 71 Chapter 16 Displaying the Extracted Data sccesssssceesssseeeeeesseeeeensseeeenenenaeens 73 1 Viewing the Color IMage cececeeeeeeeeeeeeeeeee eee eeeeee een naa EEEE A EENE EN E
66. eshold for the second marker after selecting the first marker settings Show Score Colors Select the score colors to display on the image Double Positivity First Marker or Second Marker To select the Double Positivity settings 1 If the tissue has been manually or automatically segmented select the Tissue Category as described above 2 Select Double Positivity 2 x 2 bin in the Scoring list box 3 Under First Marker Settings select the Compartment and Component and then click the Auto button 4 Under Second Marker Settings select the Compartment and Component and then click the Auto button 5 To adjust the First Marker Threshold select First Marker in the Histogram Results box and then adjust the First Marker Positivity slider or type the desired value in the Positivity text box Positive regions are indicated by the Single Positive 1 color Negative regions are indicated by the Double Negative color 6 To adjust the Second Marker Threshold select Second Marker in the Histogram Results box and then adjust the Second Marker Positivity slider or type the desired Score IHC or IF Settings research use only Tissue Category Category1 Z Scoring Double Positivity 2x2bin First Marker Settings Compartment Nuclei Cyto Membrane Component Hematoxylin x Threshold Max 0 20 Ato Positivity Optical Density g 0 028 0 0 0 2 Negativity Positivity 36 77 63 23
67. ess or implied warranty and with no claim as to its suitability for any purpose The Vigra library is Copyright 2008 Ulrich Kothe Heidelberg Collaboratory for Image Processing The VIGRA License identical to the MIT X11 License Permission is hereby granted free of charge to any person obtaining a copy of this softw are and associated documentation files the Softw are to deal in the Softw are w ithout restriction including w ithout limitation the rights to use copy modify merge publish distribute sublicense and or sell copies of the Softw are and to permit persons to Software EULA 101 whom the Softw are is furnished to do so subject to the follow ing conditions The above copyright notice and this permission notice shall be included in all copies or substantial portions of the Softw are THE SOFTWARE IS PROVIDED AS IS WITHOUT WARRANTY OF ANY KIND EXPRESS OR IMPLIED INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT IN NO EVENT SHALL THE AUTHORS OR COPY RIGHT HOLDERS BE LIABLE FOR ANY CLAIM DAMAGES OR OTHER LIABILITY WHETHER IN AN ACTION OF CONTRACT TORT OR OTHERWISE ARISING FROM OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE 12 This software uses the Boost 1 53 0 library http www boost org This software includes machine executable object code generated by a source language processor
68. ew Project 27 or Configure the Open Project 28 The name of the current project displays in the inForm title bar Specify the desired project name when creating a new project or when configuring the project Create a New Project The steps available when creating a new project are determined by the software configuration used to create the project See Software Configurations 8 for information on the features available for each software configuration To create algorithms for use with Vectra see Creating Algorithms for Vectral 94 The Vectra options are only available if inForm Tissue Finder was configured during installation for use with Vectra images To create a new project 1 Select File gt New gt Project If an unsaved project is open already inForm prompts you to save the current project If you choose No all unsaved settings are lost The New Project window opens as shown in the figure below Prepare Images and Export are automatically included in all projects New Project 2 8 Prepare Images Segment Tissue Find Features Score Export Prepare the images for analysis Segment the images into diferent tissue Find features within the imagery using Score cells based on instensiy thresholds Save the table data and imagery categories thresholds or segmentation algorthms Manual Tissue Segmentation Cell Segmentation Scoring requires Cell Segmentation Trainable Tissue Segmentation Object Segmentation Skip This
69. figurations cceeeccceeeeece eee eeeeaaeeeeeeaaeeeeeeaaeeeeeaaeeeeeeaaeeeeeeaaeeeeeeaaeeeeeeaaeeeeeeaaiaees 8 BS KOY TOM E E E E scous vasueunahe cada E O E EE 10 4 Using Online Documentatiorisesssinarssiiis iasa a R EERS A RAAE 11 5 Installing and Staring MP OMe E TE EE E E ENS 11 6 AcMatng Iho ICONS Oeean N E shaxecs 12 7 Contacting PerkinEIMer ccc eeeeeeeeeeeeeeee eect cena ener ennai ee eeaaeeeeeeaaieeeeeaaaeeeeeaaeeeeeaaeeeeeeaaeeeees 13 Chapter 3 Understanding the inForm Work Area ssssssssssssnnununnunnnnnnnnunnnnnnnnnnnnnn 14 T View CORO erreien EEE E E EE EEEE OEA S E E E EEE R 19 2 Available Images WINGOW sesesssinordsiianidide ai EAR A K AASE 21 3 New Project WINGO reiasa AEA E AEE AEA E AE 22 4 Spectral Library WinNdOW sassessrssscuincnne E E E E N 24 Chapter 4 Common Image Analysis TaSkS cccsssssscseessssseneeeneeeseeneeeenennens 26 Chapter 5 Working with Projects and Algorithms e E T A ATT 27 1 Creating Projects ccecsucccceenenniedeecetieneeneutedensnid E EE EENE E EEE EEEE 27 2 Opning a POC e E EE 29 3 Opening an AIQOrith ee eee eee ce cece cece ae ee ee eeaaeeeeeaaeeeeseaaeeeeseaaeeeessaaeeeseaaeeeeseaaeeeeaeaaneeeeeags 30 4 Saving an AIQOrithm cece eceeeeeee eee eect eee e eter ee ae ttanu ttina eeeeaaeeeeeeaaeeeeeeaaeeeeseaaeeeeeeaeeeeeaeaeeeesenees 30 Chapter 6 Opening IMageS ss sssssssrrununnurrnnnununnnnnnnnnnnnnnnnnnnnunnnnnnnnnnnnnn nnmnnn neneman 31 1 O
70. for the entire cell e Cell statistics are calculated over the entire segmented area of the cell Unsegmented areas are not included e The Entire Cell statistics are calculated using only the cell components that are segmented For example if only nucleus and cytoplasm are segmented the Entire Cell statistics will not include the membrane TMA Core Info Select the options to include in the data table See TMA Core Infol 871 for descriptions of each option Position Stats Select which values to display for each segmented object See Position Stats 8 for descriptions of each option Shape Stats Select the options to include in the data table See Shape Stats 88 for descriptions of each option C Std Dev C Total TMA Core Info Show Core ID Show Slide Info lt i View Editor army Data Displayed Cell Segmentation Data x 2 Table Contents Components Tissue Categories All Components All Categories V Her2 Fast Red Category1 7 ER DAB Category2 E Hematoxylin E Tissue Component Stats Position Stats E Min X Position V Mean l Y Position O Max Process Region ID _ Distance from Process Region Edge Category Region ID Distance from Tissue Category Edge Shape Stats Area pixels Area percent Compactness Minor Axis Major Axis Axis Ratio Figure 46 View Editor Cell Segmentation Data e inForm User Help The columns bel
71. from the Boost libraries covered by the Boost license These libraries have not been modified by CRI and all rights are reserved by the copyright holder This software is provided as is without express or implied warranty and with no claim as to its suitability for any purpose The Boost license is reproduced below Boost Softw are License Version 1 0 August 17th 2003 Permission is hereby granted free of charge to any person or organization obtaining a copy of the softw are and accompanying documentation covered by this license the Softw are to use reproduce display distribute execute and transmit the Softw are and to prepare derivative w orks of the Software and to permit third parties to w hom the Softw are is furnished to do so all subject to the follow ing The copyright notices in the Softw are and this entire statement including the above license grant this restriction and the follow ing disclaimer must be included in all copies of the Softw are in whole or in part and all derivative w orks of the Softw are unless such copies or derivative w orks are solely in the form of machine executable object code generated by a source language processor THE SOFTWARE IS PROVIDED AS IS WITHOUT WARRANTY OF ANY KIND EXPRESS OR IMPLIED INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY FITNESS FOR A PARTICULAR PURPOSE TITLE AND NON INFRINGEMENT INNO EVENT SHALL THE COPY RIGHT HOLDERS OR ANYONE DISTRIBUTING THE
72. fter all of the desired settings have been selected in the Tissue Segmentation Training panel and the Tissue Segmenter has been trained process the images using any of the methods below and then advance to the next step in the project e Click the Segment Image button to segment the currently selected image e Click the Segment All button to segment all images open in the project e Click the Advance button to display the next step in the project The tissue is segmented the next time images are processed e Click the next step in the project in the Step Bar 14 at the top of the inForm window The settings panel for the next step displays The tissue is segmented the next time images are processed Trainable Tissue Segmentation 53 11 Threshold Use the Threshold step to set threshold values for any of the unmixed components in an image and then to view the regions on the image that are above the threshold Statistics are calculated for the thresholded regions and are displayed in the Quant Data table To set the thresholds and calculate the threshold statistics for the images 1 Click the Threshold button in the step bar D Threshold 2 Inthe Components to Threshold list select each of the image components you want to create a threshold map for 3 Click the Auto button to calculate a preliminary value for the Threshold Max and to choose a preliminary Threshold value The Threshold Max is set to the value of the brightest
73. g New 27 Definition 10 Open 29 Opening images 31 sawe 29 Quant data viewing 86 R Recent trainings 50 Reference white 36 Refine splitting 62 Region ID number 34 Removing images from a project 31 Re training 52 roundness minimum circularity 62 roundness minimum circularity 62 Tissue segmentation 47 S Tissue segmentation data viewing 78 Tissue Segmenter definition 10 Sample format 35 Tissue Segmenter training 49 Saving Toolbars 16 algorithms 30 Training regions drawing 48 projects 29 Training Set Score data viewing 82 Adang magas al Scoring IHC or IF 66 Training the tissue segmenter 49 Scoring type 66 Transferring the license 12 Segmentation Priority 65 U Segmentation resolution 50 TONET Segmenter training 50 Segmenting cells 59 Unmixing spectra for 36 Segmenting cytoplasm 63 User interface 14 Segmenting membrane 64 Segmenting nuclei 59 V Segmenting tissue 47 Segmenting tissue training 51 Select Reported Units 39 Signal scaling 60 W Specifications 7 Spectra for unmixing selecting 37 View Editor 19 Spectral Library Weighting 39 loading 36 Welcome 6 viewing 39 Spectral Library window 24 Support Technical 13 Supported image formats 7 T Technical Support 13 Ten bin scoring 69 Threshold 54 Threshold maximum 66 Tissue automatically classifying 47 manually classifying 44 Tissue Categories Creating 47 Deleting 47 Tissue Category Definition 10 For scoring 66 Selec
74. ges Mean The average of all signals of the selected component across all wavelengths Available for all units except Counts for Monochrome images This is the equivalent of the settings in versions of inForm prior to V2 0 2 Examples For RGB color images inForm reports the average signal across the three color planes red green and blue A pure blue pixel with 240 counts would have a component strength of 80 240 counts 3 planes For multispectral images the signal is averaged across all wavelengths measured So a pixel having a DAB signal of OD 0 60 in the blue OD 0 20 in the green and OD of 0 10 in the red would report a signal of 0 30 OD units 0 60 0 20 0 10 summed OD 3 planes Peak The peak signal of the selected component across all wavelengths Available for images converted to Optical Density and for fluorescent multispectral images Default for brightfield images 40 inForm User Help Weighting Example The simplified example below explains the difference between the three weighting options The example uses only two fluors and ignores noise and other sources of error This example uses a spectral library with just two elements Fluor A and Fluor B and eight wavelengths 420 460 500 540 580 620 660 700 The spectral library is shown in the figure below Fluors 054 Fluor A Fluor B Signal 0 27 0 1 7 500 e00 Wavelength Figure 15 Spectral Library Exam
75. ges to export RGB RGB with Tissue Segmentation Map E RGB with Cell Segmentation Map RGB with Scoring Map E RGB with All Segmentation Maps Composite Image Component Images multiimage TIFF Maps to export Tissue Segmentation Map Cell Segmentation Map TIFF Review Merge Batch Analysis Manual Analysis Tables To Export F Tissue Segmentation Data Cell Segmentation Data Score Data Table Fields To Export Export all fields Use view settings File Name Options Copy 0 level s of directory structure from source image Figure 53 Batch Analysis Tab 1 Click the Batch Analysis tab to configure the batch settings and select images for processing You are prompted to save the open project Click Yes to save the open project or No to close without saving 2 Algorithms that contain the Manual Classification step are not compatible with batch processing If the open project contains this step the software will post a warning Open a Batch Processing project that does not contain steps that require user interaction and then click the Batch tab 3 Select a Batch Algorithm or Project When you first switch to the Batch panel the Algorithm settings are copied from the active project You can run a batch using these settings or click the Browse button to select an Algorithm ifp or Project ifr file that contains the desired settings When
76. h the weighting values a2 inForm User Help Components of Unmixed Signal 1514 aa Fluor B peak 1 52 Fluor A Fluor B total area under curve 7 75 FluorB 107 R NII S Fluor A total area under curve 2 55 3 2 a O Fluor A peak 0 57 0 55 R o FluorA mean Ae 7 gt 0 05 500 e00 700 Wavelength Figure 18 Weighting Values 8 5 Completing the Prepare Images Step After all of the desired settings have been selected in the Image Preparation Settings panel process the images using any of the methods below and then advance to the next step in the project e Click the Prepare Image button to apply the selected settings to the currently selected image e Click the Prepare All button to apply the selected settings to all images open in the project e Click the Advance button to apply the selected settings to all images open in the project and then display the next step in the project e Click the next step in the project in the Step Barl 14 at the top of the inForm window All images are processed using the current settings and then the settings panel for the next step displays Preparing Images a 9 1 9 2 Manual Tissue Segmentation The topics in this section describe how to manually draw tissue categories on an image The tissue categories can then be used to generate data or can be used during the Cell Segmentation Scoring or Object Counting steps Algorithms that include a Manual
77. has been created the settings in the steps can be changed Changing settings may require you to reprocess the images for later steps in the project Working with Projects and Algorithms 5 3 5 4 Opening an Algorithm Algorithms include all steps in processing an image image preparation image analysis and export You can load an algorithm into inForm and then open the images you want to process or load the images first and then apply the algorithm Algorithms can be used in Batch Processing sto automatically process multiple images at the same time using the same analysis settings If a project is opened for use in Batch Processing the project s algorithm is used to process the images To open an algorithm select File gt Open gt Algorithm If a project or image is already loaded inForm warns you that all images will be reset to their initial state Click Yes to discard all training regions tissue regions cell regions object regions thresholds and component images Select the desired algorithm and click Open Open the desired images if they are not already open To prepare the image s click the Prepare Image or Prepare All button To review or change the settings in each algorithm step click the Advance button to advance through each step or click the step buttons at the top of the window When algorithm development is complete see Saving an Algorithm 30 Saving an Algorithm inForm Algorithms allow you to
78. he project Configure a Project You can configure a project to add or remove steps from the project inForm retains as many settings as possible depending on which steps are added or removed from the project For example if you created a project without tissue segmentation and later decide that you want the project to include manual tissue segmentation you can modify the project to add Manual Tissue Segmentation Settings that are not affected by the changes are retained but some settings may need to be adjusted due to the changes to the project You cannot reconfigure a Vectra project and you cannot convert a non Vectra project into a Vectra project The steps available when configuring a project are determined by the software configuration used to configure the project See Software Configurations 8 for information on the features available for each software configuration 1 Select Tools gt Configure Project or click the Configure button 2 Use the Configure Project window to select the desired steps The Configure Project window contains the same options as the New Project Windowl 221 3 Click the Configure button and inForm reconfigures the project 4 Verify that all of the desired settings are selected in each step 2 inForm User Help 5 2 Opening a Project To open a saved inForm project 1 Select File gt Open gt Project select the name of the project and click the Open button The images in the project
79. hors make NO WARRANTY or representation either express or implied with respect to this softw are its quality accuracy merchantability or fitness for a particular purpose This software is provided AS IS and you its user assume the entire risk as to its quality and accuracy This softw are is copyright C 1991 2010 Thomas G Lane Guido Vollbeding All Rights Reserved except as specified below Permission is hereby granted to use copy modify and distribute this softw are or portions thereof for any purpose without fee subject to these conditions 1 If any part of the source code for this softw are is distributed then this README file must be included w ith this copyright and no w arranty notice unaltered and any additions deletions or changes to the original files must be clearly indicated in accompanying documentation 2 If only executable code is distributed then the accompanying documentation must state that this softw are is based in part on the work of the Independent JPEG Group 3 Permission for use of this software is granted only if the user accepts full responsibility for any undesirable consequences the authors accept NO LIABILITY for damages of any kind These conditions apply to any softw are derived from or based on the IJG code not just to the unmodified library If you use our work you ought to acknow ledge us Permission is NOT granted for the use of any UG author s name or company name in adverti
80. ield or Fluorescence Monochrome images do not use a Sample Format For all other image formats inForm automatically detects the Sample Format For some image types you can change this selection if desired If desired you can select the Image Format and Sample Format before adding images In this case only images that match the selected Image Format and Sample Format can be opened Change Resolution Select the Change Resolution check box to change the image resolution Lower the image resolution to increase the speed of training and tissue segmentation If the set of images is so large that all of the images cannot be held in memory lower the image resolution prior to unmixing Preparing Images 35 8 2 Loading a Spectral Library inForm uses a process called spectral unmixing to decompose the image into its various components For a simple example a color image can be unmixed into its red green and blue components But more powerfully an image can also be unmixed into its stain components such as Hematoxylin and DAB for Brightfield images which describe the contribution of each stain to the overall signal The contributions of each stain are based on a spectral library that shows the characteristic color distribution of light associated with each stain A fluorescence example could consist of stain components such as DAPI and FITC the spectral library can additionally include the color distribution associated with tissue autofluoresce
81. imple threshold This approach is purely pixel based This approach can also be used for other image analysis needs that can be satisfied with a simple threshold such as detecting all pixels within a tissue category that stain positive for an IHC stain e Select the Object Based approach if the nuclear counterstain does not provide consistent and specific staining of nuclear objects and more advanced morphometry based approaches are needed to detect nuclei Applying Signal Scaling e Auto Scale Select Auto Scale if you want the software to automatically scale each nuclear counterstain component plane individually before performing nuclei segmentation This approach is a good place to start and is useful if nuclear counterstain signals vary widely Auto Scale is often sufficient for fluorescence applications Cell Segmentation Settings Compartments to Segment V Nuclei V Cytoplasm Membrane Edge Rules V Discard Cell f Touching An Edge Nuclei Cytoplasm Membrane General Tissue Category Approach All categories Tissue Figure 25 Cell Segmentation Settings for Nuclei Cell Segmentation Settings Compartments to Segment V Nuclei V Cytoplasm V Membrane Edge Rules Discard Cell f Touching An Edge Nuclei Cytoplasm Membrane General Tissue Category Tumor X Approach Object Based Signal Scaling for Pixel Based Threshold Figure 26 Cell Segmentation Approach Signa
82. in the Cell Segmentation and Count Data Tables if tissue has been segmented Distance from Tissue Category Edge Pixels The distance in pixels from the segmented object to the nearest edge of the tissue category Note that if the distance is reported as N A this indicates that the distance is unknown or indeterminate for example if a cell is in a tissue region that abuts the edge of the image The cell may be closer to the edge of the image than to any other region in the image so it cannot be determined whether the distance to an in image point would be nearer than a point that lies beyond the edge of the image Only available in the Cell Segmentation and Count Data Tables if tissue has been segmented 16 14 Shape Stats The data tables contain the options below Area pixels The size in pixels of the region Area percent The region size as a percent of the total processed area If there are no processing regions it is the total image area Otherwise it is the total area of the processing regions Compactness The compactness of the region Compactness is the ratio of the area of the region to the area of a circle the most compact shape having the same perimeter A circle has a compactness of 1 Objects that are elongated or irregular are less compact as are objects with irregular or reticulated edges Minor Axis The smaller of the two axes of the minimum area bounding box enclosing the region Major Axis The larger of
83. information from images of stained tissue sections Getting started new users gt Be sure to review the topics in the Introduction 7 to familiarize yourself with the features and specifications of the software 6 inForm User Help 2 1 Introduction The topics in this section provide basic information about inForm what it is for and what you can do with it About inForm inForm operates on monochrome color or multispectral images of tissue sections Sections can be stained with standard stains immunohistochemical stains IHC and immunofluorescence stains IF including Qdots Stains can be single or multiplexed for multi analyte analyses to support applications such as signaling pathway research in oncology inForm is available in different configurations each providing additional capabilities for different applications inForm Tissue Finder can be trained to find virtually any tissue type or structure such as tumor fibrosis inflammation stroma granuloma or vessels inForm Cell Analysis can be used to assess IHC or IF staining levels on a cell by cell basis for cell prototyping It can also generate object counts inForm Basic Analysis can generate area statistics and quantitative stain levels for hand drawn or threshold generated regions inForm Viewer can visualize component data and reconstruct the unmixed signals into a composite image Features gt Compatibility with PerkinElmer s mult
84. ing the tissue masks makes it easier to see the colors on the original image If Processing Regions have been defined use the Process Regions button to show or hide the process region masks Click and drag to outline the area to be included intersecting the defined tissue classification areas as desired Release the mouse button to connect the end point to the start point The outlined area is assigned to the specified tissue category To remove an area from the tissue category 1 Click the Select Regions to Unclassify button Gg Click and drag to enclose the area from which you want to remove the tissue classification Release the mouse button to connect the end point to the start point The outlined area is removed from the tissue category Manual Tissue Segmentation s To remove single pixels from the tissue category 1 Zoom in on the image 2 Click the Select Pixels to Unclassify button 3 Click on a single pixel to remove it from the tissue classification region 4 Click and drag to remove a line of pixels from the tissue classification region a6 inForm User Help 10 Trainable Tissue Segmentation The topics in this section describe how to automatically classify tissue using the tissue segmenter To set up the tissue segmenter add tissue categories and draw training regions on images in the training set train the tissue segmenter segment images and then fine tune the segmenter based on initi
85. inly marked as such and must not be misrepresented as being the original softw are 3 This notice may not be removed or altered from any source distribution Should you have any question concerning this Agreement you may contact CRI by writing to CRI 68 Em St Hopkinton MA 01748 You may also call 1 508 435 9500 106 inForm User Help Index 0 0 3 4 bin scoring 68 A About inForm 7 Adding images toa project 31 Algorithm 10 30 Creating for Vectra 94 Definition 10 Opening 30 Saving 30 Analysis process editor 15 Auto scale 60 Available Images Window 21 Batch Analysis tab 19 Batch processing 91 ay oe Calculating Fractional Tissue Area 97 Categories adding 47 Cell segmentation data viewing 80 Cells segmenting 59 Classifiers 94 Classifying Tissue automatically 47 manually 44 Colocalization 57 Colocalization data viewing 84 Color Image view 73 Compartment 66 Component data Viewing 73 Component for scoring 66 Component units 39 component view 74 Components for training 49 Components to include 64 Composite Image viewing 77 Contact PerkinElmer 13 Count data viewing 85 Counting objects 71 Cytoplasm segmenting 63 D Data Displaying extracted 73 Exporting 89 Merging 93 Down sample 35 Drawing Processing Regions 34 Drawing training regions 48 E Editing masks 52 Excel 97 Export settings 89 Exporting the data 89 Extracted Data viewing 73 F et Feature
86. ion of the values for the component in the region The summary line displays the standard deviation of the individual means e Total Displays the sum of the value of the component across all pixels in the region 16 12 TMA Core Info The data tables contain the options below e Show Core ID If the image is of a core on a TMA slide scanned with a Vectra scanner the Sector Row Column and Field of the core display in the table If there is no core information the columns contain zeros e Show Slide Info If the image was taken with a Vectra scanner the Lab ID and Slide ID from the file name display in the table If there is no slide information the columns are blank 16 13 Position Stats The data tables contain the options below Units are in pixels e X Position and Y Position The X and Y location of the center of the bounding box for the region e Process Region ID If process regions are defined for the image displays the ID number of the process region containing the tissue category Displays N A if an object spans multiple process regions e Distance from Process Region Edge Pixels The distance in pixels from the center of this region s bounding box to the nearest edge of the containing process region Displays N A if the Displaying the Extracted Data distance cannot be determined Category Region ID If tissue is segmented displays the Region ID of the tissue category the segmented object is in Only available
87. ispectral images im3 and im4 and with conventional color images tif jpg etc gt Full range of software configurations which can include capabilities from simple tissue area measurements to cell as a unit multiplexed molecular phenotyping To view the levels of inForm and the capabilities of each level see Software Configurations 8 gt Computationally efficient runs on a standard laptop Specifications gt Operates on PerkinElmer s multispectral images im3 and im4 and monochrome or color images TIFF BMP JPEG PNG gt Computer Specifications Microsoft Windows 7 64 bit with 83GB RAM recommended Minimum Microsoft Windows 7 32 bit with 2GB RAM Introduction 2 2 Software Configurations inForm is available in four different software configurations Each configuration enables specific tasks for viewing selecting segmenting and counting The following inForm software configurations are available Viewer inForm Viewer includes the following features e View images e Unmix images e View components and their signals e View a composite of the component images e Export images Basic Analysis inForm Basic Analysis includes all of the features available in inForm Viewer plus the following features e Manually segment tissue e Create threshold maps e Create colocalization maps e Generate quantitative data for threshold and colocalization maps e Save projects and algorithms e Bat
88. just if necessary 8 If desired change the threshold map view 58 or edit the threshold maps 5e 9 To view the statistics for the generated regions in the Quant data table see Viewing the Quant Data Tablel s or to export the data see Exporting the Datal 8 sa inForm User Help Changing the Threshold Map View The threshold maps for each component are stacked in layers on top of the original graphic with the top component in the Simple Threshold Settings panel as the uppermost layer in the image display You can change the order of the threshold maps or hide threshold maps 1 To hide a threshold map clear the Visible check box under the component that you want to hide Hides Her2 Fast x Red Component l Hides ER DAB Component 2 To move a threshold map up or down in the stack of maps click the Up or Down arrows on the left of the Color button Her2 Fast Red Threshold Max 0 20 s Ato Moves map toward bottom of stack Moves map toward top of stack Threshold 55 Editing Threshold Maps Threshold maps can be edited manually Additional areas can be drawn onto the threshold map or undesired areas can be erased from the threshold map This procedure describes how to use the editing tools on the Simple Threshold Settings panel The Add Thresholded Regions and Erase Thresholded Regions buttons on the toolbar can also be used 1 To view editing options click the More button under
89. l Scaling for Segmentation Only Auto Scale values range from 0 1 Fixed Scale Counts Figure 27 Signal Scaling e Fixed Scale If you need better nuclear segmentation performance and counterstain signals are consistent and reliable you might get better results by selecting the Fixed Scale option The units are in parentheses This requires entering a scaling value which can be determined by 6 inForm User Help choosing the View Component Data tool from the toolbar and mousing around on the brightest nuclei to see the signal levels of the component being used for nuclear segmentation Everything in a nuclear segmentation indicated by a green region should be nuclei o Reduce the maximum OD or Counts value if nuclear segmentations are smaller than the nuclei or if no nuclear segmentations occur when the segmenter is finished o Increase the scale if segmentation is too sensitive and is finding many objects that are weakly stained and not nuclei Adjust the scale until the unwanted regions are not detected Selecting Components for Nuclear Segmentation You can further add to what you consider nuclear pixels by including other components in the underlying signal Pick an individual component signal as the primary and then select secondary and tertiary components if desired Enter a minimum signal value for each component In Auto Scale mode the signal range for each component is 0 to 1 If you selected the Fixe
90. lect enough sample slides so that heterogeneity of the entire slide set is represented in the sample slides The sample slides can contain up to 20 slides and should contain examples of all of the tissue types the algorithms will encounter when classifying images during the Vectra slide scanning protocol You may want to scan two sets of sample slides the first set can be used to train the algorithm and the second set can be used to test the algorithm In Vectra load and scan the sample tissue slides manually one at a time Refer to the Vectra User s Manual for detailed instructions Selecting Images for Algorithm Training When you have finished scanning the sample slides review the monochrome and color image data sets and select images for algorithm training and testing as follows 1 Select from 3 to 10 monochrome images from the Monochrome fullres Vectra directory These images should contain examples of all of the tissue types that the algorithm will encounter Group images that have similar phenomena into pairs and divide each pair into two separate folders Images from one folder will be used for training the algorithm Images from the second folder will be used to test the new algorithm 2 Review the color image data set and select from 3 to 10 color images from the LPF fullres Vectra directory representing all phenomena the algorithm might encounter Again group these images into two folders of up to 20 images each 3
91. lick Yes Appendix A Calculating Fractional Tissue Area s 22 Software EULA The follow ing is an agreement the Agreement betw een you and Cambridge Research amp Instrumentation Inc 68 Elm St Hopkinton MA 01748 CRI for software known as inForm and its accompanying documentation collectively the Softw are By installing and or using the Softw are you agree to the follow ing terms and conditions If you do not agree to all of the terms and conditions in this Agreement you may not install or use the Softw are 1 Single Use License The Software is licensed to you and not sold Subject to the terms and conditions of this Agreement CRI hereby grants to you a restricted non exclusive non transferable non assignable non sublicensable and revocable license to use for your internal purposes only the executable code version of the Software and the accompanying documentation in hard copy or electronic format CRI RESERVES ALL RIGHTS NOT EXPRESSLY GRANTED BY THIS AGREEMENT 2 Specific Restrictions You may use the Software only on a single computer You may make only one 1 copy of the Softw are solely for backup purposes You agree that except as expressly permitted by applicable law neither you nor a third party acting on your behalf will i decompile disassemble or reverse engineer the Softw are ii modify or create derivative w orks of the Softw are iii use the Softw are in any manner to provide service bure
92. limit on nuclear Components for Nuclear Segmentation Minimum Signal 0 1 Primary hem z oo g Boi PR Her2 T hem dab ER her Le dark red PR Her2 CO Tertiary Figure 28 Components for Nuclear Segmentation Size Range Minimum Size pixels ho g C Maximum Size pixels 10000 8 Figure 29 Size Range Segmenting Cells at segmentations in number of pixels Nuclear segmentations larger than this number are ignored This can be useful if you want to ignore large clusters of nuclei or if there are other objects in the classified region of interest that are not nuclei Entering Nuclei Clean up Settings e Select Fill Holes when some nuclei end up segmented with eet holes in them This results from a common effect in tissue FA fil Holes sections where nuclei appear hollow Checking this check f W Max Hole Size pixels 35 box automatically fills holes in nuclei Refine Splitting requires Maximum Size e If you want only those nuclei holes that are smaller than a Roundness Minimum circularity 0 20 certain size to be filled in select Max Hole Size pixels and enter the maximum number of pixels Nuclei with holes that are larger than this value are not filled Figure 30 Clean up options e If you select Refine Splitting nuclei that are touching each other are split into separate nuclei instead of being detected as one large nucleus This option is available only if Maximum Size is sel
93. lt button Click to reset the Rendering Options back to the default settings Image Options See Image Options 8 p 1 View Editor Data Displayed Color Image Rendering Options Adjust Brightness U E Use Enhanced Contrast Image Options V Tissue Segmentation Map Nuclear Segmentation Map Cytoplasm Segmentation Map V Membrane Segmentation Map Show Score Colors V Training Regions V Processing Regions E Equalize Display Histogram Figure 41 View Editor Color Image Displaying the Extracted Data 16 2 Viewing the Component Image Select a Component in the Data Displayed list then choose the desired options The following options are available for the Component Image Rendering Options Show As e Brightfield Select to display the component signal on a white background e Fluorescence Select to display the component signal on a black background Component Color e Black and White Select to show the component as a black and white image The component signal color is determined by the Brightfield Fluorescence setting e Color Select to show the component in color This color is calculated from the spectral shape of the component i e the color you would see when viewed through a microscope Scaling Scale Views for each Image Individually Select to show each image scaled individually Dim images are brightened to show
94. m this softw are without specific prior written permission THIS SOFTWARE IS PROVIDED BY THE AUTHOR AS IS AND ANY EXPRESS OR IMPLIED WARRANTIES INCLUDING BUT NOT LIMITED TO THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED INNO EVENT SHALL THE AUTHOR BE LIABLE FOR ANY DIRECT INDIRECT INCIDENTAL SPECIAL EXEMPLARY OR CONSEQUENTIAL DAMAGES INCLUDING BUT NOT LIMITED TO PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES LOSS OF USE DATA OR PROFITS OR BUSINESS INTERRUPTION HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY WHETHER IN CONTRACT STRICT LIABILITY OR TORT INCLUDING NEGLIGENCE OR OTHERWISE ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE Julian Sew ard Cambridge UK jsew ard bzip org bzip2 libbzip2 version 1 0 4 of 20 December 2006 17 This software uses the libjpeg turbo 1 5 13 library http libjpeg turbo virtualgl org This softw are makes use of the libjpeg turbo library This library has not been modified by CRI and all rights are reserved by the copyright holder This softw are is provided as is without express or implied w arranty and with no claim as to its suitability for any purpose The library is covered by the libjpeg turbo License Additionally libjpoeg turbo incorporates parts of libjpeg which is covered by the BSD license The libjpeg turbo license is reproduced below Redistribution and use in source and bin
95. mage using either thresholded or algorithmic segmentation See Counting Objects 7 for details e Colocalization Identifies regions on an image where multiple unmixed component thresholds overlap Threshold values for each component are specified individually See Colocalization 5 for details e Threshold Identifies regions on an image where a component is above a specific threshold See Threshold 54 for details Score Score cells based on intensity thresholds e Scoring Scores the segmented nuclei cytoplasm and or membrane to count the percent positive of a specific component above a specified threshold See Scoring IHC or IFI 6 for details Export Required for all projects Specifies the export settings for images and data tables after processing is complete See Exporting the Datal 8s for details Understanding the inForm Work Area 23 3 4 Spectral Library Window Use the Spectral Library window to view charts of the emission spectral curve for each stain selected for unmixing Viewing the spectral library can help when troubleshooting unexpected unmixing results Ideally the spectra should be easily distinguishable in each of the imaging bands Typically one filter is used for each stain If the Spectral Library uses multiple filters one chart displays for each filter The figure below shows a spectral library with DAPI CY3 and CY5 with good separation in each chart The Spectral Library window updates dynamicall
96. matically identifies regions of tissue or regions that correspond to specific types of tissue When you train a tissue segmenter it learns how to identify structures in the images based on the training regions within each tissue category The tissue regions or areas detected with the tissue segmenter can then be segmented to detect cells and perform IHC or IF scoring or can be used for object counting Only available with inForm Tissue Finder The set of images that is used to train the Tissue Segmenter to identify tissue categories automatically Images are automatically added to the training set ifa training region is drawn on the image To remove an image from the training set right click on the large image in the image display area and click to deselect Training Set When an image is removed from the training set the training regions on the image are not used when training the Tissue Segmenter The training regions are not deleted from the image so the image can be added back to the training set if desired You can also use the Available Images window to add or remove images from the training set 10 inForm User Help 2 4 2 5 Using Online Documentation This manual contains detailed information about the inForm application It is designed to be used as a reference tool in your everyday work with inForm The manual explains how to use inForm through detailed explanations of features and step by step procedures for common
97. may add Your own copyright statement to Your modifications and may provide additional or different license terms and conditions for use reproduction or distribution of Your modifications or for any such Derivative Works as a whole provided Your use reproduction and distribution of the Work otherwise complies w ith the conditions stated in this License 5 Submission of Contributions Unless You explicitly state otherw ise any Contribution intentionally submitted for inclusion in the Work by You to the Licensor shall be under the terms and conditions of this License without any additional terms or conditions Notw ithstanding the above nothing herein shall supersede or modify the terms of any separate license agreement you may have executed with Licensor regarding such Contributions 6 Trademarks This License does not grant permission to use the trade names trademarks service marks or product names of the Licensor except as required for reasonable and customary use in describing the origin of the Work and reproducing the content of the NOTICE file 7 Disclaimer of Warranty Unless required by applicable law or agreed to in writing Licensor provides the Work and each Contributor provides its Contributions on an AS IS BASIS WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND either express or implied including w ithout limitation any warranties or conditions of TITLE NON INFRINGEMENT MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE Y
98. mpactness descriptions of each option aes ajor is Shape Stats Select the options to include in the data Ae to table See Shape Stats 88 for descriptions of each option Figure 50 View Editor Quant Data 16 10 Image Options The Image Options below are available on the View Editor 19 when an image is selected in the Data Displayed list Select or clear the check boxes described below to show or hide maps or regions on all images in the project The buttons on the left side of the image display area also show or hide the maps or regions Show Score Colors is also accessible on the Score Settings panel The Equalize Display Histogram can only be shown or hidden here e Tissue Segmentation Map Select to display the Tissue Segmentation Map on the image Only for projects that include a Manual Classification or Segment Tissue step e Nuclear Segmentation Map Select to display the Nuclear Segmentation Map on the image Only for projects that include a Segment Cells step e Cytoplasm Segmentation Map Select to display the Cytoplasm Segmentation Map on the image Only for projects that include a Segment Cells step e Membrane Segmentation Map Select to display the Membrane Segmentation Map on the image Only for projects that include a Segment Cells step e Show Score Colors Select to display the Score Colors Map on the image Only for projects that include a Score IHC or IF step e Object Segmentation Map Select to di
99. n image 1 Right click anywhere on the image 2 Select Delete Regions gt Delete Processing Regions on the shortcut menu 3a inForm User Help 8 8 1 Preparing Images The first processing step in an algorithm is the Prepare Images step This section describes how to prepare the images for segmentation or analysis by changing image resolution converting brightfield images to optical density and unmixing the images Selecting the Image Format The Image Preparation Settings panel displays the Image Format Sample Format and Change Resolution options Image Preparation Settings Image Format Multispectral im3 v Sample Format Brghtfield Brightfield Change resol Fluorescence Figure 1 0 Image Format and Sample Format Image Format Before image analysis can begin inForm needs to know the type of images in the project Possible image types are Multispectral im3 TRIO im4 RGB or Monochrome If you open a new project and then immediately add new images to the project inForm automatically selects the correct Image Format Monochrome RGB Multispectral im3 or TRIO im4 If you open TIFF BMP or PNG images inForm recognizes the images as either RGB or Monochrome images Sample Format All images in the project must be the same sample format For any RGB images and for Multispectral im3 images created with early versions of Nuance you must select the correct Sample Format Brightf
100. nce TRIO multispectral im4 images contain the spectral library which is automatically selected when the first im4 file is opened in the project For Fluorescence multispectral im3 images the spectral library is created in the Nuance software For Brightfield multispectral im3 images the spectral library is created in the Nuance software For RGB images the spectral library can be created in the Nuance software or you can use the default RGB spectral library that is included with inForm For Monochrome Images spectral libraries are not used To select the spectral library Spectral Library Browse 1 For Brightfield RGB or for multispectral im3 images BF_Her2 ER hem _lib cs acquired with early versions of Vectra convert to Spectra for Uriin l i for Unmixing Optical Density as described in Converting to Optical Ta Herd Faa Red Density s below Images acquired with recent iv ER DAB versions of Vectra are already converted to Optical 5 a Density 2 Click the Browse button and select the desired Spectral Library csl The name of the selected Vew library displays The components in the library display Select Reported Units Weighting in the Spectra for Unmixing list box on z Pex 2 3 For fluorescence images see Remove Hazel 38 for Figure 11 Load Spectral Library details on removing the background haze before unmixing 4 Inthe Spectra for Unmixing list box select the check boxes for
101. ncer tissue category and enter the number of pixels you want to trim from the borders of the cancer regions The Minimum Segment Size option is an adjustable parameter in pixels that removes segments that have fewer pixels than this value The removed segment is replaced by the tissue segment that is most abundant around the segment that was removed lf the Discard if Touching Image Border check box is selected tissue regions touching the edge of the image or process region are discarded If not selected all tissue regions are included 10 4 Adding Images to the Training Set When selecting images for the training set it is recommended that you start with one image per distinct appearance The training set should represent the range of staining levels and tissue architectures that you want to train the tissue segmenter to identify Whenever you draw a training region on an image see Drawing Training Regions 48 inForm automatically adds the image to the training set You can also add or remove images from the training set manually as described below 1 Click the Image List Editor button to open the Available Images Window 2 f ff 7 Select the Training Set check box next to the images to add the image to the Training set Clear the Training Set check box next to an image to remove the image from the Training set Close the Available Images window ao FPF Ye KN When you switch to Single mode a small blue square on each thum
102. nd reporting Note Cells are considered to be an assembly of associated cellular components nucleus cytoplasm and membrane If you are segmenting tissue some cells may straddle two segmented tissue regions If a cell is not entirely contained within a single region the containing region is undefined and the Region ID in the Data tables is N A see Displaying the Extracted Datal 73 The Segment Cells step is only available in inForm Cell Analysis and inForm Tissue Finder Use the Segment Cells panel to set parameters for segmenting cells into their respective categories so that data can be extracted and analyzed on a per cell basis To segment cells 1 Select the desired cell components to segment Nuclei Cytoplasm and Membrane Cytoplasm and Membrane can only be selected after Nuclei is selected 2 Select Discard if Touching an Edge to discard any cell where any cell component touches the edge of the image process region or tissue region if applicable Only the cells that are entirely within the image a process region if applicable and the selected tissue category if applicable are included in the segmentation map If not selected cells are clipped at the edge of the image or region If a cell is clipped at the edge of a process region or tissue category region the size of the nucleus or cytoplasm inside the region may be smaller that the specified Minimum Size 3 If Nuclei is selected click the Nuclei tab and
103. nged Hovering over the line in a spectral curve displays the name wavelength x value and signal strength y value of the nearest data point By default the charts open with Unmix Units and Scale to Max equal scales selected e X Axis Spectral wavelength of light in nanometers e Y Axis The units selected in the Units drop down list a inForm User Help Units drop down list Specifies the units on the Y Axis on the charts Only the units valid for the spectral library are available For example if a spectral library does not contain Raw units the Raw option is not available e Raw Counts Pixel intensity in unit counts Typical Fluorescence values are from 0 to 4095 Typical Brightfield values are from 0 to 255 e Unmix Units Displays the signal used for unmixing Can show artifacts that are not visible in other views Usually the best unit for evaluating a spectral library e Normalized for Exposure Signal is displayed in counts second Used to compare signal strength across bands Fluorescence libraries only e OD Optical Density units Brightfield libraries only Display drop down list Specifies the Y Axis scaling option for the charts e Scale to Max equal scales All charts are scaled to the same value which is the maximum signal value in all of the bands e Scale to Max varying scales Each chart is scaled individually to the maximum signal unit for that chart This option is not available when there i
104. nt stain and object pixels can be found by applying a simple threshold This approach is purely pixel based Select the Object Based approach if the stain does not provide consistent and specific staining of objects and more advanced morphometry based approaches are needed Select the desired Signal Scaling e Auto Scale Automatically scales each component plane individually before performing object segmentation Fixed Scale If you need better segmentation performance and stain signals are consistent and reliable you might get better results by selecting the Fixed Scale option Select the Primary component for object segmentation from the drop down list Adjust the Minimum Signal value for the primary component to the desired threshold value If desired select Secondary and Tertiary components and minimum signal values Object Counting Settings General Tissue Category al categories X Approach Object Based M Signal Scaling for Segmentation Only Auto Scale values range from 0 1 Fixed Scale OD Components for Object Segmentation Minimum Signal 0 1 Primary Her2 Fast Red z 00 l Secondary E Tertiary Size Range Minimum Size pixels 1 E Maximum Size pixels 10000 Clean up V Fill Holes Max Hole Size pixels Refine Splitting requires Maximum Size E Roundness Minimum circularity 0 20 Edge Rules E Discard Object f Touching An Edge To exclude obje
105. o display data An entry row is created for each Region ID in the tissue category Component Stats Select the values to display for each component signal See Component Stats 8 for descriptions of each option Position Stats Select which values to display for each segmented object See Position Stats 87 for descriptions of each option TMA Core Info Select the options to include in the data table See TMA Core Info 8 for descriptions of each option Shape Stats Select the options to include in the data table See Shape Stats 88 for descriptions of each option ay View Editor Sy Data Displayed Tissue Segmentation Data KA Table Contents Components Tissue Categories E All Components E All Categories V Her2 Fast Red Category1 V ER DAB V Category2 E Hematoxylin E Tissue Component Stats Position Stats E Min V X Position V Mean Y Position V Max Process Region ID V Std Dev Distance from Process Region Edge C Total TMA Core Info Shape Stats E Show Core ID Area pixels E Area percent E Show Slide Info Compactness E Minor Axis Major Axis F Axis Ratio Figure 45 View Editor Tissue Segmentation Data Displaying the Extracted Data 16 5 Viewing the Cell Segmentation Data Table The Cell Segmentation Data Table is available when an algorithm contains a cell segmentation step The table provides information on the segmen
106. o the desired tissue category 2 Click and drag to outline the desired area on the image 3 Ifthe region intersects another region in the same tissue category the two regions are merged into one region za inForm User Help If the new region intersects a region in a different tissue category the area becomes part of the new tissue category The unselected parts of the image file are covered with a gray mask This may make it difficult to see the image to select additional tissue categories To temporarily hide the tissue categories and the gray mask click the Show Hide Tissue Segmentation Map button After the next tissue category is drawn on the image the tissue segmentation map displays again See Editing Tissue Category Regions 48 if necessary When all of the desired tissue categories have been drawn continue with the next processing step in the project Either click the Advance button or click the next step button 9 3 Editing Tissue Category Regions After tissue categories have been drawn on an image you can edit the regions to add areas to a region remove areas from a region or remove individual pixels from a region To add an area to manually selected tissue categories 1 2 3 Zoom in or out on the image to the desired magnification Click the option button next to the tissue category that you want to add an area to lf desired use the Tissue Segmentation Map button to show or hide the tissue masks Hid
107. ocessing 97 To draw processing regions on an image 1 In the open project click the Single Mode button and then click on the desired image in the image thumbnails 2 If desired use the Zoom In button to zoom in on the image or the Pan button to move the image in the image display area 3 Click the Draw Processing Region button on the toolbar 4 Click and drag to outline the desired area of the image Release the mouse button to connect the endpoint to the start point of the processing region An ID number is assigned to the processing region and displays in the middle of the region The ID number identifies the region in data tables 5 Repeat step 4 to draw multiple processing regions on the same image 6 Click on each image and repeat steps 2 through 5 to define all desired processing regions on each image in the project Deleting Processing Regions If a processing region doesn t enclose the desired region delete the processing region and draw a new processing region To delete a processing region 1 Click the Remove Processing Regions button on the toolbar 2 Click on the processing region that you want to delete The processing region is removed and the remaining processing regions with ID numbers higher than the deleted region are re numbered You can also delete a processing region by right clicking on the processing region and selecting Delete Region on the shortcut menu To delete all processing regions on a
108. olumn starting with the cell just entered 4 Type CTRL D to fill the column range with the desired formula Saving the Table You can save the entire Excel workbook and or export the PivotTable data as a text file for further processing with other programs If exporting the PivotTable you may want to refine the table for cleaner output Below are some suggestions Remove the Grand Total Row at the Bottom 1 Open PivotTable Options by clicking the Options button at the left of the PivotTable Tools Options ribbon 2 Inthe Totals amp Filters tab deselect Show grand totals for columns Display Fractional Area To display the tissue fraction as a decimal fraction rather than as a percent click the drop down next to the field name in the Values box Select Value Field Settings In the dialog box click Number Format and set the desired formatting o inForm User Help Save as Tab Separated Values To save the file as tab separated text 1 Inthe File ribbon choose Save As 2 Select Text Tab delimited in the Save as type drop down list To save as a CSV file select CSV Comma delimited 3 Type the desired file name select the desired location and click Save 4 Two warning messages display First The selected file type does not support workbooks that contain multiple sheets Click OK Then FileName may contain features that are not compatible with Text Tab delimited C
109. ominal Level slider specifies the lower threshold for the 50th bin The remaining thresholds are equally spaced between zero and the 50th bin threshold For display the 50 values are summed into 10 values each value being the sum of 5 bins The actual values for all 50 individual bins display in the Score table inForm User Help Score IHC or IF Settings research use only Tissue Category Category2 Z Scoring 50 bin v Marker Settings Compartment Nuclei Cyto Memb Component ER DAB Threshold Max 0 30 z 50th Bin Nominal Level Optical Density AL g 0 19 Change Color 15 6 10 EE 11 15 16 20 21 25 26 30 31 35 36 40 Figure 40 50 bin Scoring 15 Counting Objects Count Objects locates individual irregularly shaped objects within an image field or within a selected tissue category if tissue has been segmented 1 10 Click the Count Objects button in the step bar to display the Object Counting Settings panel B Count Objects If tissue has been segmented select the Tissue Category in which you want to find the objects Objects outside of the selected tissue category are not counted This option only displays if the project includes a Manual Classification or Segment Tissue step Select the desired Approach to use to identify objects Object Based or Pixel Based Threshold Select the Pixel Based Threshold approach when there is a reliable or consiste
110. onent ER DAB positivities and selected denominator 7 Hematoxylin Component Stats Position Stats In Channel Table Shows the positivity of Min F X Postion one component within another for all selected i Mean Mins colocalization components alin E Ran Std Dev Distance from Process Region Edge pixels e Percent Stats Shows the percentage of ka inter component positivity Mna a a P x Show Slide Infi Compact e Pixel Stats Shows the pixel count of the ieee en a component positivity F Major Avis E Axis Ratio Quant Table Shows stats for the whole image and colocalization region Components Select all or individual components for which to display component statistics One column per numerator component is shown Component Stats Select the values to display for each component signal See Component Stats 8 for descriptions of each option TMA Core Info Select the options to include in the data table See TMA Core Infol 871 for descriptions of each option Position Stats Select which values to display for each segmented object See Position Stats 87 for descriptions of each option Shape Stats Select the options to include in the data table See Shape Stats 8d for descriptions of each option Figure 48 View Editor Colocalization Data 16 8 Viewing the Count Data Table The Count Data Table is available when an algorithm contains an Object Segmentation step Select Count Data in the Data
111. oolbar enables you to adjust the image zoom level draw processing regions draw training regions and adjust the view settings The tools available on the toolbar change based on the current processing step The toolbar displays in the Manual Analysis tab D Select Return the cursor to the default pointer mode View Component Data Hover over the image to display the signal intensity of each component at any location Zoom In Click on the image to re center the image and to zoom in Click and drag to zoom in on a specific region of the image e E Zoom Out Click on the image to re center the image and to zoom out Double click the image with this tool to return magnification to 100 Zoom to Full Click to view the image at 100 Pan Click and drag the image to view the desired region in the window Draw Processing Regions Click and drag to outline the desired processing region Only the areas of the image within the processing regions are analyzed Delete a Region Click inside a processing or training region to remove the region Draw Training Regions Click and drag to identify regions of the image that represent the selected tissue category when training the tissue segmenter This tool selects automatically when you select a tissue category from the category list When in draw mode you can continue drawing training regions Only available during the Segment Tissue step N Qs a A Draw Classification Regions Click and
112. open in Gallery mode as shown in the figure below The View Editor and the Available Images window close if either one was open in the previous project r 1 inForm 2 0 Manual Analysis New Project File Edit Views Tools License Help Image Size l J 5 Multispectral im3 Brightfield Change resolution No change Browse E Review Merge Batch Analysis Manual Analysis oD Peak Vectra 2 Tutorial _BF_Batch2_Tissue1_HP_IM3_11_ 21796 7 16613 4 im Vectra 2 Tutorial _BF_Batch2_Tissue1_HP_IM3_15_ 18343 5 17645 4 im Editing functions are not available in Gallery Mode Project Galle TE Prepare Al p 228 J Y Figure 7 Gallery View 2 To edit algorithm settings click the Single button to change to Single mode Saving Projects Saved projects include the images the algorithm to process the images and any processing regions training regions or manual tissue regions 1 Select File gt Save gt Project 2 Type a new file name if desired select the desired location and click Save A project file ifr is saved in the specified location A folder named lt ProjectName gt _Traininglmages is created in the same folder as the project file Local copies of all the images in the project are saved in the training images folder To save only the algorithm and not the images see Saving an Algorithm 301 Changing Settings After a project
113. order and the components are listed in order in the description tag of each of the component files in the stack Select the desired Maps to Export Maps are exported alone without the original image behind the map Note that the Cell Segmentation Map is a label image that can be used in other image processing tools Label images are 16 bit gray scale TIFF images where all pixels from a segmented object have the same unique gray scale value up to 65 000 objects Select the desired Tables to Export The tables available for export depend on which steps are included in the algorithm or project Data tables are Export Settings Export Directory C Program Files Caliper Lie Browse Image export options image Outpt Format ner images to export RGB AGB wth Tissue Segmertation Map AGB wah Cal Segmertation Map RGB wth Scoring Map RGB wth Al Segmertation Maps Composte image Componert images uuiqnage TIFF Maps to export J Tissue Segmentation Map Cel Segmentation Map TIFF Tables To Epot J Tissue Segmentation Data Cal Segmentation Data Score Data Table Relds To Export Epot al feks Use view settings Fie name options in case of fle name corfict Overwrte eosting fies Rename new files Directory Structure Copy 0 leveli of directory structure from source mage Figure 51 Export Settings Exporting the Data exported to tab separated text files which can be opened in Microsoft Excel
114. ou are solely responsible for determining the appropriateness of using or redistributing the Work and assume any risks associated with Your exercise of permissions under this License 8 Limitation of Liability In no event and under no legal theory whether in tort including negligence contract or otherw ise unless required by applicable law such as deliberate and grossly negligent acts or agreed to in writing shall any Contributor be liable to You for damages including any direct indirect special incidental or consequential damages of any character arising as a result of this License or out of the use or inability to use the Work including but not limited to damages for loss of goodwill w ork stoppage computer failure or malfunction or any and all other commercial damages or losses even if such Contributor has been advised of the possibility of such damages 9 Accepting Warranty or Additional Liability While redistributing the Work or Derivative Works thereof You may choose to offer and charge a fee for acceptance of support warranty indemnity or other liability obligations and or rights consistent w ith this License How ever in accepting such obligations You may act only on Your own behalf and on Your sole responsibility not on behalf of any other Contributor and only if You agree to indemnify defend and hold each Contributor harmless for any liability incurred by or claims asserted against such Contributor by reason of
115. ow are always visible in the Cell Segmentation Data table e Cell ID Displays a unique ID number for each cell based on the location of the nuclei within each tissue category if tissue has been segmented e Total Cells The total number of cells scored in the image e Tissue Category Area Pixels If tissue has been segmented either automated or manual the area of the selected tissue category measured in number of pixels If there is no tissue segmenting step this column is the total number of pixels in the whole image e Cell Density per megapixel If tissue has been segmented either automated or manual displays the cell density for each tissue category the number of cells in the tissue category area the number of pixels in the tissue category area 1 000 000 If there is no tissue segmenting step displays the cell density for the entire image Total cells the number of pixels in the image 1 000 000 Displaying the Extracted Data iat 16 6 Viewing the Score Data Table The Score Data Table is available when an algorithm contains a Score step The available data changes depending on the type of scoring Select Score Data in the Data Displayed list then choose the desired options under Table Contents Each option you select adds a data column to the table Table Contents Select the options to include in the p data table See TMA Core Infol s for descriptions of Dien Econ Data Displayed each option
116. pening images IN a PrO SCt corsier a E A EARE EAA ERATA 31 2 Viewing Magos ensec e EEE E ENEE EAEE EEEE NEEE EEE AEE NE 32 Chapter 7 Drawing Processing RegionS sssssssssssnnnsrnnnnnnnnnnnnnnnnnnnnnnnnnnnnnn nennen 34 Chapter 8 Preparing ImageS ssssssssssesrnnnununnennnnnuunnnnnnnnununnnnnnnnnnnnnnnnnnnnnnnn nenna 35 1 Selecting the Image Format ucrssieissinsisesiese nae n E aE EE E E 35 2 Loading a Spectral LID Ary ascensio EE E E EEEE EE EEE EEE EN 36 3 Miewinig a Spectral Daear a ER 39 4 Gomponent UNIS eceescansrencain eein EE EEES EEEE EREE EEE 39 5 Completing the Prepare Images Step cceececceeeeeeeeeeeeeeeeaaeeeeeeaaeeeeeeaaeeeeeaaeeeeeeaaeeeeeeaaeeeeees 43 Contents 3 Chapter 9 Manual Tissue Segmentation cccceesesseeeneneeseseeeeeeeneessneeeeeeennnnees 44 1 Creating Tissue Categories cece cece eceeeeeeeeeeeeceaeeeeeceaaeeeeceaaeeeeceaaeeeeaaaaeeeeaaaaeeesaeaaeeesaaaaees 44 2 Drawing Tissue Category Regdi niS escrire ciar aaa R EARE RERAN ER EE EN 44 3 Editing Tissue Category Rogno ssessssrree insien a AER ESS AKEE EAEE EEKE 45 Chapter 10 Trainable Tissue Segmentation ssssssuusssssrnnunnnnnnnnnnununnennnnnnnnnne nnmnnn 47 1 Adding Tissue Categories cc eececcceeeeeeeeeeeeeeeeeeeeee ee eeeeeeeeeeeeeeaaeeeeaaeeeeesaeaeeeeeaaeeeeeeaaeeeeesaaeeees 47 2 Drawing Training ReQiOns ceeceeeeceeeeeeeeeeeee tees eeeeeeeeeeaeeeeeeeaaeeeeseaaeeeeeeaaeeeeeeeaeeeeaeaeeeeseee
117. ple If an image pixel expresses one part Fluor A and three parts Fluor B the signal at the pixel is shown in the figure below Measured Response at Image Pixel 1 257 Signal 1 00 5 0 50 4 r r 500 800 700 Wavelength Figure 16 Signal at a Pixel Unmixing recovers the original spectra scaled according to the measured expression at the pixel as Preparing Images shown in the figure below Components of Unmixed Signal Fluor A Fluor B 1 274 Signal 0 475 500 600 Wavelength Figure 17 Original Spectra Scaled The challenge for the unmixed component image is to represent the unmixed spectrum using a single value The three weighting options correspond to three possibilities for how to do this e Peak weighting represents the spectrum as the signal level at the brightest wavelength of the spectrum In the example this gives component values of Fluor A 0 5739 Fluor B 1 5216 e Total weighting represents the spectrum as the sum of the signals at all wavelengths In the example this gives component values of Fluor A 2 551 Fluor B 7 7514 e Mean weighting represents the spectrum as the average signal across wavelengths In the example this gives component values of Fluor A 0 3189 Fluor B 0 9689 Notice that Total weighting is the Mean weighting multiplied by the number of wavelengths eight in this example The figure below shows scaled spectra annotated wit
118. r Excel 2010 and may need modification for other versions of Excel Create a PivotTable to Compute Fractional Tissue Area To create the PivotTable 1 Start Excel and open the data file In the Open dialog select Text Files in the file type drop down then select the inForm data file The file name should end with tissue_seg_data_summary txt 2 Inthe Text Import Wizard if you are using a file from inForm V1 4 change the value for Start import at row to 3 This skips the first two rows For inForm V2 0 or higher leave the value at 1 Click Finish to open the file 3 In the Insert ribbon click PivotTable Leave the defaults use the entire table and put the table on a new worksheet Click OK to create the PivotTable A new empty PivotTable opens On the right side the PivotTable Field List shows all the fields in the source table 4 Click the check box next to the name of the field to summarize by usually Sample Name or Slide ID It could also be TMA row and column The selected field is shown in the Row Labels box and as a column in the pivot table 5 In the PivotTable Field List click the check box next to Tissue Category to add Tissue Category to the Row Labels list 6 Drag Tissue Category from Row Labels to Column Labels Appendix A Calculating Fractional Tissue Area 7 Some tissue categories in the data may not be used in the tissue fraction calculation for example a Background
119. r all within a selected category 4 When you are finished drawing regions for the first category switch to the next category e g non tumor regions and draw regions around cells characteristic of the next tissue category Be sure to capture visual anomalies such as areas slightly out of focus folds bubbles etc Also capture stroma regions that have differing qualities of shape or orientation 5 Repeat this process for all of the images in the training set 6 When all training regions have been drawn see Training the Tissue Segmenterl 49 as inForm User Help 10 3 Training the Tissue Segmenter Only images that are included in the Training set are used to train the tissue segmenter Images that have been removed from the training set are not used to train the tissue segmenter even if the images include training regions Selecting Components for Training Select the components that you want to use for training the tissue segmenter commoners foros V ER DAB Components that provide distinguishing pattern information that 7 Hemataxyiin is independent of signals to be quantified are preferred For lees example the counterstain component plane e g DAPI or Hematoxylin should be tried first for training because the Pattem Scale Large x counterstain reveals the structure of the tissue and is less likely to produce a segmenter biased by the presence of the _ Tran Tissue Segmenter molecular marker you a
120. r instructions Manual Classification Use to create tissue categories and manually draw tissue regions on the images in the project Exported data includes position shape and y S amp S component signal statistics for each segmented tissue region See Manual Tissue Manual 9 Segmentation 44 for instructions This feature is available with inForm Basic Analysis 14 inForm User Help or higher Segment Tissue Use to train the automated tissue segmenter and automatically segment the images Exported data includes position shape and component signal statistics for each segmented tissue region See Trainable Tissue Segmentation 47 for instructions This feature is only available with inForm Tissue Finder Threshold Use to create and analyze threshold maps Exported data includes O position shape and component signal statistics for each thresholded region See Threshold 54 for instructions This feature is available with inForm Basic Analysis or Threshold higher position shape and component signal statistics for the colocalization region as well as position shape and component signal statistics for the underlying component Colocalize regions See Colocalization 5 for instructions This feature is available with inForm Basic Analysis or higher Colocalize Use to create and analyze colocalization maps Exported data includes Segment Cells Use to segment cell nuclei and optionally segment cytoplasm
121. re The columns below are always visible in the Score Data TETERE table Show Core ID Show Slide Info Tissue Category If tissue has been segmented either automated or manual the tissue category used for scoring displays in the table If there is no tissue segmenting step this column is blank Tissue Category Area Percent If tissue has been segmented either automated or manual the area of the selected tissue category displays in the table If there is no tissue segmenting step this column is 100 the whole image Number of cells The number of cells scored in the image First Cell Compartment The cell compartment used for scoring First Stain Component The stain used to create the threshold for scoring Figure 47 View Editor Score Data The following additional columns are visible when Positivity 2 bin scoring is selected e Positivity The scored positivity e Positivity Threshold The threshold value used to set the positivity The following additional columns are visible when Double Positivity 2x2 scoring is selected Second Cell Compartment The secondary cell compartment used for scoring Second Stain Component The secondary stain used to create the threshold for scoring Double Negative The percent of cells that is negative less than the threshold for both components Single lt Componentt gt The percent of cells that is positive for component 1 and negative
122. re trying to quantify For example if Recent Trainings Attempt 1 99 94 v measuring the positivity of cells stained with Alexa Fluor in a sample stained with both DAB and Alexa Fluor you would not want the segmenter to be dependent on the Alexa stain Segmentation Resoktion iad assis because that would bias the outcome Hi Tie Egos By cae 6 Tissue Category Categoyi sy V Minimum Segment Size pixels 500 Discard if touching image border Figure 21 Components for Training On the other hand sometimes in brightfield applications where the presence of an IHC stain such as DAB blocks the staining of the counterstain using both the counterstain and the IHC stain component planes for training yields a more robust tissue segmenter even for the negative cases In these cases where the target label is included in the training it is important to include abundant negative cases in the training set Also you should generally avoid using components that provide information unrelated to the desired segmentation Selecting a Pattern Scale The recommended approach is to start with the default Large pattern scale Once you have drawn training regions and trained a tissue segmenter if segmentation accuracy for the training regions is low lt 85 increasing Pattern Scale may improve accuracy However the pattern scale needs to be compatible with the size of the training regions Pattern scale refers to the spatial extent of
123. reuse the same settings to process compatible images or sets of images without re selecting the settings for each new project Saving the project saves the images with the steps see Saving a Project 29 To save the algorithm select File gt Save gt Algorithm Or to save the algorithm in a protected format that cannot be edited select Locked Algorithm Type the desired algorithm name select the desired location and then click Save 30 inForm User Help 6 Opening Images The topics in this section describe how to open images in a project and describes the viewing options available 6 1 Opening Images in a Project Add the images to the project as described below The images in a project can be included in the Processing set the images to be processed by the algorithm and or can be included in the Training set if training a Tissue Segmenter The number of images that can be opened in a project depends on the size of the images width and height in pixels and the operating system Windows 7 32 bit systems can open up to 20 Nuance sized images 1392x1040 Windows 7 64 bit systems can open up to 400 Nuance sized images Opening Images To open the desired images 1 Select File gt Open gt Image The Select Image window opens The Files of Type selector at the bottom of this window specifies the types of image files to display Multispectral files im3 TRIO images im4 or other common image files such as TIFFs B
124. ring is selected e Bin X where X ranges from 1 10 or 1 50 per scoring type Columns showing the percent of scored values that fall in each scoring bin The user sets the lower threshold for the last bin The last bin is the percent of cells whose component strength is greater than the threshold the user set The thresholds for the remaining bins are spaced evenly from 0 For example the 49th bin has a lower limit of 48 threshold 49 Displaying the Extracted Data s 16 7 Viewing the Colocalization Data Table The Colocalization Data Tables are available when an algorithm contains a Colocalization step Select Colocalization Data in the Data Displayed list then choose the desired options for each table aa inForm User Help The following tables are included in the ap vi Edi SEE colocalization data Gadi Data Displayed Colocalization Table Shows how much the sseslzstion Data Z 2 numerator components are colocated with the oo sion Tene denominator 7 Colocalization Table V Percent Stats V Pixel Stats e Percent Stats Shows the percent colocalization where all colocalization 7 In Channel Table pixels are positive within the selected EI Pacar Siats F Pool Stats denominator and the percent positivity of piacere each selected component in colocalization mponents All Components e Pixel Stats Shows the actual pixel count 7 Her2 Fast Red for the colocalization comp
125. rties relating to the subject matter hereof No modification or waiver of any provision of this Agreement shall be effective unless in writing and signed by the parties No delay or failure on the part of any party in exercising any right under this Agreement shall impair any such right or any remedy of such party nor shall it be construed to be a waiver of any continuing breach or default under this Agreement In the event any provision of this Agreement is held to be unenforceable the remaining provisions of this Agreement will remain in full force and effect This Agreement shall be governed by the laws of the State of New York without regard to principles of conflicts of laws Any disputes relating hereto shall be adjudicated only in the state or federal courts in New York County New York State and you hereby consent to the exclusive jurisdiction of those courts This Agreement shall not be governed by the United Nations Convention on Contracts for the International Sale of Goods the application of w hich is expressly excluded You may not assign or otherwise transfer this Agreement or any of your rights or obligations therein without the prior written consent of CRI You may not use the Software for any unlawful purpose nor export or re export the Software except as authorized by law 10 TIFF module This software uses the LibTIFF 4 0 3 library http www libtiff org misc html The TIFF module used is Copyright 1988 1997 Sam Leffler Copyright
126. run can be merged into a single file For example all Cell Segmentation data files can be merged into a single file that contains the Cell Segmentation data from all the images in the batch 1 After running a batch click the Review Merge Tab 2 The Batch directory defaults to the directory of the 3 amde T z current batch process or the last directory used for a poms batch processing Click the Browse button to choose E 7 rar CADE a a different directory of images slides and data l aie j zga 3 The data sets that can be merged display in the list H All data sets are selected by default p 4 Review the list of data sets and select the data sets to be reviewed e Select or clear the check marks in the list to F include or exclude specific data sets e Click the Include All button to select alldatasets in the list e Click the Include None button to clear the check boxes for all data sets in the list 5 When a data set name is selected in the list Inchude Al lnchsde Hone thumbnails of the exported images for each data set display at the bottom of the window Click on a thumbnail to view the image in the large image display area Viewing image 1 of 2 Merge Figure 54 Review Merge Tab 6 Review the images and accept or reject each data set e Use the left or right arrow buttons next to the Gallery thumbnails to scroll through the exported images for each data set e Click the Yes button to include the
127. s for descriptions of each option Figure 49 View Editor Count Data Shape Stats Select the options to include in the data table See Shape Stats 88 for descriptions of each option Displaying the Extracted Data e 16 9 Viewing the Quant Data Table The Quant Data Table is available when an algorithm contains a Threshold or Colocalization step Select Quant Data in the Data Displayed list then choose the desired options under Table Contents Each option you select adds one or more data columns or rows to the table The following options are available for the Quant Data p gt 9 op Q i View Editor So Table fer ata Displaye Components Select all or individual components for Quant Data zj o which to display component statistics Each selected EER component creates one column for each component stat components selected z a V Her2 Fast Component Stats Select the values to display foreach H 53 248 component signal See Component Stats 87 for Laman er Component Stats Position Stats descriptions of each option ii FI X Fodion V Mean Y Position TMA Core Info Select the options to include in the 5 z EE data table See TMA Core Infol 87 for descriptions of F Sd Dev SEE AEA aE each option Total TMA Core Info Shape Stats Position Stats Select which values to display for each Show Core ID Area pixels segmented object See Position Stats 8 for Show Side Info Co
128. s If selected the score colors display on the image and around the results percentage text boxes If not selected the colors do not display on the image or around the results percentage text boxes Change Color Opens the Change Color window to select the colors for each bin or to restore the default colors to the bins When you score the image s the results display at the bottom of the Score IHC or IF settings panel See Scoring Types 6 for descriptions of the results for each scoring type 6s inForm User Help Scoring Types Positivity This bins the spectrally unmixed nuclear cytoplasm or membrane signals into two bins negative or positive with a single threshold See Figure 30 above It provides data in percent Decreasing the Positivity Threshold slider increases the percent positivity The Score Results Percent box shows the negativity and positivity percentages of the cell nuclei cytoplasm or membrane Double Positivity 2x2 bin Select this scoring method to classify cells according to co expression into four classes double negative single positive for one marker single positive for another marker or double positive The Negativity and Positivity percentages for each of the two markers are also shown Double Positivity contains the same options as Positivity with the addition of the following Second Marker Settings Select the Compartment Component Threshold Max and Positivity Thr
129. s 7 Fifty bin scoring 70 Fill holes 62 Fine tuning segmenter 51 Fixed scale 60 Fractional Tissue Area calculating 97 H Haze spectra removing 38 HPF finder algorithm 94 ID number 34 IF scoring 66 IHC scoring 66 Image Display area 18 Index 107 Image Displaying 73 Format 35 Viewing 32 Image analysis tasks Summary of 26 Images Adding to a project 31 Adding to a training set 51 Opening 31 Opening ina project 31 inForm Algorithms 94 inForm window 14 Inner distance to nucleus 63 Installing inForm 11 Introduction to inForm 7 K Key Terms 10 L License activating 12 Masks editing re training 52 Maximum size 61 65 Maximum size pixels 62 Membrane segmenting 64 Merge editor 19 Merging the data 93 Minimum size 61 63 65 N New project 27 New Project Window 22 Nuclear segmentation components 61 64 Nuclei size range 61 65 Nuclei segmenting 59 0 Object based approach 59 Objects counting 71 inForm User Help Online documentation 11 Open Algorithm 30 Project 29 Opening images 31 Opening images in a project 31 Optical density converting to 37 Outer distance to nucleus 63 P Pattern scale selecting 49 Percent total area 97 PerkinElmer Technical Support 13 PivotTable 97 Pixel based approach 59 Positivity scoring 67 Printed documentation 11 Process step bar 14 Processing Regions 31 Deleting 34 Drawing 34 Project 10 Adding images 31 Creatin
130. s only one imaging band when the library is for a brightfield image or when the units are OD e Normalized Each chart is scaled individually on the Y Axis to 1 Additionally the signals are scaled so each spectrum has a maximum of 1 This option is not available when units are Unmix Units Save Chart Button Opens the Save Chart window to specify the location name and file format for saving a graphic of the charts displayed in the Spectral Library window The chart can be saved as a bmp gif png tif or jpg and saved to a computer or network location The graphic file includes the legend Legend Displays the color for each stain in the chart The colors are selected when the Spectral Library is created and cannot be changed in this view Understanding the inForm Work Area 25 Common Image Analysis Tasks Use inForm to create automated image analysis routines to accurately extract data from PerkinElmer s TRIO Nuance or Vectra images or conventional RGB images TIF BMP or PNG While it is also possible to process JPG images this is not recommended JPGs can contain compression artifacts that could adversely affect segmentation performance Some common inForm image analysis tasks inForm performs pixel based analysis to quantitatively obtain component data inForm can set thresholds pixel based for all component signals and quantitate the thresholded component signals inForm performs colocalization which locates
131. sing or publicity relating to this softw are or products derived fromit This softw are may be referred to only as the Independent JPEG Group s softw are We specifically permit and encourage the use of this software as the basis of commercial products provided that all warranty or liability claims are assumed by the product vendor 18 This software uses the zlib1 1 2 7 library http www zlib net This softw are makes use of the zlib1 library This library has not been modified by CRI and all rights are reserved by the copyright holder This softw are is provided as is without express or implied w arranty and with no claim as to its suitability for any purpose The library is covered by the zlib1 License The zlib 1 license is reproduced below C 1995 2012 Jean loup Gailly and Mark Adler This softw are is provided as is without any express or implied warranty In no event will the authors be held liable for any damages arising from the use of this softw are Permission is granted to anyone to use this softw are for any purpose including commercial applications and to alter it and redistribute it freely subject to the follow ing restrictions 1 The origin of this softw are must not be misrepresented you must not claim that you wrote the original software If you use this software in a product an acknowledgment in the product documentation would be appreciated but is not required 2 Altered source versions must be pla
132. spectral nay Sample Fomat Brightfield z C Change resolution Factor Nochange Spectral Library Browse Spectra for Unmixing Review Merge Batch Analysis Manual Analysis Select Reported Units Weighting oD gt Pex gt m ectra 2 Tutorial _BF_Batch2_Tissue1_HP_IM3_11 21796 7 16613 4 im Vectra 2 Tutorial _BF_Batch2_Tissue1_HP_IM3_15_ 18343 5 17845 4 im Project Gallery rom na Figure 8 Gallery View Single Mode To view an image in Single Mode click the Single button at the lower left of the display area E Click on the thumbnail of the image to view the image in the display area Drama Mena maps a File Edit Views Tools License Help Image Preparation Settings Format Multispectral im3 ig i i Spectral Library Spectra for Unmbing Review Merge Batch Analysis Manual Analysis Figure 9 Single View Opening Images 33 7 Drawing Processing Regions Processing regions enclose areas of the image that will be processed using the steps in the project or algorithm Areas of the image outside of the processing regions are ignored and are not included in any calculations or statistics Processing regions are manually drawn on the images and cannot be used for Batch Pr
133. splay the Object Segmentation Map on the image Only for projects that include a Count Objects step a6 inForm User Help e Training Regions Select to display the Training Regions on the image Only for projects that include a Segment Tissue step e Processing Regions Select to display the Processing Regions on the image e Equalize Display Histogram Select to map the pixels so the image histogram has approximately the same number of pixels assigned to each tonal value of the histogram This gives the best display of the whole dynamic range of dim and bright signals If not selected the lowest 0 01 of the pixels are mapped to 0 the highest 0 01 are mapped to 255 and the remaining pixels are linearly interpolated between those values This prevents a few bright or saturated pixels from skewing the display See Displaying the Extracted Datal 73 for descriptions of other options available for each image type or data table 16 11 Component Stats The data tables contain the options below e Min Displays one column containing the minimum value of the component in the region and one column displaying the e Mean Displays the mean of the component in the region The summary line displays the weighted mean value of the component across all regions which is the average of the region means weighted by region size e Max Displays the maximum value of the component in the region e Std Dev Displays the standard deviat
134. splays a restart message 12 Click the OK button inForm restarts automatically The purchased options are now available in the software Transferring a Software License You can transfer the license from the current computer to another computer When the license is transferred to the new computer the license is removed from the current computer leaving only the inForm Viewer options Transferring a license requires an administrator account on both machines a USB thumb drive and an available USB port on each machine To transfer the license log into the unlicensed computer with an administrator user name and password Start inForm and select License Activate Choose Transfer Follow the instructions provided in the dialog boxes to complete the transfer 12 inForm User Help 2 7 Contacting PerkinElmer For more information contact PerkinElmer or your local authorized PerkinElmer distributor PerkinElmer Inc 68 Elm Street Hopkinton MA 01748 USA Phone 800 762 4000 or 1 203 925 4602 Fax 1 203 944 4904 PerkinElmer Web Site http www PerkinElmer com Technical Support Email amp global techsupport perkinelmer com Introduction 3 Understanding the inForm Work Area The inForm work area contains the step bar process editor panel image display toolbar and image display area The example below shows the Manual Analysis tab with a brightfield image that has been segmented into two tissue categori
135. t form provided that You meet the follow ing conditions a You must give any other recipients of the Work or Derivative Works a copy of this License and b You must cause any modified files to carry prominent notices stating that You changed the files and c You must retain in the Source form of any Derivative Works that You distribute all copyright patent trademark and attribution notices fromthe Source form of the Work excluding those notices that do not pertain to any part of the Derivative Works and d If the Work includes a NOTICE text file as part of its distribution then any Derivative Works that You distribute must include a readable copy of the attribution notices contained w ithin such NOTICE file excluding those notices that do not pertain to any part of the Derivative Works in at least one of the follow ing places within a NOTICE text file distributed as part of the Derivative Works within the Source form or documentation if provided along with the Derivative Works or within a display generated by the Derivative Works if and wherever such third party notices normally appear The contents of the NOTICE file are for informational purposes only and do not modify the License You may add Your own attribution notices w ithin Derivative Works that You distribute alongside or as an addendum to the NOTICE text from the Work provided that such additional attribution notices cannot be construed as modifying the License You
136. t here makes it easier to find the algorithm from the Vectra software The file name should identify it as the HPF Finder algorithm for the specific set of tissue slides 7 The Tissue Finder and HPF Finder algorithms can now be selected in the Vectra tissue protocol 96 inForm User Help 21 Appendix A Calculating Fractional Tissue Area Fractional tissue areas can be calculated from inForm data using an Excel PivotTable Examples of this type of calculation are e Tumor load tumor area tumor stroma e Lung fibrosis fibrosis area normal lung normal collagen fibrosis e Breast tissue lobular involution acini area acini inflammation lobular stroma It is also possible to calculate tissue fractions where the numerator and denominator are different tissue categories such as e Kappa lambda ratio in lymphoid tissue kappa tissue area lambda tissue area This calculation can be performed per image or per slide with inForm V2 0 1 or higher and Vectra Initial Setup These instructions assume that you already have an inForm algorithm capable of automatically segmenting the tissue categories of interest and that you have processed several images using the algorithm and merged the tissue segmentation tables into a single table To compute tissue fraction by slide you must have a column in the data table for slide ID inForm includes this column for Vectra images The instructions in this section are fo
137. t the Scale Views Based on Selected Images option The scaling limits are set based on the positive control images that are open in the project Save the algorithm and or the project Use the settings for batch processing by loading the algorithm or project in the Batch tab If the positive controls change open the project and click the Reset button on the View Editor Select the new positive controls in the Scaling Image Selection windowl 7e to set the new scaling limits Displaying the Extracted Data Scaling Image Selection Window Use the Scaling Image Selection Window to choose the images used to scale the brightness of all images in the project The images that are brightest in at least one component are at the top of the list and are selected by default Images must be unmixed to be used for the scaling calculations If an image is not unmixed it displays at the bottom of the list and the Select For Scaling check box is disabled Select For Scaling Check boxes The scaling limits are calculated using all selected images These scaling limits are saved and used when processing images in Batch Mode Use the Thumbnail and Image Name columns to identify the images in the project The Image Name column also identifies images that are the brightest for each component Cancel button Closes the Scaling Image Selection Window without recalculating any image scaling limits The limits do not change Done button Recalc
138. tains an object segmentation step The table provides information on the segmented objects See Viewing the Count Datal 85 for detailed information Score Data This table is available when an algorithm contains a score step The available data changes depending on the type of scoring positivity 0 3 10 bin 50 bin and Double Positivity See Viewing the Score Datal 82 for detailed information Views Available In Tissue Finder Includes all views in Cell Analysis Tissue Segmentation data displays in the Tissue Segmentation Table See Viewing the Tissue Segmentation Datal 78 for detailed information 2 inForm User Help 3 2 Available Images Window Use the Available Images Window to view a list of ie Sax all images in the project and to add images to or remove images from the processing set or the Process Training image training set The training set is only supported in a Bi Vocke 2 Titosal BF Batch Tamve 1 HP_O03_11 1717567 inForm Tissue Finder when training the Tissue m Segmenter Click the Image List Editor button to open the Available Images window Je LA d He To load additional images into the project click the Load Image button The new images must be compatible with the Prepare Images settings in the current project New images must have the same spectral wavelength characteristics as the existing images in the project To remove an image right click on the image name and click
139. ted cells Select Cell Segmentation Data in the Data Displayed list then choose the desired options under Table Contents Each option you select adds a data column or row to the table Cells are assigned sequential ID numbers according to the location of nuclei within each tissue category if tissue has been segmented ID numbering restarts at 1 in each tissue category The table only shows statistics for cell compartments that have been segmented nuclei cytoplasm membrane For example if cytoplasm is not segmented the Cytoplasm column is not included in the table The following options are available for the Cell Segmentation Data Table Components Select all or individual components for which to display component statistics Each selected component creates one column for each component stat selected If there is no data for a component the component will not be displayed in the list Tissue Categories Select all or individual tissue categories for which to display data If selected a Tissue Category column displays in the Cell Segmentation table to specify the Tissue Category for each cell This selection does not display if there is only one Tissue Category defined or if there is no Segment Tissue step in the project Component Stats Select the values to display for each component signal See Component Stats 87 for descriptions of each option Each selection adds one column for each segmented cell component and one column
140. the components that you want to unmix 5 To view the selected Spectral Library click the View button under the Spectra Unmixing list box and see Viewing a Spectral Library 39 36 inForm User Help Converting to Optical Density Brightfield multispectral and RGB images of IHC stained samples need to be converted to optical density if the goal is to quantify stain intensity optical density Typically brightfield multispectral images acquired with Nuance and recent versions of Vectra are already converted to optical density and do not need conversion as part of Prepare Images Brightfield multispectral images acquired with early versions of Vectra are not converted to optical density when acquired and require conversion as part of Prepare Images Selecting a Reference White for Optical Density conversion V Convert to Optical Density y3 Figure 12 Convert to Optical Density 1 Select the Convert to Optical Density check box 2 If the white areas on the image are small zoom in on a white area 3 Click the White Picker button 4 Move the cursor a white box over a white region on the image and click White Selected displays under the Convert to Optical Density check box If the image appears very light you may have clicked on an area that was not completely white Clear the Convert to Optical Density check box to return to the original image and then repeat the steps above Note The optical densi
141. ting 60 Tissue finder algorithm 94 index 109 For more information contact PerkinElmer or your local authorized PerkinElmer distributor PerkinElmer Inc 68 Elm Street Hopkinton MA 01748 USA Phone 800 762 4000 or 1 203 925 4602 Fax 1 203 925 4904 Email global techsupport perkinelmer com Web site http www perkinelmer com P gt PerkinElmer inForm and Vectra are registered trademarks of PerkinElmer Incorporated Microsoft and Excel are trademarks of Microsoft Corporation in the United States and other countries All other trademarks or registered trademarks are the property of their respective owners 2012 2013 PerkinElmer Inc All rights reserved
142. tissue_seg_map age b Move this file 2 13036_8160_5100_tissue_seg_data txt 13KB Text Document Copy this file 13036_8160_5100_nuc_seg_map tif 2 049KB TIF Image 8 Publish this file to the 13036_8160_5100_image_with_all_seg tif 2 684KB TIF Image Web 13036_8160_5100_cyto_seg_map tif 2 049KB TIF Image 3 E mail this file 13143_12280_26600_cell_seg_data txt 677KB Text Document amp Print this file i 13036_8160_5100_image_with_tissue_seg tif 3 319KB TIF Image X Delete this fle 13036_8160_5100 tif 3 327KB TIF Image Figure 52 Exported Data Files 9 inForm User Help 18 Batch Processing Batch Analysis processes a large number of images quickly and effectively with just one click of the batch Run button Before beginning batch analysis create and test an algorithm or project on a set of images that is representative of the entire data set Adjust the algorithm or project as necessary until the images are processed as desired and then save the algorithm or project Input Files Batch Algorithm or Project rojects Seg Tissue SegCell Score ifr Add Slides Ronove Remove All Export Options C Documents Data inForm V2 0 Projects Export 1 Vectra 2 Tutorial _BF_Batch2_Tisst E Create separate directories for each image C Documents Data inForm V2 0 Projects Eport 1 Vectra 2 Tutorial _BF_Batch2_Tissi Export Directory CAData Image export options Image Output Format TIFF Ima
143. ty conversion is required if you are working with brightfield images and plan to segment cells and measure stain intensity Choosing Spectra for Unmixing The Spectra for Unmixing list box is populated from the selected spectral library Select the check boxes for the signals to unmix For fluorescence images see the Remove Hazel 34 topic first to remove the background haze before unmixing For Brightfield RGB images if you do not load a library the spectra for unmixing default to Blue Green and Red Spectral Library Browse BF_Her2 ER hem _lib csl Spectra for Unmixing V Her2 Fast Red 7 ER DAB V Hematoxylin F Tissue View Select Reported Units Weighting op z Peakur 2 Figure 13 Spectra for Unmixing Preparing Images Removing the Haze Signal In fluorescence images image contrast and unmixing accuracy can sometimes be improved if the general haze Spectral Library Browse in the image which has a characteristic spectra is breast IF 2488 a647 lib csl subtracted from the images first When creating the spectral library for the images create the background or haze spectrum and add it to the library When the library is loaded you can select the haze spectrum and remove it from the images Remove Haze To remove haze O roa 1 Check the Remove Haze check box Rigid 1A Tem OVE maze Spectrum 2 Select the Haze Spectrum to remowe Haze is removed numerically
144. u can select Coarse for the Segmentation Resolution This allows the algorithm to classify images more quickly 8 Click Train Tissue Segmenter to train the new Tissue Finder algorithm When training is finished segment the current image If the segmentation looks good segment all of the images 9 To test and verify the Tissue Finder algorithm a Click the Available Images button to open the Available Images Window 2 b Remove the first set of images from the processing set but leave them in the training set c Load and segment the second set of images using the same algorithm d Click Segment All and examine the resulting tissue segmentation e Ifthe algorithm needs improving draw additional training regions on the second set of images Click Train Tissue Segmenter again and then segment all of the images again Creating Algorithms for Vectra s 10 When you are satisfied with the Tissue Finder algorithm click the Advance button or the Export button Export Settings Export Directory 11 In the Export step click the Browse button to select an export a directory for the segmented monochrome images The remaining export options are automatically selected and cannot be changed 12 Click Export for All to export the segmented monochrome images to the specified directory Q a gt z 3 Z 9 fa 2 z 2 a o 2 5 E ec 13 Select File gt Save gt Algorithm to save the new
145. ulates the image scaling limits using the selected images and Figure 43 View Editor Color Image then closes the window inForm User Help 16 3 Viewing the Composite Image Select Composite Image in the Data Displayed list then choose the desired options The following options are available for the Composite j gt Image 1 View Editor 2 j i Data Dimeyed a Rendering Options Composte z 2 e Show As Rendering Options Show As e Brightfield Select to display the composite Te esata image on a white background Scaling Scale Views for Each Image Individually e Fluorescence Select to display the composite Scale Views Equally for All Images in the Project image on a black background Scale Views Based on Selected Images e Scaling Component Display e Scale Views for each Image Individually Display Intensity True Adjustable Select to show each image scaled individually Dim images are brightened to show the Diy Cpe ss CR ea er coer a components The relative brightness of 7 Herd Fast Red MM Ri components cannot be compared between Intensity g TT images e Scale Views Equally for All Images in the 7 ER DAB E Gree g Project Select to show the brightness of Intensity g 100 i components for each image all scaled relative to each other Bright components appear bright and 7 Hematoxylin MMMM Ee dim components appear dim All images are rescaled relative to all images in the project E
146. umor mask over non tumor cells for example use this tool to draw the desired mask over an area Only available during the Segment Tissue step Select Clear Edits to clear all manual edits to the tissue masks Clear Edits amp View Editor Click to open the View Editor Windowl 19 to choose the type of image or table to view and to change the display settings You can also view or hide cells masks and regions Buttons to quickly show or hide image maps are available on the left side of the image display area Image List Editor Click to opens the Available Images Window 2 to add images to or 99 remove images from the open project the processing set or the training set Show Hide Map Options Click to show or hide the image maps You can show or hide segmentation maps training regions and processing regions The Show Hide options are also available on the View Editor 191 The buttons include 2 Tissue Segmentation Map Click to show hide the Tissue Segmentation Map 2 Nuclear Segmentation Map Click to show hide the Nuclear Segmentation Map e Cytoplasm Segmentation Map Click to show hide the Cytoplasm Segmentation Map E Membrane Segmentation Map Click to show hide the Membrane Segmentation Map Training Regions Click to show hide the Training Regions g Processing Regions Click to show hide the Processing Regions Understanding the inForm Work Area D Image Display Area The
147. ure to choose Yes if you want to be able to create algorithms to use with the Vectra software Starting inForm Select Start gt All Programs gt PerkinElmer gt inForm gt inForm 2 0 2 OR double click the inForm shortcut on the Windows desktop To use any software configuration other than inForm Viewer see Activating the Licensel 12 Introduction 2 6 Activating the License After installing inForm only the options for inForm Viewer are available The customer license must be activated to enable any purchased software options License activation requires an administrator account To activate a customer license 1 Log into the computer using a Windows Administrator account Start inForm Select License Activate or click the Activate button The Activation window opens Enter your 20 character license number If available use the online activation method to complete your activation O oa A WO N If online activation is not available click the Copy to Clipboard button to copy the text onto the Windows clipboard 7 Email the copied text to the email address shown in the Activation window 8 Click the Cancel button to close the Activation window 9 Wait to receive the email that contains the text for the response key 10 Start inForm and select License Activate The Activation Window opens 11 Type or paste the response key from the email into the Activation window and click the OK button inForm di
148. ution as defined by Sections 1 through 9 of this document Licensor shall mean the copyright ow ner or entity authorized by the copyright ow ner that is granting the License Legal Entity shall mean the union of the acting entity and all other entities that control are controlled by or are under common control with that entity For the purposes of this definition control means i the pow er direct or indirect to cause the direction or management of such entity whether by contract or otherwise or ii ow nership of fifty percent 50 or more of the outstanding shares or iii beneficial ow nership of such entity You or Your shall mean an individual or Legal Entity exercising permissions granted by this License Source form shall mean the preferred form for making modifications including but not limited to softw are source code documentation source and configuration files Object form shall mean any form resulting from mechanical transformation or translation of a Source form including but not limited to compiled object code generated documentation and conversions to other media types Work shall mean the w ork of authorship whether in Source or Object form made available under the License as indicated by a copyright notice that is included in or attached to the work an example is provided in the Appendix below Derivative Works shall mean any work whether in Source or Object form that is based on or deri
149. ved from the Work and for which the editorial revisions annotations elaborations or other modifications represent as a whole an original w ork of authorship For the purposes of this License Derivative Works shall not include w orks that remain separable from or merely link or bind by name to the interfaces of the Work and Derivative Works thereof Contribution shall mean any w ork of authorship including the original version of the Work and any modifications or additions to that Work or Derivative Works thereof that is intentionally submitted to Licensor for inclusion in the Work by the copyright ow ner or by an individual or Legal Entity authorized to submit on behalf of the copyright ow ner For the purposes of this definition submitted means any form of electronic verbal or written communication sent to the Licensor or its representatives including but not limited to communication on electronic mailing lists source code control systems and issue tracking systems that are managed by or on behalf of the Licensor for the purpose of discussing and improving the Work but excluding communication that is conspicuously marked or otherw ise designated in writing by the copyright ow ner as Not a Contribution Contributor shall mean Licensor and any individual or Legal Entity on behalf of whoma Contribution has been received by Licensor and subsequently incorporated within the Work 2 Grant of Copyright License Subject to the terms
150. with the selected steps e Vectra Tissue Finder Algorithm Creates an algorithm to use in the Vectra software to identify areas of tissue See Creating Algorithms for Vectral 9 for details e Vectra HPF Finder Algorithm Creates an algorithm to use in the Vectra software to identify areas of tissue where images should be taken using the high power objective See Creating Algorithms for Vectral 94 for details Prepare Images Required for all projects Specifies the Image Format and Sample format changes the image resolution loads a spectral library unmixes the spectral components and removes haze from the images See Preparing Images 39 for details Segment Tissue Segment the image into specific tissue categories e Manual Tissue Segmentation Tissue categories are manually drawn on each image See Manual Tissue Segmentation 441 for details e Trainable Tissue Segmentation The tissue segmenter is trained to identify different tissue types in the images Processing the images automatically assigns the appropriate tissue 22 inForm User Help categories See Trainable Tissue Segmentation 47 for details Find Features Find features in the images using thresholds or segmentation algorithms e Cell Segmentation Identifies individual cells or subcellular objects in the image using either thresholded or algorithmic segmentation See Segmenting Cells 5s for details e Object Segmentation Identifies objects on an i
151. would be misclassified The software continues to try to improve the accuracy Click the Stop button to stop training when an acceptable level of accuracy is reached or the accuracy has stabilized In general segmenters should be at least 80 accurate If the accuracy reaches 100 a Done button displays to close the Algorithm Training dialog Recent Trainings The Recent Trainings box records a history of segmenters trained using the current segmentation settings For example the first time you train is named Attempt 1 the second training is named Attempt 2 and so on To re classify the images using a previous segmenter select the desired training attempt i e segmenter from this drop down box If you change any settings the history of segmenters is cleared Segmentation Resolution The Segmentation Resolution determines the resolution of the tissue category segmentation There is a trade off between resolution and segmentation speed with an approximately 4x increase in segmentation speed for every step in resolution reduction Select Coarse if you want to segment the images more quickly Although the edges of the segmented regions appear pixelated the segmenter still accurately differentiates structures well e g tumor from stroma Select Fine if the segmented regions need to have smoother edges such as images for use in publications Segmenting with Fine resolution takes longer than with Coarse resolution Also if
152. xtremely difficult to estimate and ascertain thus making any remedy at law or in damages inadequate Therefore you agree that CRI shall be entitled without the necessity of posting of any bond or security to the issuance of injunctive relief by any court of competent jurisdiction enjoining any breach or threatened breach of such covenants and for any other relief such court deems appropriate This right shall be in addition to any other remedy available to CRI at law or in equity 4 Termination The license granted in Section 1 above is effective until terminated This Agreement is conditioned upon your continued compliance w ith the terms and conditions hereof and will terminate automatically without notice from CRI if you fail to comply with any term or condition of the Agreement Furthermore CRI may terminate this Agreement at any time upon thirty 30 days notice Upon termination of this Agreement you shall immediately destroy all copies of the Softw are including all accompanying documentation and any other confidential and proprietary information you have received during or in connection w ith this Agreement 5 Limited Warranty CRI warrants that the media on which the Softw are is provided will be free from defects in materials and faulty workmanship under normal use for a period of ninety 90 days from the date of delivery Your exclusive remedy under this Section 6 shall be at CRI s option a refund of the price paid for the Software or
153. y if the Image Preparation Settings are changed To open this window click the View button under the Spectral Library list box in the Prepare Images step after selecting a spectral library r 1 Spectral Library camy Units Spectral Lib csl Unmix Units z Display 0 6 Scale to Max equal scales Mi 04 Save Chat 02 Stains EE Cy5 0 Gl Cy3 Mg DAP 06 ME autofluorescence 2 04 2 02 c 0 p P SS 0 6 0 4 0 2 TE 420 460 500 540 580 620 660 700 740 Wavelength nm d Figure 5 Spectral Library with multiple filters Spectral Library Name For multispectral im3 images and RGB images displays the name of the current spectral library above the charts The Spectral Library name is not displayed for TRIO im4 images because the spectral library is embedded in the image cube and does not have a separate file name Charts The spectral library display can contain multiple charts one chart for each imaging band Each chart displays the spectral segments within an imaging band Brightfield images only have one imaging band so only one chart displays After loading a library select the check boxes for the spectra you want to use for unmixing in the Prepare Images step The Spectral Library window automatically updates to show only the selected spectra For TRIO im4 images all spectra are selected for unmixing and the selections cannot be cha
154. you selected a small Pattern Scale the Resolution must be set to Medium or finer so inForm User Help Segmenting the Tissue in the Training Images Once you have trained the segmenter you can segment the tissue in the current image or all of the images in the training set You can segment just the current image first to test the tissue segmenter Then make any desired adjustments to the training parameters and retrain the segmenter To see the Tissue Category mask on all images click the Segment All button When you are ready advance to the next step in the project Fine Tuning the Segmenter Observe the training regions following segmenter training As mentioned above if the software grays out any region this indicates that the region is too small for the selected Pattern Scale and was not used to train the tissue segmenter You can ignore this if segmentation performance is adequate delete the region and redraw a larger one or try selecting a smaller pattern scale The Trim Edges option can be used to erode a tissue segmentation by removing the specified number of pixels from the periphery of the selected tissue category making it smaller This is useful if the tissue segmenter has a tendency to include pixels extending beyond the edges of objects or tissue categories of interest e g tumor or vessels For example if you notice cancer regions extending into non cancerous areas select the Trim Edges option choose the ca

inForm - User Help

Contents

Download Pdf Manuals

Related Search

Related Contents