Home

here - MITK

1. 20 3 1 Internal presets 5 6 24 se 20 3 2 Saving and loading custom presets 20 4 Interactively create transferfunctions 20 4 1 Threshold lt o lt gt 20392 Bell eg beet Ha de ee amp Eos 20 5 Customize transferfunctions in detail 20 5 1 Choosing grayvalue interval to edit 20 5 2 tay vale lt gt OPA oo kek ls eR a RG Ae A E E E 20 5 3 Grayvalue gt Color 20 5 4 Grayvalue and Gradient gt Opacity The Image Navigator 93 94 94 95 96 97 98 98 98 99 100 100 101 101 103 104 105 105 106 106 106 106 107 107 108 109 109 109 110 111 111 113 Generated on Thu Nov 3 17 31 07 2011 for DI_App_docu by Doxygen Chapter 1 Using The Diffusion Imaging Application 1 1 What is the Diffusion Imaging Application The Diffusion Imaging Application contains selected views for the analysis of images of the human brain These encompass the views developed by the Neuroimaging Group of the Division Medical and Bio logical Informatics as well as basic image processing views such as segmentation and volumevisualization For a basic guide to MITK see General MITK Manual Using The Diffusion Imaging Application Generated on Thu Nov 3 17 31 07 2011 for DI_App_docu by Doxygen Chapter 2 The Segmentation Module 4 The Segmentation Module sale Figure 2 1 Icon of the Module Some of the features described below are not available in the
2. 18 2111 Tnrod ch i c c se sw ioa e e ee PRA a be etek be ee 18 2 11 2 What might be part of an extension s ks we eR t epep ioter 18 ii CONTENTS S21 Topless ranea a e e Ee a BRS 19 21122 GUI classes for toal 2 4 2 4 e aed 4 A cs e 19 21123 Pom fles ag ac aR do ee we OE a NR 19 2 11 3 Writing a CMake file for a tool extension o e 19 2114A Compiling the extension ua a a a 20 2113 Configuring ITK amtolodd 2 65 cw hee Sea ee EE ee ee ee eS 21 3 The Brain Network Analysis Module 23 Syl SUMDAN NAAA 24 Se Del dr AR A A ESOS A EEE ES Hae eS 24 E III boo ia ds Fe BN eee OO le ee oo eo ee ee ci 24 OA URGE AI 26 39 Troubleshooting 2 00 ba ke ee Re ee eee eee ee ee ee 26 4 The DataManager 27 Al AIN 28 2 Mace ate coc a HAs RRR PEARS SSE ASS SORE EAA Ao 29 do DUMAS AE eR Eee ss E G 30 4 4 Working with the Datamanager 00 0000 000 30 44 1 List of Data Elements 2 020022 eee eee eee 30 442 Visibility of Data Blemenmts gt gt c c o seca mecra ka RAS eS 30 4 4 3 Representation of Data Elements 2 0 31 AA AAN 32 AS Propemy list lt e ce bea eh awe a ERASE ha we be 33 5 General MITK Manual 35 Il AWM MIRK ooo clo o ee ae a Oe Be oe ee hoe de SO AR 36 32 Whe User IMENE nk ela we A ae ba a eee es 36 rA Four Window View nok ke eek a eS ee eee be ER eS 37 Sk EIN 2c 6 A SS A eee Gwe EEE eS 37 Dost WWANIORIOM cep KS AS AA ER RAE REESE RSE
3. work that is extensible with new tools description at Technical design of QmitkSegmentation The usual way to create new tools since it is mostly used inside DKEZ is to just add new files to the MITK source code tree However this requires to be familiar with the MITK build system and turnaround time during development might be long recompiling parts of MITK again and again For external users who just want to use MITK as a library and application there is a way to create new segmentation tools in an MITK external project which will compile the new tools into a shared object DLL Such shared objects can be loaded via the ITK object factory and its autoload feature on application startup This document describes how to build such external extensions Example files can be found in the MITK source code in the directory MITK_SOURCE_ DIR QApplications ToolExtensionsExample 2 11 2 What might be part of an extension The extension concept assumes that you want to create one or several new interactive segmentation tools for Segmentation or another MITK functionality that uses the tools infrastructure In the result you will create a shared object DLL which contains several tools and their GUI counterparts plus optional code that your extension requires The following sections shortly describe each of these parts Generated on Thu Nov 3 17 31 07 2011 for DI_App_docu by Doxygen 2 11 How to extend the Segmentation bundle
4. 108 90 gt Time 0 00 ms Pixelvalue 0 00 1 38 GB 17 86 To create a ROI the mean FA skeleton typically called mean_FA_skeleton nii gz that is created by FSL should be loaded in to the datamanager and selected By using the Pointlistwidget points should be set on the skeleton make sure to select points with relatively high FA values Points are set by first selecting the button with the and than shift leftclick in the image When the correct image is selected in the datamanager the Create ROI button is enabled Clicking this will create a region of interest that passes through the previously selected points The roi appears in the datamanager Before doing so the name of the roi and the information on the structure on which the ROI lies can be set This will be saved as extra information in the roi image Before the ROI is calculated a pop up window will ask the user to provide a threshold value This should be the same threshold that was previously used in FSL to create a binary mask of the FA skeleton When this is not done correctly the region of interest will possible contain zero valued voxels Generated on Thu Nov 3 17 31 07 2011 for DI_App_docu by Doxygen 98 The Tbss Analysis Module 18 6 Profile plots File Edit Window Help Fm oven FE save Project BX close Project M undo Ci redo EE image Navigator sen Sai Data Manager Display ES e es roiname a x FSL import E groups Subject data G
5. A IA ULZ Detalla lt lt es a Se eee ee a E A ate bee le dated BAe ae eee ER UU DISSES se bo ae eG ee a ee ork ee RO eS 10 5 Troubleshooting cocidos ea ee A The Measurement Module ULA ESRIMES La a la e we ebb ke ed ILLI Daw Line seia raain a BE He Ea ILIZ Drow Path c es k Ar a FS Ge a MAS Dew Anel poso nre Ea ee bbb E 11 1 4 Draw Four Point Angle 53 54 55 56 57 59 61 62 62 62 63 64 64 64 66 66 Generated on Thu Nov 3 17 31 07 2011 for DI_App_docu by Doxygen iv CONTENTS ILIS Dior Ce ie ee ee ee se CA Be Re oe Ae Oe EEA Me 69 ILLO Diiw Recimil coo oia a eee el BAS E ee ke 69 1617 Diaw PONE o be ke be ee Bee be ee eo Se bak ee Ge 69 le Rates obs Be a Gna he e le Be ta a andes de eg 70 11 2 1 Work with measurement figures 2 2 2 2 000002 eee eee 70 11 2 2 Save the image with measurement information 71 11 2 3 Remove measurement figures or image e 71 12 The Movie Maker Module 73 A E ooe ee OE Re OPES BO Se SERS eS Shes 74 Li PEE oo dad a Boe AAA ee ee Sie Y 74 24 Usage eo Pe eh eee ee PE EER ee bed eh ee a a 75 123 1 Window o A EA 76 1232 Recowdme DIODOS caian a ee eee Od AR S 76 123 Paving OPONE e A eR o gta GRE 76 13 ODF Details View 77 ESE ES o o A A ee ee ie ee ER 78 14 The Image Statistics Module 79 AL SUMMA coc a a a A A a ee aS 80 142 Deals 3 Oh a ee ee Sw Phe eR ee ee dd a Oe RS 80 AAA 80 LA e se pe Lan
6. FSL creates output images that typically have names like all_FA_skeletonized nii gz that are 4 dimensional images that contain registered images of all subjects By loading this 4D image into the datamanager and listing the groups with the correct number of subjects in the order of occurrence in the 4D image in the TBSS View using the Add button and clicking the import subject data a TBSS file is created that contains all the information needed for tract analysis The diffusion measure of the image can be set as well Generated on Thu Nov 3 17 31 07 2011 for DI_App_docu by Doxygen 18 5 Regions of interest 97 18 5 Regions of interest File Edit Window Help feb open Fi save Project JX close Project E undo A Redo image Navigator SONS aL Data Manager Display E S TBSS roiname Y FSL import Pointset Es mean_FA_skeleton Subject data Group Size Healthy 23 Diseased 24 W mie channel Remove Diffusion measure Fractional Anisotropy Tools ROIs Measuring mean_FA_skeleton 3D Points onRoi 9 1 000 24 236 11 887 1 1 000 21 064 15 002 al e a a ay E Create ROI 9115084 9115085 T IS 9114985 load a tbss meta image into the datamanager 9114986 9114886 9114887 Show Welcome Screen 91147 87 Transversal 90 Sagittal 91 Name roiname Coronal S 109 Structure info Structure Time 0 Position lt 1 00 18 00 18 00 gt mm Index lt 91
7. Input Parameters e References Generated on Thu Nov 3 17 31 07 2011 for DI_App_docu by Doxygen 89 Data Manager seg2 segl Diffusion_Raw E Diffusion_Raw3500 Stochastic Tracking A Selected Image s Status FileName 1 OK Diffusion_Raw 2 OK seg2 3 OK segl Tract_len TotalTracts Lkhd cache 50 1 500 O Execute FibreTracking Figure 17 1 Stochastic Tracking View Generated on Thu Nov 3 17 31 07 2011 for DI_App_docu by Doxygen 90 Stochastic Tracking View y nh Pv Analysis Quantification Measurement Segmentation image testl sy Contouring Editing tools Remember Contour Positions Contour interpolation for all slices Figure 17 2 Segmentation View to define ROIs 17 1 Input Data Mandatory Input Generated on Thu Nov 3 17 31 07 2011 for DI_App_docu by Doxygen 17 2 Input Parameters 91 e For a successful execution of the stochastic tractography filter a DWI input and a binary image defining the desired ROI are required The ROI serves as the origin from where on the fibers are beeing tracked Both DWI and ROI data can be imported via drag n drop or via the opening dialog box provided by MITK Alternatively the segmentation view offers tools for generating ROI data e One DWI Image image selected in the datamanager e One or more ROIs selected in the datamanager 17 2 Input Parameters e Parameters such as max tract le
8. Load Indicator in the lower right hand corner of the screen gives information about the memory currently required by the MITK application Keep in mind that image processing is a highly memory intensive task and monitor the indicator to avoid your system freezing while constantly swapping to the hard drive 5 7 Perspectives The different tasks that arise in medical imaging need very different approaches To acknowledge this circumstance MITK supplies a framework that can be build uppon by very different solutions to those tasks These solutions are called perspectives each of them works independently of others although they might be used in sequence to achieve the solution of more difficult problems It is possible to switch between the perspectives using the Window gt Open Perspective dialog using the Welcome Screen or by using the bar on top of the editor area See Menu for more information about switching perspectives Generated on Thu Nov 3 17 31 07 2011 for DI_App_docu by Doxygen Chapter 6 MITK Diffusion Imaging MITK DI 42 MITK Diffusion Imaging MITK DI This module provides means to diffusion weighted image reconstruction visualization and quantification Diffusion tensors as well as different q ball reconstruction schemes are supported Q ball imaging aims at recovering more detailed information about the orientations of fibers from diffusion MRI measurements and in particular to resolve the orientations of cr
9. Stefano J M Brady and P M Matthews Advances in functional and structural MR image analysis and implementation as FSL NeuroImage 23 S1 208 219 2004 Generated on Thu Nov 3 17 31 07 2011 for DI_App_docu by Doxygen Chapter 19 Intra voxel incoherent motion estimation IVIM 100 Intra voxel incoherent motion estimation IVIM The required input for the Intra voxel incoherent motion estimation IVIM is a diffusion weighted image dwi or hdwi that was acquired with several different b values Window Help 3 open Fel saveProject DX Close Project undo Qt Redo Image Navigator ONO LLE Data Manager i Welcome 3 y B DIETMAR El E kidney DES W MIP channel3 gt VIM Intra Voxel Incoherent Motion Estimation neglect b lt 170 neglect Si lt Do Outputimages Y f D D gt Generate Output Images Averaging 310 voxels display voxel wise results 1 0 8 rrt T T T T T 1 O 100 200 300 400 500 600 700 800 F 0 16897 D 0 00163796 D 0 00845 ignored measurement points additional points second Fit Position lt 79 05 25 81 5 09 gt mm Index lt 30 52 7 gt Time 0 00 ms Pixelvalue 1 00 2 27 GB 29 19 Figure 19 1 The IVIM View Once an input image is selected in the datamanager the IVIM view allows for interactive exploration of the dataset click around in the image and watch the estimated parameters in the figure of the view as well as generation o
10. The width parameter correspondens to the width of the bell 20 5 Customize transferfunctions in detail 20 5 1 Choosing grayvalue interval to edit b Reset To navigate across the grayvalue range or to zoom in some ranges use the range slider Generated on Thu Nov 3 17 31 07 2011 for DI_App_docu by Doxygen 110 The Volume Visualization Module All three function editors have in common following e By left clicking a new point is added e By right clicking a point is deleted e By left clicking and holding an exisiting point can be dragged e By pressing arrow keys the currently selected point is moved By pressing the DELETE key the currently selected point is deleted e Between points the transferfunctions are linear interpolated There are three transferfunctions to customize 20 5 2 Grayvalue gt Opacity Grayvalue gt Opacity Figure 20 2 grayvalues will be mapped to opacity An opacity of 0 means total transparent an opacity of means total opaque Generated on Thu Nov 3 17 31 07 2011 for DI_App_docu by Doxygen 20 5 Customize transferfunctions in detail 111 20 5 3 Grayvalue gt Color Grayvalue gt Color 1024 Figure 20 3 grayvalues will be mapped to color The color transferfunction editor also allows by double clicking a point to change its color 20 5 4 Grayvalue and Gradient gt Opacity Grayvalue Gradient gt Ope Here the influence of
11. allowing you to start editing the new segmentation right away 2 4 2 Selecting Segmentations for Editing As you might want to have segmentations of multiple structures in a single patient image the application needs to know which of them to use for editing You select a segmenation by clicking it in the tree view of Data Manager Note that segmentations are usually displayed as sub items of their patient image In the rare case where you need to edit a segmentation that is not displayed as a a sub item you can click both the original image AND the segmentation while holding down CTRL on the keyboard When a selection is made the Segmentation View will hide all but the selected segmentation and the corresponding original image When there are multiple segmentations the unselected ones will remain in the Data Manager you can make them visible at any time by selecting them If you want to see all segmenations at the same time just clear the selection by clicking outside all the tree items in the Data Manager 2 4 3 Selecting Editing Tools If you are familiar with MITKApp you know that clicking and moving the mouse in any of the 2D render windows will move around the crosshair that defines what part of the image is displayed This behavior is disabled while any of the manual segmentation tools are active otherwise you might have a hard time concentrating on the contour you are drawing To start using one of the editing tools click its
12. example would look like this Generated on Thu Nov 3 17 31 07 2011 for DI_App_docu by Doxygen 20 The Segmentation Module PROJECT FIND_PAC FIND_PAC FIND_PAC ExternalTool AGE ITK AGE MITK AGE Qt3 ADD_DEFINITIONS QT_DEFINITIONS SET SET SET SET MITK_GENERAT TOO L_OT3GUI_FILES QmitkExternalToolGUI cpp TOO L_FILES mitkExternalTool cpp TOO L_ADDITIONAL_CPPS externalalgorithm cpp externalalgorithmsolver cpp TOOL_ADDITIONAL_MOC_H TE_TOOLS_LIBRARY mitkExternalTools Basically you only have to change the definitions of TOOL_FILES and optionally TOOL_OT3GUI_ L_CPPS and TOOL_ADDITIONAL_MOC_H For all cpp files in TOOL_ FILES TOOL_ADDITIONA FILES and TOOL_OT3GUI described listed in TOOL_ADDITIONAI _FILES there will be factories created assuming the naming conventions in the sections above Files listed in TOOL_ADDITIONAL_CPPS will just be compiled Files L_MOC_H will be run through Qts meta object compiler moc this is nec cessary for all objects that have the Q OBJECT macro in their declaration moc will create new files that will also be compiled into the library 2 11 4 Compiling the extension For compiling a tool extension you will need a compiled version of MITK We will assume MITK was compiled into home user mitk debug You need to build MITK with BUILD_SHARED_CORE turne
13. made up of volume of physical measurements volume elements are called voxels In CT images for example the gray value of each voxel corresponds to the mass absorbtion coefficient for X rays in this voxel which is similar in many parts of the human body The gray value does not contain any further information so the computer does not know whether a given voxel is part of the body or the background nor can it tell a brain from a liver However the distinction between a foreground and a background structure is required when e you want to know the volume of a given organ the computer needs to know which parts of the image belong to this organ e you want to create 3D polygon visualizations the computer needs to know the surfaces of structures that should be drawn e as a necessary pre processing step for therapy planning therapy support and therapy monitoring Creating this distinction between foreground and background is called segmentation The Segmentation perspective of MITKApp uses a voxel based approach to segmentation i e each voxel of an image must be completely assigned to either foreground or background This is in contrast to some other applications which might use an approach based on contours where the border of a structure might cut a voxel into two parts Generated on Thu Nov 3 17 31 07 2011 for DI_App_docu by Doxygen 6 The Segmentation Module The remainder of this document will summarize the features of the Se
14. supports undo and redo operations as well as the image navigator which gives you sliders to navigate through the data quickly The Preferences dialog allows you to adjust and save your custom settings Generated on Thu Nov 3 17 31 07 2011 for DI_App_docu by Doxygen 40 General MITK Manual 5 4 3 Help This dialog contains this help the welcome screen and information about MITK Diffusion 5 5 Levelwindow Once an image is loaded the levelwindow appears to the right hand side of the four window view With this tool you can adjust the range of grey values displayed and the gradient between them Moving the lower boundary up results in any pixels having a value lower than that boundary to be displayed as black Lowering the upper boundary causes all pixels having a value higher than it to be displayed as white The pixels with a value between the lower and upper boundary are displayed in different shades of grey This way a smaller levelwindow results in higher contrasts while cutting of the information outside its range whereas a larger levelwindow displays more information at the cost of contrast and detail You can pick the levelwindow with the mouse to move it up and down while moving the mouse cursor to the left or right to change its size Picking one of the boundaries with a left click allows you to change the size symmetrically Holding CTRL and clicking a boundary adjusts only that value 5 6 System Load Indicator The System
15. the files In case your dicom images are readable by MITK DI select one or more input dicom folders and click import Each input folder must only contain DICOM images that can be combined into one vector valued 3D output volume Different patients must be loaded from different input folders The folders must not contain other acquisitions e g T1 T2 localizer In case many imports are performed at once it is recommended to set the the optional output folder argu ment This prevents the images from being kept in memory Generated on Thu Nov 3 17 31 07 2011 for DI_App_docu by Doxygen 48 MITK Diffusion Imaging MITK DI Each input folder must only contain DICOM images that can be combined into one vector valued 3D output volume Different patients must be loaded from different input folders The folders must not contain other Recursive C Avg dupl grad blur 0 001000 Set optional out folder E Import DICOM as dwi Figure 6 8 Dicom import The option Average duplicate gradients accumulates the information that was acquired with multiple repetitions for one gradient Vectors do not have to be precisely equal in order to be merged if a blur radius gt 0 is configured 6 6 Quantification The quantification view allows the derivation of different scalar anisotropy measures for the reconstructed tensors Fractional Anisotropy Relative Anisotropy Axial Diffusivity Radial Diffusivity o
16. visualization of the ODF as well as the ODF values and according statistical information are displayed 6 Preprocessing Tensors XK Q Balls ODF Details Sum 1 Mean 0 00396826 Min 0 00294184 Max 0 0060185 GFA 0 204762 Pos 7 12 0 0 006 0 005 0 004 0 003 0 002 0 001 0 FAA AA AA AS AO AAA 0 50 100 150 200 ODF Values 1 2 3 4 1 0 00504804 0 00524651 0 00497723 0 00424586 0 0035 Figure 13 1 The ODF Details View 13 1 Issues At the moment this view can opnly process Q Ball images but not tensor images Also the normalization properties etc of the image as well as the correct rotation of the ODF are currently not incorporated into the views visualization Generated on Thu Nov 3 17 31 07 2011 for DI_App_docu by Doxygen Chapter 14 The Image Statistics Module 80 The Image Statistics Module Figure 14 1 Icon of the Module 14 1 Summary This module provides an easy interface to quickly compute some features of a whole image or a region of interest This document will tell you how to use this module but it is assumed that you already know how to use MITK in general Please see Details for more detailed information on usage and supported filters If you encounter problems using the module please have a look at the Troubleshooting page 14 2 Details Manual sections e Overview e Usage e Troubleshooting 14 3 Overview This module provides an easy interface to quickly co
17. 16 0 VTE 1 4 2 14 5 9 MITK nda Datamanager Figure 4 2 How MITK looks when starting 4 2 Loading Data There are three ways of loading data into the Datamanager as so called Data Elements The user can just drag and drop data into the Datamanager or directly into one of the four parts of the Standard View He can as well use the Open Button in the right upper corner Or he can use the standard File gt Open Dialog on the top A lot of file formats can be loaded into MITK for example e 2D images 3D volumes with or without several timesteps dcm ima pic e Surfaces x stl vtk e Pointsets mps The user can also load a series of 2D images e g image001 bmp image002 bmp to a MITK 3D volume To do this just drag and drop one of those 2D data files into the Datamanager by holding the ALT key After loading one or more data into the Datamanager they appear as Data Elements in a sorted list inside the Datamanager Data Elements can also be sorted hierarchically as a parent child relation For example Generated on Thu Nov 3 17 31 07 2011 for DI_App_docu by Doxygen 30 The DataManager after using the Segmentation Module on Data Elementl the result is created as Data Element2 which is a child of Data Elementl see Screenshot1 The order can be changed by drag and drop E Segmentation Mitk3M3 ITK 3 16 0 VTK 5 4 2 Ot 4 5 3 MITK 20536 File Edit Window Help 4 E Open Save P
18. Be ER a 37 Soave COMZE cocoa eRe ee bea ed e 38 Sd MI hs o Sad ate By ee AS e BS BR ES Be ee ae we Ee SS 39 Bel PG ce ou a OR en es Bs E E E ee ee Pe A RD 39 SA2 AP 39 SA BI o Sco oh ee E Boe ea be ee ot da la 40 SS Levelwmdow lt o corans sesa ad aa ee Eee ER EE Ee Ee LG 40 56 System Load Indiestol lt n 6c ek ee eS Bae ee Re oe eS SSE RR OD 40 Generated on Thu Nov 3 17 31 07 2011 for DI_App_docu by Doxygen CONTENTS iii 6 7 8 10 11 XI IETSPOGLIVES cs kone ce tok Ree ON See ee ee MITK Diffusion Imaging MITK DI 6 1 Keown snes ok oe ee ea GS ee Sa ee ee G2 Prepocessine eo ha a a A SE a EY 6 3 TensorReconstruction lt gt e spree esoo ek neoe 6 4 Q Ball Reconstruction o a 63 Dicom Mpo s 2 0 6 64 4 swane a ie ee a dos 6 6 Qiantificdion 6 65 sea ke es ee eR RE a ee ew a ee 6 7 ODF Visualization Setting 2 6 ee ee eses DS IBEISIEDOSS ear g SS eee ee ew 6 9 Technical Information for Developers Diffusion Bundles in Development 74 Experimental Bundles s o ne pa ka a Fiber Processing View 8 1 Fiber Bundle Manipulation 8 2 Generation of additional information from fiber bundles Gibbs Tracking View Ol TUD cis codi oi Sk PORE e a NS ES 9 2 Q Ball Reconstruction gt ss e ssec desi crete dma 9 3 Suryeilanc of the tracking process YA RE STSCES os oa A A The Image Statistics Module ai
19. Circle Draws a circle by setting two points whereas the first set point is the center and the second the radius of the circle The measured values are the radius and the included area 11 1 6 Draw Rectangle Draws a rectangle by setting two points at the opposing edges of the rectangle starting with the upper left edge The measured values are the circumference and the included area 11 1 7 Draw Polygon Draws a polygon by setting three or more points The measured values are the circumference and the included area Add the final point by double left click Below the buttonline the statistics window is situated it displays the results of the actual measurements from the selected measurement figures The content of the statistics window can be copied to the clipboard with the correspondig button for further use in a table calculation programm e g Open Office Calc etc Generated on Thu Nov 3 17 31 07 2011 for DI_App_docu by Doxygen 70 The Measurement Module W Measurements Mitk3M3 ITK 3 16 0 TK 5 4 2 Qt 4 5 3 MITK 20586 lol x File Edit Window Help hr gt E cs 4 H Open m Save Project EA Close Project K Undo Redo gt Image Navigator Data Manager E DICOM MRI Image Y Ml Rectanglet a Angle1 Y Circlet M Linet jae Image DICOM MRI Image SAK MS y LA Rectangle1 Circumference 148 78 mm Area 1370 59 mm Angle1 Angle 79 61 Circle1 R
20. Ge eee gee eee Shee ee es Pha we eS on 82 14 5 Troubleshoo ng s op oda EAE RSS See RE RS PERK a AO 82 15 Partial Volume Analysis 83 ESE ERP o a A A Be bee eB Se ee AR os 84 Z BR PO e egos Dawe Ge ee eb dum ee eps Shae wee ees Pha we eS ae 84 3 5 EXPO y co 28s e444 4240 A a See DEE TES TERRES EE ES 84 16 The Screenshot Maker 85 HOA UBRES oy A A Re Oe ee Re Se ee oe ee AE oS 86 17 Stochastic Tracking View 87 PVA Wiper Da sn cria be ea ea ee ea ee ee ee ee aS 90 UEZ Input Parsinetere ogo se ce po ae ES RRS a ees PAE ee eS 91 ELA Revie ia e ee oe ee ee tee eee Be 91 Generated on Thu Nov 3 17 31 07 2011 for DI_App_docu by Doxygen CONTENTS 18 19 20 21 The Tbss Analysis Module 181 SUMMA ccoo ee ee ee 18 2 Demis cocinar ee a ao US VEIVIEN ass onan a 18 4 FSL IMPO 20 18 5 Regions of interest 13 6 Profe plete o corso 18 7 Troubleshooting o s 24 542 65 5 6 e444 18 8 References coooa crek tarkkaa Intra voxel incoherent motion estimation IVIM 19 1 Region of interest analysis 19 2 Region of interest analysis 19 3 Region of interest analysis 19 4 Region of interest analysis The Volume Visualization Module 20 1 OVGIVIEW o o Se ee es HE ee ey 20 2 Enable Volume Rendering 20 2 1 Loading an image into the application 20 2 2 Enable Volumerendering 20 2 3 The LOD amp GPU checkboxes 20 3 Applying premade presets
21. Is communicate with their associated tools via direct method calls they know their tools See mitk BinaryThresholdTool for examples Again a naming convention if the convenience macros for tool extension shared ob jects are used you have to put a tool GUI called OmitkExternalToolGUI into a files named QmitkExternalToolGUI cpp and QmitkExternalToolGUI h The convenience macro will create a factory called QmitkExternalToolGUIFactory into a file named QmitkExternalToolGUIFactory cpp 2 11 2 3 Additional files If you are writing tools MITK externally these tools might depend on additional files e g segmentation algorithms These can also be compiled into a tool extension shared object 2 11 3 Writing a CMake file for a tool extension Summing up the last section an example tool extension could comprise the following files mitkExternalTool h mitkExternalTool xpm gt implementing mitk ExternalTool header icon implementation mitkExternalTool cpp QmitkExternalToolGUI h 7 implementing a GUI for mitk ExternalTool QmitkExternalToolGUI cpp externalalgorithm h externalalgorithm cpp externalalgorithmsolver h gt a couple of files not related to MITK tools externalalgorithmsolver cpp This should all be compiled into one shared object Just like ITK VTK and MITK we will use CMake for this purpose I assume you either know or are willing to learn about www cmake org A CMake file for the above
22. MITK Diffusion Documentation 2011 Generated by Doxygen 1 6 2 Thu Nov 3 17 31 07 2011 Contents 1 Using The Diffusion Imaging Application 1 1 1 What is the Diffusion Imaging Application 0048 1 2 The Segmentation Module 3 2 ERVIN eb coe as ee a ee ee ee ae SERS 4 22 TechmicalIss es lt eccora e eee BE EE ye ae eee ee ee ek 6 2o a AA AA 6 ZA Mantal Contours coo opio a a A aed ee ees 7 24 1 Creating New Sesmentations o 2 456s cee ER ee eR 8 2 4 2 Selecting Segmentations for Editing o e seso eme comise criias 8 2a Selectma Eding Toole ces cio bwin se Se eS 8 244 Using Editing Tools sc eresten ee eee 8 2A 0 e 34 5s ee ek da bee eee bak ee See baw be ees a 11 2 0 Organ Segmentation np a kk SEES See eae PhS Re eo 12 ool Jer Ge EAN o crei erani BAe ee eels ele wana de lotes 12 2 5 2 Heart Lung and Hippocampus on MRI 12 Zo OTOS ong A A Se Be ees Se 13 28 Lesion SQCGMEMAUON 2 4 0 25 55 0a dae RAGE RR EGE ES ORAS S 13 2 7 Things you can do with segmentations 2 2 2 ee 13 28 UMSS Masking pie a RR YA ee es Ee eee Ee A 14 29 Technical Information for Developers 005 5 6 ee ee ee 15 2 10 Technical design of QmitkSegmentation o e e 16 210 1 ITA oo cisco a a A a a Y 16 2 10 2 IVETE aask lt a ic E a ees A E ARANA 16 2103 Classes involved co somos cross a ESE Re 16 2 11 How to extend the Segmentation bundle with external tools
23. The Image Statistics Module Figure 10 1 Icon of the Module 10 1 Summary This module provides an easy interface to quickly compute some features of a whole image or a region of interest This document will tell you how to use this module but it is assumed that you already know how to use MITK in general Please see Details for more detailed information on usage and supported filters If you encounter problems using the module please have a look at the Troubleshooting page 10 2 Details Manual sections e Overview e Usage e Troubleshooting 10 3 Overview This module provides an easy interface to quickly compute some features of a whole image or a region of interest Generated on Thu Nov 3 17 31 07 2011 for DI_App_docu by Doxygen 10 3 Overview 65 1835 42 1020 052 604 516 345 180 Figure 10 2 The interface 66 The Image Statistics Module 10 4 Usage After selection of an image or a binary mask of an image in the datamanader the Image Statistics module shows some Statistical information If a mask is selected the name of the mask and the name of the image to which the mask is applied are shown at the top Below it is the statistics window which displays the calculated statistical features such as mean standard deviation and the histogram At the bottom of the module are two buttons They copy their respective data in csv format to the clipboard 10 5 Troubleshooting No kno
24. a range of 50 or more slices The Segmentation view offers a helpful feature for these cases Generated on Thu Nov 3 17 31 07 2011 for DI_App_docu by Doxygen 12 The Segmentation Module Interpolation creates suggestions for a segmentation whenever you have a slice that e has got neighboring slices with segmentations these do not need to be direct neighbors but could also be a couple of slices away AND e is completely clear of a manual segmentation i e there will be no suggestion if there is even only a single pixel of segmentation in the current slice Interpolated suggestions are displayed in a different way than manual segmentations are until you accept them as part of the segmentation To accept single slices click the Accept button below the toolbox If you have segmented a whole organ in every x slice you may also review the interpolations and then accept all of them at once by clicking all slices 2 5 Organ Segmentation The manual contouring described above is a fallback option that will work for any kind of images and structures of interest However manual contouring is very time consuming and tedious This is why a ma jor part of image analysis research is working towards automatic segmentation methods The Segmentation View comprises a number of easy to use tools for segmentation of CT images Liver and MR image left ventricle and wall left and right lung 2 5 1 Liver on CT Images On CT image volu
25. adius 22 78 mm Diameter 45 55 mm Area 1629 67 mm Linel Length 45 25 mm Copy to Clipboard angth a A a ki K E okom Length 45 2 Position lt 16 6064 13 5681 20 0121 gt mm Index lt 289 164 17 gt Time O ms Pixelvalue 115 52 18 MB 1 57 11 2 Usage 11 2 1 Work with measurement figures The measurement module comes with seven measurement figures see picture below that can be ap plied to the images The results of the measurement with each of these figures is shown in the statistics window and in the lower right corner of the view plane Generated on Thu Nov 3 17 31 07 2011 for DI_App_docu by Doxygen 11 2 Usage 71 E Measurements Mitk3M3 ITK 3 16 0 TK 5 4 2 Qt 4 5 3 MITK 20586 s jaj x File Edit Window Help E Open Save Project EA Close Project E Undo Redo z Image Navigator Data Manager Standard view xd M Fe DICOM MRI Image M H Rectangle1 M Angle1 Selected Image DICOM MRI Image Circle1 Radus 22 78 mm Diameter 45 55 mm Ares 1629 67 mm Copy to Clipboard LENTA YE ES Position lt 16 6064 13 5681 20 0121 gt mm Index lt 289 164 17 gt Time O ms Pixelvalue 115 36 67 MB 1 10 When applying more then one measurement figure to the image the actual measurement figure is depicted in red and the displayed values belong to this measurement figure All measurement figures become
26. age 2006 Jun 31 2 53 1 42 Epub 2006 Feb 14 PMID 16478665 Generated on Thu Nov 3 17 31 07 2011 for DI_App_docu by Doxygen Chapter 16 The Screenshot Maker 86 The Screenshot Maker This module provides the functionality to create and save screenshots of the data Available sections e Usage 2D Screenshots 3D Screenshots Upsampling 5 gt Options x z Background Color IN Figure 16 1 The Screenshot Maker User Interface 16 1 Usage The first section offers the option of creating a screenshot of the last activated render window thus the one which was last clicked into Upon clicking the button the Screenshot Maker asks for a filename in which the screenshot is to be stored The multiplanar Screenshot button asks for a folder where screenshots of the three 2D views will be stored with default names The high resolution screenshot section works the same as the simple screenshot section aside from the fact that the user can choose a magnification factor In the option section one can rotate the camera in the 3D view by using the buttons Furthermore one can choose the background colour for the screenshots default is black Generated on Thu Nov 3 17 31 07 2011 for DI_App_docu by Doxygen Chapter 17 Stochastic Tracking View 88 Stochastic Tracking View This view provides the user interface for the Stochastic Fibertracking algorithm proposed by Ngo 1 Available sections e Input Data e
27. avoxel incoherent motion imaging Lemke A Stieltjes B Schad LR Laun FB Magn Reson Imaging 2011 Jul 29 6 766 76 Epub 2011 May 5 PMID 21549538 Differentiation of pancreas carcinoma from healthy pancreatic tissue using multiple b values comparison of apparent diffusion coefficient and intravoxel incoherent motion derived parameters Lemke A Laun FB Klauss M Re TJ Simon D Delorme S Schad LR Stieltjes B Invest Radiol 2009 Dec 44 12 769 75 PMID 19838121 Generated on Thu Nov 3 17 31 07 2011 for DI_App_docu by Doxygen 102 Intra voxel incoherent motion estimation IVIM Generated on Thu Nov 3 17 31 07 2011 for DI_App_docu by Doxygen Chapter 20 The Volume Visualization Module 104 The Volume Visualization Module Figure 20 1 Icon of the Module Available sections e Overview e Enable Volume Rendering e Applying premade presets e Interactively create transferfunctions e Customize transferfunctions in detail 20 1 Overview The Volume Visualization Module is a basic tool for visualizing three dimensional medical images MITK provides generic transfer function presets for medical CT data These functions that map the gray value to color and opacity can be interactively edited Additionally there are controls to quickly generate common used transfer function shapes like the threshold and bell curve to help identify a range of grey values Generated on Thu Nov 3 17 31 07 2011 for DI_App_docu
28. button the the displayed toolbox The selected editing tool will be active and its corresponding button will stay pressed until you click the button again Selecting a different tool also deactivates the previous one If you have to delineate a lot of images you should try using shortcuts to switch tools Just hit the first letter of each tool to activate it A for Add S for Subtract etc 2 4 4 Using Editing Tools All of the editing tools work by the same principle you use the mouse left button to click anywhere in a 2D window any of the orientations transversal sagittal or frontal move the mouse while holding the mouse button and release to finish the editing action All tools work on the original slices of the patient image i e with some rotated tilted MR image volumes you need to perform a reinit option in the Data Manger before you are able to use the editing tools Multi step undo and redo is fully supported by all editing tools Use the application wide undo button in the toolbar to revert erroneous actions Generated on Thu Nov 3 17 31 07 2011 for DI_App_docu by Doxygen 2 4 Manual Contouring 9 6 Subtract Figure 2 4 Add and Subtract Tools Use the left mouse button to draw a closed contour When releasing the mouse button the contour will be added Add tool to or removed from Subtract tool the current segmentation Hold down the CTRL key to invert the operation this will switch tools temporarily to a
29. by Doxygen 20 2 Enable Volume Rendering 105 20 2 Enable Volume Rendering 20 2 1 Loading an image into the application Load an image into the application by e dragging a file into the application window e selecting file load from the menu Volume Visualization imposes following restrictions on images e It has to be a 3D Image Scalar image that means a normal CT or MRT e 3D T are supported for rendering but the histograms are not computed e Also be aware that volume visualization requires a huge amount of memory Very large images may not work unless you use the 64bit version Generated on Thu Nov 3 17 31 07 2011 for DI_App_docu by Doxygen 106 The Volume Visualization Module 20 2 2 Enable Volumerendering Select an image in datamanager and click on the checkbox left of Volumerendering Please be patient while the image is prepared for rendering which can take up to a half minute 20 2 3 The LOD GPU checkboxes Volume Rendering requires a lot of computing resources including processor memory and graphics card Torun volume rendering on smaller platforms enable the LOD checkbox level of detail rendering Level of detail first renders a lower quality preview to increase interactivity If the user stops to interact a normal quality rendering is issued The GPU checkbox tries to use computing resources on the graphics card to accelerate volume rendering It requires a powerful graphics card an
30. currently selected Data Element has Which properties these are depends on the Data Element Examples are opacity shader visibility These properties can be changed by clicking on the appropriate field in the value column Generated on Thu Nov 3 17 31 07 2011 for DI_App_docu by Doxygen 34 The DataManager opacity outline binary r la Figure 4 7 Screenshot5 Property List Generated on Thu Nov 3 17 31 07 2011 for DI_App_docu by Doxygen Chapter 5 General MITK Manual 36 General MITK Manual Welcome to the basic MITK user manual This document tries to give a concise overview of the basic functions of MITK and be an comprehensible guide on using them Available sections e About MITK e The User Interface e Perspectives 5 1 About MITK MITK is an open source framework that was originally developed as a common framework for Ph D students in the Division of Medical and Biological Informatics MBI at the German Cancer Research Center MITK aims at supporting the development of leading edge medical imaging software with a high degree of interaction MITK re uses virtually anything from VTK and ITK Thus it is not at all a competitor to VTK or ITK but an extension which tries to ease the combination of both and to add features not supported by VTK or ITK Research institutes medical professionals and companies alike can use MITK as a basic framework for their research and even commercial thor
31. d on You build the tool extension just like any other CMake based project extension debug e set home user mitk debug the non know where your source code is e g home user mitk tool extension src change into the directory where you want to compile the shared object e g home user mitk tool invoke cmake ccmake home user mitk tool extension src configure press c or the configure button set the ITK_DIR variable to the directory where you compiled ITK MITK_DIR variable to the directory where you compiled MITK configure press c or the configure button generate press g or the generate button This should do it and leave you with a or project file or Makefile that you can compile using make or VisualStudio Generated on Thu Nov 3 17 31 07 2011 for DI_App_docu by Doxygen 2 11 How to extend the Segmentation bundle with external tools 21 2 11 5 Configuring ITK autoload If the compile succeeds you will get a library mitkExternalTools dll or libmitkExternalTools so This library exports a symbol it kLoad which is expected by the ITK object factory On application startup the ITK object factory will search a list of directories from the envi ronment variable ITK_AUTOLOAD_PATH Set this environment variable to your binary directory home user mitk tool extension debug The ITK object factory will load all shared objects that it finds in the specified directories and will test i
32. d OpenGL hardware support for shaders but achieves much higher frame rates than software rendering 20 3 Applying premade presets 20 3 1 Internal presets There are some internal presets given that can be used with normal CT data given in Houndsfield units A large set of medical data has been tested with that presets but it may not suit on some special cases Click on the Preset tab for using internal or custom presets CT Generic is the default transferfunction that is first applied CT Black amp White does not use any colors as it may be distracting on some data CT Cardiac tries to increase detail on CTs from the heart CT Bone emphasizes bones and shows other areas more transparent Generated on Thu Nov 3 17 31 07 2011 for DI_App_docu by Doxygen 20 4 Interactively create transferfunctions 107 CT Bone Gradient is like CT Bone but shows from other organs only the surface by using the gradient MR Generic is the default transferfunction that we use on MRT data which is not normalized like CT data CT Thorax small tries to increase detail CT Thorax large tries to increase detail 20 3 2 Saving and loading custom presets After creating or editing a transferfunction see Customize transferfunctions in detail or Interactively create transferfunctions the custom transferfunction can be stored and later retrieved on the filesystem Click Save respectively Load button to save
33. d be undoable no matter whether a tool or the interpolation mech anism changed something GUI Integration of everything 2 10 3 Classes involved The above blocks correspond to a number of classes Here is an overview of all related classes with their responsibilities and relations Generated on Thu Nov 3 17 31 07 2011 for DI_App_docu by Doxygen 2 10 Technical design of QmitkSegmentation 17 Reference image Toolmanager offers different tools ExtractImageFilter Active tool Working image Tool result OverwriteSliceImageFilter SegmentationInterpolation 1 Management of images mitk ToolManager has a set of reference data original images and a second set of working data segmentations mitk Tool objects know a ToolManager and can ask the manager for the currently relevant images There are two GUI elements that enable the user to modify the set of reference and working images QmitkToolReferenceDataSelectionBox and Qmitk ToolWorkingDataSelectionBox GUI and non GUI classes are coupled by itk Events non GUI to GUI and direct method calls GUI to non GUD 2 Management of drawing tools As a second task ToolManager manages all available tools and makes sure that one at a time is able to receive MITK events The GUI for selecting tools is imple mented in QmitkToolSelectionBox 3 Drawing tools Drawing tools all inherit from mitk Tool which is a mitk StateMachine There is a number of deriva
34. dvanced Settings BO Threshold 0 Output BO Image Spherical Harmonics Lambda 0 006 Figure 6 6 Advanced q ball reconstruction settings Generated on Thu Nov 3 17 31 07 2011 for DI_App_docu by Doxygen 6 5 Dicom Import 47 This is how a q ball image should initially look after reconstruction Standard q balls feature a relatively low GFA and thus appear rather dark Adjust the level window to solve this File Edit Window Help E Open Save Project 4 Close Project E Undo ES Redo 5 Check for updates Image Navigator Data Manager Display Ed X Diffusion Image_QAO k Diffusion Image E X Q Balls Tensors N Quantification 5 DICOM Import 3 Reconstruction Advanced Settings Standard SH ooo Spherical harmonics recon Desoteaux2007 E Start Reconstruction 100 Transversal 19 Sagittal 48 Coronal 41 Time 0 WAE 1 52 GB 19 79 Figure 6 7 q ball image after reconstruction 6 5 Dicom Import The dicom import does not cover all hardware manufacturers but only Siemens dicom images MITK DI is also capable of reading the nrrd format which is documented elsewhere 1 2 These files can be created by combining the raw image data with a corresponding textual header file The file extension should be changed from x nrrd to dwi or from nhdr to x hdwi respectively in order to let MITK DI recognize the diffusion related header information provided in
35. e Project E4 Close Project e Undo E Redo O Check for updates Image Navigator Data Manager Display x Diffusion Image_dti Ly 3 Diffusion Image_QAO Diffusion Image ME El 1500 min max ize v Scaling 1 00 ional scaling y Opacity Min 0 41 Opacity Max 0 09 Q Balls Tensors Quantification 5 DICOM Import Reconstruction Advanced Settings Standard SH xj Spherical harmonics recon Desoteaux2007 Transversal 19 Sagittal 48 li Coronal a Time 0 NAE 2 26 GB 29 43 Figure 6 10 Q ball image with ODF glyph visibility toggled ON 6 8 References 1 http teem sourceforge net nrrd format html 2 http www cmake org Wiki Getting_Started_with_the_NRRD_Format 3 C F Westin S E Maier H Mamata A Nabavi F A Jolesz R Kikinis Processing and visualization for Diffusion tensor MRI Medical image Analysis 2002 pp 93 108 5 Tuch D S 2004 Q ball imaging Magn Reson Med 52 1358 1372 6 Descoteaux M Angelino E Fitzgibbons S Deriche R 2007 Regularized fast and robust analyti cal Q ball imaging Magn Reson Med 58 497 510 7 Aganj I Lenglet C Sapiro G 2009 ODF reconstruction in q ball imaging with solid angle consid eration Proceedings of the Sixth IEEE International Symposium on Biomedical Imaging Boston MA 8 Goh A Lenglet C Thompson P M Vidal R 2009 Estimating Orientation Distribution Functions with Pr
36. e detection removal feature If leakage happens you can left click into the leakage region and the tool will try to automatically remove this region see illustration below Generated on Thu Nov 3 17 31 07 2011 for DI_App_docu by Doxygen 10 The Segmentation Module Figure 2 7 Leakage correction feature of the Region Growing tool Figure 2 8 Correction Tool You do not have to draw a closed contour to use the Correction tool and do not need to switch between the Add and Substract tool to perform small corrective changes The following figure shows the usage of this tool e if the user draws a line which starts and ends outside the segmenation a part of it is cut off left image if the line is drawn fully inside the segmentation the marked region is added to the segmentation right image Generated on Thu Noy 3 17 31 07 2011 for DI_App_docu by Doxygen 2 4 Manual Contouring 11 Figure 2 9 Actions of the Correction tool illustrated Le Fill Figure 2 10 Fill Tool Left click inside a segmentation with holes to completely fill all holes oP Erase Figure 2 11 Erase Tool This tool removes a connected part of pixels that form a segmentation You may use it to remove so called islands see picture or to clear a whole slice at once hold CTRL while clicking 2 4 5 Interpolation Creating segmentations for modern CT volumes is very time consuming because strucutres of interest can easily cover
37. eatures e Usage 12 1 Overview MovieMaker is a functionality for easily creating fancy movies from scenes displayed in MITK widgets It is also possible to slide through your data automatically rotate 3D scenes and take screenshots of widgets 12 2 Features The Movie Maker allows you to create movies and screenshots from within MITK It can automatically scroll thorugh timesteps and slices while recording a movie This way you can record visualizations like a beating heart or a rotating skull Generated on Thu Nov 3 17 31 07 2011 for DI_App_docu by Doxygen 12 3 Usage 75 12 3 Usage Write Movie MS Windows only EN Spatial Temporal Combined S T Relation Forward 0 Backward Ping Pong Cyde sec Figure 12 2 A view of the command area of QmitkMovieMaker Generated on Thu Nov 3 17 31 07 2011 for DI_App_docu by Doxygen 76 The Movie Maker Module 12 3 1 Window selection With the first two drop down boxes you can choose which window you want to step through and which window you want to record in Left clicking inside a window will set both drop down boxes to that window but you can choose different windows for stepping and recording The first drop down box defines the window along which slices will be stepped through if stepping is set to spatial see below The second denotes the window from which the content will be recorded 12 3 2 Recording Options The slider can be used t
38. ew This module is currently under heavy development and as such the interface as well as the capabilities are likely to change significantly between different versions This documentation describes the features of this current version Generated on Thu Nov 3 17 31 07 2011 for DI_App_docu by Doxygen 3 3 Overview 25 ak Brain Network Analysis Create Network Network Statistics of vertices 11 of edges 10 of self loops D average degree 1 81818 connection density 0 181818 efficiency 0 308989 Histograms Hoda degr Shortest patr Figure 3 2 The interface Generated on Thu Nov 3 17 31 07 2011 for DI_App_docu by Doxygen 26 The Brain Network Analysis Module 3 4 Usage To create a network select first a parcellation of the brain e g as provided by freesurfer by CTRL Leftclick and secondly a fiber image as created using tractography module Then click on the Create Network button To calculate network statistics select a network in the datamanager At this time the following statistics are calculated for the entire network The number of vertices in the network The number of edges in the network The number of edges which have the same vertex as beginning and end point The average degree of the nodes in the network The connection density the network the number of edges divided by the number of possible edges The unweighted efficiency of the network 1 divided by average path
39. f f D and Dx maps activate the checkmarks and press Generate Output Images The neglect b lt threshold allows you to ignore b values smaller then a threshold for the initial fit of f and D Dx is then estimated using all measurements The exact values of the current fit are always given in the legend underneath the figure 19 1 Region of interest analysis Create region of interest To create a new segmentatin open the quantification perspective select the tab Segmentation and create a segmentation of the structure of interest Alternatively of course you may also load a binary image from file or generate your segmentation in any other possible way IVIM in region of interset Go back to the VIM perspective and select both the diffusion image and the segmentation holding the CTRL key A red message should appear Averaging N voxels 19 2 Region of interest analysis All model parameters and corresponding curves can be exported to clipboard using the buttons underneath the figure Generated on Thu Nov 3 17 31 07 2011 for DI_App_docu by Doxygen 19 3 Region of interest analysis 101 19 3 Region of interest analysis Advanced users that know what they are doing can change the method for the model fit under Advanced Settings on the very bottom of the view 3 param fit linear fit of f D and fix Dx are among the options 19 4 Region of interest analysis Toward an optimal distribution of b values for intr
40. f the tensor image as BO Generated on Thu Nov 3 17 31 07 2011 for DI_App_docu by Doxygen 46 MITK Diffusion Imaging MITK DI The gradient images are afterwards generated using the standard tensor equation 6 4 Q Ball Reconstruction The q ball reonstruction bundle implements a variety of reconstruction methods The different reconstruc tion methods are described in the following e Numerical The original numerical q ball reconstruction presented by Tuch et al 5 e Standard SH Descoteaux s reconstruction based on spherical harmonic basis functions 6 Solid Angle SH Aganj s reconstruction with solid angle consideration 7 e ADC profile only The ADC profile reconstructed with spherical harmonic basis functions e Raw signal only The raw signal reconstructed with spherical harmonic basis functions gt QBalls Tensors Quantification DICOM Import 3 Reconstruction Advanced Settings SH recon with solid angle consideration Aganj2009 E Start Reconstruction Figure 6 5 The q ball resonstruction view BO threshold works the same as in tensor reconstruction The maximum l level configures the size of the spherical harmonics basis Larger l values e g 1 8 allow higher levels of detail lower levels are more stable against noise e g 1 4 Lambda is a regularisation parameter Set it to O for no regularisation lambda 0 006 has proven to be a stable choice under various settings w A
41. f they contain a symbol function pointer itkLoad which is expected to return a pointer to a 1tk ObjectFactoryBase instance If such a symbol is found the returned factory will be registered with the ITK object factory If you successfully followed all the steps above MITK will find your mitk ExternalTool on application startup when the ITK object factory is asked to create all known instances of mitk Tool Furthermore if your mitk ExternalTool claims to be part of the default group there will be a new icon in Segmentation which activates your tool Have fun And Windows users welcome to the world of DLLs Generated on Thu Nov 3 17 31 07 2011 for DI_App_docu by Doxygen 22 The Segmentation Module Generated on Thu Nov 3 17 31 07 2011 for DI_App_docu by Doxygen Chapter 3 The Brain Network Analysis Module 24 The Brain Network Analysis Module alts Figure 3 1 Icon of the Module 3 1 Summary This module can be used to create a network from a parcellation and a fiber image as well as to calculate and display network statistics This document will tell you how to use this module but it is assumed that you already know how to use MITK in general Please see Details for more detailed information on usage and supported filters If you encounter problems using the module please have a look at the Troubleshooting page 3 2 Details Manual sections e Overview e Usage e Troubleshooting 3 3 Overvi
42. gmentation perspective and how they are used 2 2 Technical Issues The Segmentation perspective makes a number of assumptions To know what this module can be used for 1t will help you to know that e Images must be 2D 3D or 3D t e Images must be single values i e CT MRI or normal ultrasound Images from color doppler or photographic RGB images are not supported e Segmentations are handled as binary images of the same extent as the original image 2 3 Image Selection The Segmentation perspective makes use of the Data Manager view to give you an overview of all images and segmentations Generated on Thu Nov 3 17 31 07 2011 for DI_App_docu by Doxygen 2 4 Manual Contouring 7 Data Manager a E ms Patient ma ne Liver E E Spleen Figure 2 3 Data Manager is used for selecting the current segmentation The reference image is selected in the drop down box of the control area To select the reference image e g the original CT MR image use the drop down box in the control area of the Segmentation view The segmentation image selected in the Data Manager is displayed below the drop down box If no segmentation image exists or none is selected create a new segmentation image by using the New segmentation button Some items of the graphical user interface might be disabled when no image is selected In any case the application will give you hints if a selection is needed 2 4 Manual Contouring With manual c
43. gmentations The class also performs this interpolation for single slices on demand Again we have a second class responsible for the GUI QmitkSlicesInterpolator enables disables interpolation and offers to accept interpolations for one or all slices 6 Undo Undo functionality is implemented in mitk OverwriteSliceImageFilter since this is the cen tral place where all image modifications are made The filter stores a binary difference image to the undo stack as a mitk ApplyDifflmageOperation When the user requests undo this ApplyDifflma geOperation will be executed by a singleton class DifflmageApplier The operation itself observes the image which it refers to for itk DeleteEvent so no undo operation will be executed on for images that have already been destroyed 7 GUI The top level GUI is the functionality QmitkSegmentation which is very thin in comparison to ERIS There are separate widgets for image and tool selection for interpolation Additionaly there are some methods to create delete crop load and save segmentations 2 11 How to extend the Segmentation bundle with external tools e Introduction e What might be part of an extension Tool classes GUI classes for tools Additional files e Writing a CMake file for a tool extension e Compiling the extension e Configuring ITK autoload 2 11 1 Introduction The application for manual segmentation in MITK Segmentation bundle comes with a tool class frame
44. he box in front of the Data Element in the Datamanager shows the visibility A green filled box means a visible Data Element an empty box means an invisible Data Element see Screenshot1 Generated on Thu Nov 3 17 31 07 2011 for DI_App_docu by Doxygen 4 4 Working with the Datamanager 4 4 3 Representation of Data Elements There are different types of representations how to show the Data Element inside the standard view By right clicking on the Data Element all options are listed see Screenshot2 and Screenshot 3 An arbitrary color can be chosen The opacity can be changed with a slide control In case of images a texture interpolation can be switched on or off The texture interpolation smoothes the image so that no single pixels are visible anymore In case of surfaces the surface representation can be changed between points wireframe or surface Global reinit updates all windows to contain all the current data Reinit updates a single data item fits the windows to contain this data item EM Segmentation Mitk3M3 ITK 3 16 0 VTK 5 4 2 Ot 4 5 3 MITK 20559 File Edit Window Help 4 E Open Save Project Close Project K Undo Redo Data Manager Global Reinit ar im Save XX Delete O Reinit 32 Show only selected nodes A Togale visibility A Details Opacity 7 Color i Cont sx Cont Threshold meee Y Texture Interpolation Figure 4 4 Screenshot2 Pro
45. ibbs Tracking algorithm a global fiber tracking algorithm originally proposed by Reisert et al 1 Available sections e Input Data e Q Ball Reconstruction e Surveilance of the tracking process e References Generated on Thu Nov 3 17 31 07 2011 for DI_App_docu by Doxygen 9 1 Input Data 61 Stochastic Tracking Fiber Bundle Operations Mask Image gt N A GFA Image gt N A Iterations 10 7 Y Visualize Tractography Y Advanced Settings Particle Length E Particle Width Particle Weight 8 Start Temperature 0 1 _ End Temperature 0 001 _ Balance In Ex Energy 0 _ Min Fiber Length Y Subtract ODF Mean Save Parameters F Load Parameters gt Start Tractography El Stop Tractography Progress Accepted Fibers Connections Particles Proposal Acceptance Rate Tracking Time Figure 9 1 The Gibbs Tracking View 9 1 Input Data Mandatory Input e One Q Ball image selected in the datamanager Generated on Thu Nov 3 17 31 07 2011 for DI_App_docu by Doxygen 62 Gibbs Tracking View Optional Input e Mask Image Float image used as probability mask for the generation of fiber segments Usually used as binary brain mask to reduce the searchspace of the algorithm and to avoid fibers resulting from noise outside of the brain GFA Image Float image used to automatically determine the particle weight para
46. in 2D slices of 3D or 3D t images Re implementation was chosen because it seemed to be easier to write documentation and tests for newly developed code In addition the old code had some design weaknesses e g a monolithic class which would be hard to maintain in the future By now Segmentation is a well tested and easily extensible vehicle for all kinds of interactive segmentation applications A separate page describes how you can extend Segmentation with new tools in a shared object DLL How to extend the Segmentation bundle with external tools 2 10 2 Overview of tasks We identified the following major tasks Management of images what is the original patient image what images are the active segmenta tions Management of drawing tools there is a set of drawing tools one at a time is active that is someone has to decide which tool will receive mouse and other events Drawing tools each tool can modify a segmentation in reaction to user interaction To do so the tools have to know about the relevant images Slice manipulation drawing tools need to have means to extract a single slice from an image volume and to write a single slice back into an image volume Interpolation of unsegmented slices some class has to keep track of all the segmentations in a volume and generate suggestions for missing slices This should be possible in all three orthogonal slice direction Undo Slice manipulations shoul
47. ith other kind of images so here is a list of the image types that were used for training e Hippocampus segmentation Tl weighted MR images 1 5 Tesla scanner Magnetom Vision Siemens Medical Solutions 1 0 mm isotropic resolution e Heart Left ventricle inner segmentation LV Model MRI velocity encoded cine VEC cine MRI sequence trained on systole and diastole e Heart Left ventricular wall segmentation LV Inner Wall LV Outer Wall 4D MRI short axis 12 slice spin lock sequence SA_12_ sl trained on whole heart cycle Generated on Thu Nov 3 17 31 07 2011 for DI_App_docu by Doxygen 2 6 Lesion Segmentation 13 e Lung segmentation 3D and 4D MRI works best on FLASH3D and TWIST4D sequences 2 5 3 Other Organs As mentioned in the Heart Lung section most of the underlying methods are based on training The basic algorithm is versatile and can be applied on all kinds of segmentation problems where the structure of interest is topologically like a sphere and not like a torus etc If you are interested in other organs than those offered by the current version of the Segmentation view please contact our research team 2 6 Lesion Segmentation Lesion segmentation is a little different from organ segmentation Since lesions are not part of the healthy body they sometimes have a diffused border and are often found in varying places all over the body The tools in this section offer efficient ways to create 3D segmentations
48. lank It is also possible to check for negative eigenvalues The according voxels are also left blank Generated on Thu Nov 3 17 31 07 2011 for DI_App_docu by Doxygen 6 3 Tensor Reconstruction 45 6 Preprocessing Tensors x Q Balls D ODF Details ITK Reconstruction Y Advanced Settings BO Threshold 0 Check for negative eigenvalues gt ITK Tensor Reconstruction Figure 6 3 ITK tensor reconstruction A few seconds depending on the image size after the reconstruction button is hit a colored image should appear in the main window Data Manager E Diffusionimage_dti MO Diffusionimage Preprocessing Tensors X Q Balls ODF Details ITK Reconstruction MZ Advanced Settings BO Threshold o Check for negative eigenvalues Estimate Diffusion Image from Tensors B Value 1000 Gradient Directions 162 Estimate Q Ball Image from Tensors Teem Reconstruction Advanced Settings Transversal e 19 Sagittal o 48 Coronal emm 41 Time o Position lt 0 00 4 24 42 18 gt mm Index lt 48 41 19 gt Time 0 00 ms Pixelvalue 0 00 1 44 GB 6 11 Figure 6 4 Tensor image after reconstruction The view also allows the generation of artificial diffusion weighted or Q Ball images from the selected tensor image The ODFs of the Q Ball image are directly initialized from the tensor values and afterwards normalized The diffusion weighted image is estimated using the 12 norm image o
49. length this is zero for discon nected graphs Furthermore some statistics are calculated on a per node basis and displayed as histograms e The degree of each node e The unweighted betweenness centrality of each node e The spread of shortest paths between each pair of nodes For disconnected graphs the shortest paths with infinite length are omitted for readability 3 5 Troubleshooting No known problems All other problems Please report to the MITK mailing list See http www mitk org wiki Mailinglist on how to do this Generated on Thu Nov 3 17 31 07 2011 for DI_App_docu by Doxygen Chapter 4 The DataManager 28 The DataManager Figure 4 1 Icon of the Module 4 1 Introduction The Datamanager is the central componenent to manage medical data like images surfaces etc After loading one or more data into the Datamanager the data are shown in the four view window the so called Standard View The user can now start working on the data by just clicking into the standard view or by using the MITK modules such as Segmentation or Basic Image Processing Available sections Introduction e Loading Data e Saving Data e Working with the Datamanager List of Data Elements Visibility of Data Elements Representation of Data Elements Preferences e Property List Generated on Thu Nov 3 17 31 07 2011 for DI_App_docu by Doxygen 4 2 Loading Data 29 AMA Aik AA ITK 9
50. llow for quick corrections Figure 2 5 Paint and Wipe Tools Use the slider below the toolbox to change the radius of these round paintbrush tools Move the mouse in any 2D window and press the left button to draw or erase pixels As the Add Subtract tools holding CTRL while drawing will invert the current tool s behavior Region Growing Figure 2 6 Region Growing Tool Click at one point in a 2D slice widget to add an image region to the segmentation with the region growing tool Moving up the cursor while holding the left mouse button widens the range for the included grey values moving it down narrows it When working on an image with a high range of grey values the selection range can be influenced more strongly by moving the cursor at higher velocity Region Growing selects all pixels around the mouse cursor that have a similar gray value as the pixel below the mouse cursor This enables you to quickly create segmentations of structures that have a good contrast to surrounding tissue e g the lungs The tool will select more or less pixels corresponding to a changing gray value interval width when you move the mouse up or down while holding down the left mouse button A common issue with region growing is the so called leakage which happens when the structure of interest is connected to other pixels of similar gray values through a narrow bridge at the border of the structure The Region Growing tool comes with a leakag
51. load the threshold color and gradient function combined in a single xml file 20 4 Interactively create transferfunctions Beside the possibility to directly edit the transferfunctions Customize transferfunctions in detail a one click generation of two commonly known shapes is given Both generators have two parameters that can be modified by first clicking on the cross and then moving the mouse up down and left right The first parameter center controlled by horizontal movement of the mouse specifies the gravalue where the center of the shape will be located The second parameter width controlled by vertical movement of the mouse specifies the width or steepness of the shape Generated on Thu Nov 3 17 31 07 2011 for DI_App_docu by Doxygen 108 The Volume Visualization Module 20 4 1 Threshold Click on the Threshold tab to active the threshold function generator and mowe the mouse Grayvalue gt A threshold shape begins with zero and raises to one across the center parameter Lower widths results in steeper threshold functions Generated on Thu Nov 3 17 31 07 2011 for DI_App_docu by Doxygen 20 5 Customize transferfunctions in detail 109 20 4 2 Bell Click on the Bell tab to active the threshold and move the mouse Grayvalue gt Opacity function generator A threshold shape begins with zero and raises to one at the center parameter and the lowers agains to zero
52. mes preferrably with a contrast agent in the portal venous phase the Liver tool will fully automatically analyze and segment the image All you have to do is to load and select the image then click the Liver button During the process which takes a minute or two you will get visual progress feedback by means of a contour that moves closer and closer to the real liver boundaries 2 5 2 Heart Lung and Hippocampus on MRI While liver segmentation is performed fully automatic the following tools for segmentation of the heart the lungs and the hippocampus need a minimum amount of guidance Click one of the buttons on the Organ segmentation page to add an average model of the respective organ to the image This model can be dragged to the right position by using the left mouse button while holding down the CTRL key You can also use CTRL middle mouse button to rotate or CTRL right mouse button to scale the model Before starting the automatic segmentation process by clicking the Start segmentation button try placing the model closely to the organ in the MR image in most cases you do not need to rotate or scale the model During the segmentation process a green contour that moves closer and closer to the real liver boundaries will provide you with visual feedback of the segmentation progress The algorithms used for segmentation of the heart and lung are method which need training by a number of example images They will not work well w
53. meter 9 2 Q Ball Reconstruction Number of iterations More iterations causes the algorithm to be more stable but also to take longer to finish the tracking Recommended 104 7 107 9 iterations Particle length width weight controlling the contribution of each particle to the model M Start and end temperature controlling how fast the process reaches a stable state usually no change needed e Weighting between the internal affinity of the model to long and straigt fibers and external energy affinity of the model towards the data usually no change needed e Minimum fiber length constraint Fibers containing less segments are discarded after the tracking usually no change needed 9 3 Surveilance of the tracking process Once started the tracking can be monitored via the textual output that informs about the tracking progress and several stats of the current state of the algorithm If enabled the intermediate tracking results are displayed in the renderwindows each second This live visualization should usually be disabled for perfor mance reasons It can be turned on and off during the tracking process via the according checkbox 9 4 References 1 Reisert M Mader I Anastasopoulos C Weigel M Schnell S Kiselev V Global fiber recon struction becomes practical Neuroimage 54 2011 955 962 Generated on Thu Nov 3 17 31 07 2011 for DI_App_docu by Doxygen Chapter 10 The Image Statistics Module 64
54. mpute some features of a whole image or a region of interest Generated on Thu Nov 3 17 31 07 2011 for DI_App_docu by Doxygen 14 3 Overview 81 1835 42 1020 052 604 516 345 180 Figure 14 2 The interface 82 The Image Statistics Module 14 4 Usage After selection of an image or a binary mask of an image in the datamanader the Image Statistics module shows some statistical information If a mask is selected the name of the mask and the name of the image this mask is applied to are shown at the top Below that are the statistics window which displays the calculated statistical features such as mean stan dard deviation and the histogram At the bottom of the module are two buttons They copy their respective data in csv format to the clipboard 14 5 Troubleshooting No known problems All other problems Please report to the MITK mailing list See http www mitk org wiki Mailinglist on how to do this Generated on Thu Nov 3 17 31 07 2011 for DI_App_docu by Doxygen Chapter 15 Partial Volume Analysis 84 Partial Volume Analysis The Partial Volume Analysis view can be found in the Quantification perspective It allows for robust quantification of diffusion or other scalar measures in the presents of two classes e g fiber vs non fiber and partial volume between them The algorithm estimates a probabilistic segmentation of the three classes and returns a weighted average of the meas
55. ngth Tract_len number of total tracts per voxel TotalTracts and likelihood cache size in MB Lkhd chache are individually set by the user e After successfully setting necessary Input and Parameter pressing the command button executes the algorithm 17 3 References 1 Tri M Ngo Polina Golland and Tri M Ngo A stochastic tractography system and applications 2007 Generated on Thu Nov 3 17 31 07 2011 for DI_App_docu by Doxygen 92 Stochastic Tracking View Generated on Thu Nov 3 17 31 07 2011 for DI_App_docu by Doxygen Chapter 18 The Tbss Analysis Module 94 The Tbss Analysis Module Figure 18 1 Icon of the Module 18 1 Summary This module can be used to locally explore data resulting from preprocessing with the TBSS module of FSL This document will tell you how to use this module but it is assumed that you already know how to use MITK in general and how to work with the TBSS module of FSL Please see Details for more detailed information on usage and supported filters If you encounter problems using the module please have a look at the QmitkTractbasedS patialStatisticsUserManualTrouble page 18 2 Details Available sections e Overview e FSL Import e Regions of interest e Profile plots e Troubleshooting e References Generated on Thu Nov 3 17 31 07 2011 for DI_App_docu by Doxygen 18 3 Overview 95 18 3 Overview Y FSL import Subject data Grou
56. o step through the slices manually while not recording Start and stop control a preview of what a video would look like The buttons in the bottom part of this section can be used to create movies windows only or screenshots Clicking opens a file dialog where a name can be selected After confirmation a screenshot or movie is created according to the playing options 12 3 3 Playing Options The first section controls whether the movie steps through slices if a 2D view is selected rotate the shown scene if a 3D view is selected or step through time steps if set to temporal and a time resolved dataset is selected If set to combined a combination of both above options is used with their speed relation set via the S T Relation Spinbox In the second section the direction of stepping can be set Options are Forward backward and Ping Pong which is back and forth The stepping speed can be set via the spinbox total time in seconds Although stepping speed is a total time in sec this can not always be achieved As a minimal frame rate of 25 fps is assumed to provide smooth movies a dataset with only 25 slices will always be stepped through in 1 sec or faster Generated on Thu Nov 3 17 31 07 2011 for DI_App_docu by Doxygen Chapter 13 ODF Details View 78 ODF Details View This view provides detailed information about the orentation distribution function at the current crosshair position if a Q Ball image is selected A
57. obability Density Constraints and Spatial Regularity Med Image Comput Comput Assist Interv Int Conf Med Image Comput Comput Assist Interv LNCS 5761 877 ff Generated on Thu Nov 3 17 31 07 2011 for DI_App_docu by Doxygen 6 9 Technical Information for Developers 51 6 9 Technical Information for Developers The diffusion imaging module uses additional properties beside the ones in use in other modules for further information see DiffusionImagingPropertiesPage Generated on Thu Nov 3 17 31 07 2011 for DI_App_docu by Doxygen 52 MITK Diffusion Imaging MITK DID Generated on Thu Nov 3 17 31 07 2011 for DI_App_docu by Doxygen Chapter 7 Diffusion Bundles in Development 54 Diffusion Bundles in Development 7 1 Experimental Bundles The bundles in this section are not yet open source Usually this is due to the fact that they are relatively new and under heavy development which can lead to significant and sudden changes in the source code interface These will be moved open source as soon as a certain degree of code stability can be expected Generated on Thu Nov 3 17 31 07 2011 for DI_App_docu by Doxygen Chapter 8 Fiber Processing View 56 Fiber Processing View This view provides everything needed to process fiber bundles Available sections e Fiber Bundle Manipulation Generation of additional information from fiber bundles Gibbs Tracking Stochastic Tracking OO 2 Extract fr
58. of such lesions The Segmentation View currently offers supoprt for enlarged lymph nodes To segment an enlarged lymph node find a more or less central slice of it activate the Lymph Node tool and draw a rough contour on the inside of the lymph node When releaseing the mouse button a segmentation algorithm is started in a background task The result will become visible after a couple of seconds but you do not have to wait for it If you need to segment several lymph nodes you can continue to inspect the image right after closing the drawn contour If the lymph node segmentation is not to your content you can select the Lymph Node Correction tool and drag parts of the lymph node surface towards the right position works in 3D not slice by slice This kind of correction helps in many cases If nothing else helps you can still use the pure manual tools as a fallback 2 7 Things you can do with segmentations As mentioned in the introduction segmentations are never an end in themselves Consequently the Seg mentation view adds a couple of post processing actions to the Data Manager These actions are acces sible through the context menu of segmentations in Data Manager s list view Generated on Thu Nov 3 17 31 07 2011 for DI_App_docu by Doxygen 14 The Segmentation Module E Patient Mm 7 Liver cted nodes Show only selected nodes y of selected nodes Show additional information for selected nodes Apply O
59. om ROIs Y Generate ROI Image 2 Join Bundles gt Substract Bundles Fibers in Selected Bundle Tract Density Image TDI gt Generate Invert Figure 8 1 The Fiber Processing View S 1 Fiber Bundle Manipulation Fiber extraction Place ROIs in the 2D render widgets cricles or polygons and extract fibers from the bundle that pass through these ROIs by selecting the according ROI and fiber bundle in the datamanger and starting the extraction The ROIs can be combined via logical operations All fibers that pass through the thus generated composite ROI are extracted Generated on Thu Nov 3 17 31 07 2011 for DI_App_docu by Doxygen 8 2 Generation of additional information from fiber bundles 57 8 2 Generation of additional information from fiber bundles e Tract density image generate a 2D heatmap from a fiber bundle e Binary envelope generate a binary image from a fiber bundle e Fiber bundle image generate a 2D rgba image representation of the fiber bundle e Fiber endings image generate a 2D binary image showing the locations of fiber endpoints e Fiber endings pointset generate a poinset containing the locations of fiber endpoints Generated on Thu Nov 3 17 31 07 2011 for DI_App_docu by Doxygen 58 Fiber Processing View Generated on Thu Nov 3 17 31 07 2011 for DI_App_docu by Doxygen Chapter 9 Gibbs Tracking View 60 Gibbs Tracking View This view provides the user interface for the G
60. ontouring you define which voxels are part of the segmentation and which are not This allows you to create segmentations of any structeres that you may find in an image even if they are not part of the human body You might also use manual contouring to correct segmentations that result from sub optimal automatic methods The drawback of manual contouring is that you might need to define contours on many 2D slices However this is moderated by the interpolation feature which will make suggestions for a segmentation Generated on Thu Nov 3 17 31 07 2011 for DI_App_docu by Doxygen 8 The Segmentation Module 2 4 1 Creating New Segmentations Unless you want to edit existing segmentations you have to create a new empty segmentation before you can edit it To do so click the New manual segmentation button Input fields will appear where you can choose a name for the new segmentation and a color for its display Click the checkmark button to confirm or the X button to cancel the new segmentation Notice that the input field suggests names once you start typing and that it also suggests colors for known organ names If you use names that are not yet known to the application it will automatically remember these names and consider them the next time you create a new segmentation Once you created a new segmentation you can notice a new item with the binary mask icon in the Data Manager tree view This item is automatically selected for you
61. open source part of the MITK 3M3 Application Available sections e Overview Technical Issues e Image Selection e Manual Contouring Creating New Segmentations Selecting Segmentations for Editing Selecting Editing Tools Using Editing Tools e Interpolation e Organ Segmentation Liver on CT Images Heart Lung and Hippocampus on MRI Other Organs e Lesion Segmentation e Things you can do with segmentations e Surface Masking e Technical Information for Developers 2 1 Overview The Segmentation perspective allows you to create segmentations of anatomical and pathological struc tures in medical images of the human body The perspective groups a number of tools which can be used for e semi automatic segmentation of organs on CT or MR image volumes e semi automatic segmentation of lesions such as enlarged lymph nodes or tumors e manual segmentation of any strucutures you might want to delineate Generated on Thu Nov 3 17 31 07 2011 for DI_App_docu by Doxygen 2 1 Overview 5 OD Segmentation Mitk3M3 ITK 3 16 0 VTK 5 4 0 Qt 4 5 1 MITK 20157 fem cr image OOO a O alaaa a Liver o Paint g pi Region Growing Correction amp a i Fill Erase 0 gt 2 a Figure 2 2 Segmentation perspective consisting of the Data Manager view and the Segmentation view If you wonder what segmentations are good for we shortly revisit the concept of a segmentation here A CT or MR image is
62. ossing fibers Available sections E Open Known Issues Preprocessing Tensor Reconstruction Q Ball Reconstruction Dicom Import Quantification ODF Visualization Setting References Technical Information for Developers Data Manager X Qball Image _ Diffusion Image A T1 Morphology DTI Image We X Q Balls Tensors N Quantification 5 DICOM Import 3 Reconstruction Advanced Settings Solid angie SH SH recon with solid angle consideration Aganj2009 E Start Reconstruction Transversal 29 Sagittal Coronal Time MiB E gt 256 o 224 w Save Project X Close Project E Undo 5 Redo O Check for updates Image Navigator Display x 1 52 GB 19 79 Figure 6 1 The MITK Diffusion Imaging Module Generated on Thu Nov 3 17 31 07 2011 for DI_App_docu by Doxygen 6 1 Known Issues 43 6 1 Known Issues e Dicom Import The dicom import has so far only been implemented for Siemens dicom images MITK DI is capable of reading the nrrd format which is documented elsewhere 1 2 These files can be created by combining the raw image data with a corresponding textual header file The file extension should be changed from nrrd to dwi or from nhdr to x hdwi respectively in order to let MITK DI recognize the diffusion related header information provided in the files 6 2 Preprocessing The preprocessing view gives an overview o
63. ough code research needed software due to the BSD like software license Research institutes will profit from the high level of integration of ITK and VTK enhanced with data management advanced visualization and interaction functionality in a single framework that is supported by a wide variety of researchers and developers You will not have to reinvent the wheel over and over and can concentrate on your work Medical Professionals will profit from MITK and the MITK applications by using its basic functionalities for research projects But nonetheless they will be better off unless they are programmers themselves to cooperate with a research institute developing with MITK to get the functionalitiy they need MITK and the MITK applications are not certified medical products and may be used in a research setting only They must not be used in patient care 5 2 The User Interface The layout of the MITK applications is designed to give a clear distinction between the different work areas The following figure gives an overview of the main sections of the user interface Generated on Thu Nov 3 17 31 07 2011 for DI_App_docu by Doxygen 5 3 Four Window View 37 Menu Datamanager Levelwindow Active Module CETT Four Window View Memory Usage Figure 5 1 The Common MITK Application Graphical User Interface The datamanager and the Perspectives have their own help sections This document explains the use of The Four Windo
64. p Size Add Remove Import subject data L mm J Diffusion measure Fractional Anisotropy Tools ROIs Measuring mean FA skeleton Points on Roi 04 008 Create ROI Name Iroiname Structure info Structure Figure 18 2 Tbss Module Generated on Thu Nov 3 17 31 07 2011 for DI_App_docu by Doxygen 96 The Tbss Analysis Module This module is currently under heavy development and as such the interface as well as the capabilities are likely to change significantly between different versions Two limited parts of this module have been selected for inclusion in MITK DI 1 0 This documentation describes the features of this current version 18 4 FSL Import File Edit window Help E Open FEF save Project 5X close Project W Undo Redo image Navigator ON Sa LE Data Manager Display x e TBSS y all_FA_skeletonised E FSL import Subject data Group Size Healthy 23 Diseased 24 mip channel N Remove L into the MITK TBSS module Import subject data Diffusion measure Fractional Anisotropy Tools ROIs Measuring 1 000 5 800 600 4 400 4 show Welcome Screen 2004 Transversal el 90 1 Sagittal 91 o Coronal 109 om A REE a z o 200 400 60 s00 1000 Time o Position lt 1 00 18 00 18 00 gt mm Index lt 91 108 90 gt Time 0 00 ms Pixelvalue 0 00 2 33 GB 30 11 The FSL import allows to import data that has been preprocessed by FSL
65. part of the Data Manager as a node of the image tree 11 2 2 Save the image with measurement information After applying the wanted measurement figures the entire scene consisting of the image and the measure ment figures can be saved for future use Therefore just click the right mouse button when over the image item in the Data Manager and choose the item Save in the opening item list Following to that a save dialog appears where the path to the save folder can be set Afterwards just accept your choice with the save button 11 2 3 Remove measurement figures or image Tf the single measurement figures or the image is not needed any longer it can be removed solely or as an entire group The image can t be removed without simultaneously removing all the dependent measurement figures that belong to the image tree in the Data Manager To remove just select the wanted items in the data manager list by left click on it or if several items wanted to be removed left click on all wanted by simultaneously holding the ctrl button pressed For more detailed usage of the save remove functionality refer to the Data Manager User Manual Generated on Thu Nov 3 17 31 07 2011 for DI_App_docu by Doxygen 72 The Measurement Module Generated on Thu Nov 3 17 31 07 2011 for DI_App_docu by Doxygen Chapter 12 The Movie Maker Module 74 The Movie Maker Module Me Figure 12 1 Icon of the Module Available sections e Overview e F
66. perties for images Generated on Thu Nov 3 17 31 07 2011 for DI_App_docu by Doxygen 32 The DataManager EM Segmentation Mitk3M3 ITK 3 16 0 VTK 5 4 2 Ot 4 5 3 MITK 20559 File Edit Window Help E Open Save Project Close Project D Undo 6 Redo Data Manager E 4 Image_3D Global Reinit fm Save Delete O Reinit A Show only selected nodes A Toggle visibility Q Details Opacity Color Select an image gss Contouring Surface Representation Points Wireframe New segmentation v Surface Editing tools Figure 4 5 Screenshot3 Properties for surfaces 4 4 4 Preferences For the datamanager there are already some default hotkeys like the del key for deleting a Data Element The whole list is seen in Screenshot4 From here the Hotkeys can also be changed The preference page is found in Window gt Preferences Generated on Thu Nov 3 17 31 07 2011 for DI_App_docu by Doxygen 4 5 Property List 33 X Preferences Themes General Delete selected nodes Del B DataManager Global Reinit Ctrl R Segmentation Load Ctrl L Make all nodes invisible Ctrl Yo Reinit selected nodes R Save selected nodes Ctrl S Show Node Information Ctrl I Toggle visibility of selected nodes Y Import Export amp M close Figure 4 6 Screenshot4 4 5 Property List The Property List displays all the properties the
67. r q balls Gen eralized Fractional Anisotropy Generated on Thu Noy 3 17 31 07 2011 for DI_App_docu by Doxygen 6 7 ODF Visualization Setting 49 X Q Balls f Tensors Quantification 5 DICOM Import 3 Q Ball Imaging BR Tensor Imaging E FA Fractional Anisotropy E RA Relative Anisotropy E AD Axial Diffusivity E RD Radial Diffusivity E 1 A2 A3 2 A1 Figure 6 9 Anisotropy quantification 6 7 ODF Visualization Setting In this small view the visualization of ODFs and diffusion images can be configured Depending on the selected image in the data storage different options are shown here For tensor or q ball images the visibility of glyphs in the different render windows T ransversal S agittal and C oronal can be configured here The maximal number of glyphs to display can also be configured here for This is usefull to keep the system response time during rendering feasible The other options configure normalization and scaling of the glyphs In diffusion images a slider lets you choose the desired image channel from the vector of images each gradient direction one image for rendering Furthermore reinit can be performed and texture interpolation toggled This is how a visualization with activated glyphs should look like Generated on Thu Nov 3 17 31 07 2011 for DI_App_docu by Doxygen 50 MITK Diffusion Imaging MITK DID File Edit Window Help E Open m Sav
68. roject Close Project a Undo Redo Data Manager S a 4 Data Element1 M Data Element2 C Data Element3 C Data Element4 Figure 4 3 Screenshot1 The listed Data Elements are shown in the standard view Here the user can scale or rotate the medical objects or he can change the cutting planes of the object by just using the mouse inside this view 4 3 Saving Data There are two ways of saving data from the Datamanger The user can either save the whole project with all Data Elements by clicking on File gt Save Project or he can save single Data Elements by right clicking gt Save directly on a Data Element When saving the whole project the sorting of Data Elements is saved as well By contrast the sorting is lost when saving a single Data Element 4 4 Working with the Datamanager 4 4 1 List of Data Elements The Data Elements are listed in the Datamanager As described above the elements can be sorted hierar chically as a parent child relation For example after using the Segmentation Module on Data Element1 the result is created as Data Element2 which is a child of Data Element1 see Screenshot1 By drag and drop the sorting of Data Elements and their hierarchical relation can be changed 4 4 2 Visibility of Data Elements By default all loaded Data Elements are visible in the standard view The visibility can be changed by right clicking on the Data Element and then choosing Toogle visibility T
69. roup Size Add W mip jx Channel o Remove Diffusion measure Fractional Anisotropy Tools ROIs Measuring profiles on the Structure peda Healthy os AD pe AEN y gt J To plot tbss image with subject information and a region of int rresponding to the study and em both Show welcome Screen Transversal e 8s 0 68 Y Sagittal a g 0 66 Coronal 67 l Time o Position lt 1 00 24 00 13 00 gt mm Index lt 91 150 85 gt Time 0 00 ms Pixelvalue 1 00 2 83 GB 36 54 By selecting a tbss image with group information and a region of interest image as was created in a previous stap A profile plot is drawn in the plot canvas By clicking in the graph the crosshairs jump to the corresponding location in the image 18 7 Troubleshooting No known problems All other problems Please report to the MITK mailing list See http www mitk org wiki Mailinglist on how to do this 18 8 References 1 S M Smith M Jenkinson H Johansen Berg D Rueckert T E Nichols C E Mackay K E Watkins O Ciccarelli M Z Cader P M Matthews and T E J Behrens Tract based spatial statistics Voxelwise analysis of multi subject diffusion data NeuroImage 31 1487 1505 2006 2 S M Smith M Jenkinson M W Woolrich C F Beckmann T E J Behrens H Johansen Berg P R Bannister M De Luca I Drobnjak D E Flitney R Niazy J Saunders J Vickers Y Zhang N De
70. te the window menu It consists of three buttons Show crosshair Reset view No crosshair rotation Crosshair rotation Coupled crosshair rotation Swivel mode Figure 5 2 Crosshair The crosshair button allows you toggle the crosshair reset the view and change the behaviour of the planes Generated on Thu Nov 3 17 31 07 2011 for DI_App_docu by Doxygen 5 4 Menu 39 Activating either of the rotation modes allows you to rotate the planes visible in a 2D window by moving the mouse cursor close to them and click and dragging once it changes to indicate that rotation can be done The swivel mode is recommended only for advanced users as the planes can be moved freely by clicking and dragging anywhere within a 2D window The middle button expands the corresponding window to fullscreen within the four window view standard layout 2D images top 3D bottom 2D images left 3D right Big 3D Transversal plane Sagittal plane Coronal plane Coronal top 3D bottorn Coronal left 3D right Sagittal top Coronal n 3D bottom Transversal n Sagittal left 3D right Transversal n 3D left Sagittal right Figure 5 3 Layout Choices The right button allows you to choose between many different layouts of the four window view to use the one most suited to your task 5 4 Menu 5 4 1 File This dialog allows you to save load and clear entire projects this includes any nodes in the data manager 5 4 2 Edit This dialog
71. the gradient is controllable at specific grayvalues Generated on Thu Nov 3 17 31 07 2011 for DI_App_docu by Doxygen 112 The Volume Visualization Module Generated on Thu Nov 3 17 31 07 2011 for DI_App_docu by Doxygen Chapter 21 The Image Navigator 114 The Image Navigator Figure 21 1 Icon of the Module Image Navigator Es Transversal 100 Sagittal iw E Frontal P s7 E Time p 4 Figure 21 2 Image Navigator Fast movement through the available data can be achieved by using the Image Navigator By moving the sliders around you can scroll quickly through the slides and timesteps By entering numbers in the relevant fields you can jump directly to your point of interest Generated on Thu Nov 3 17 31 07 2011 for DI_App_docu by Doxygen
72. tions from Tool each offering some helper methods for specific sub classes like manipulation of 2D slices Tools are instantiated through the itk ObjectFactory which means that there is also one factory for each tool e g mitk AddContourToolFactory For the GUI represen tation each tool has an identification consisting of a name and an icon XPM The actual drawing methods are mainly implemented in mitk SegTool2D helper methods and its sub classes for region growing freehand drawing etc 4 Slice manipulation There are two filters for manipulation of slices inside a 3D im age volume mitk ExtractImageFilter retrieves a single 2D slice from a 3D volume mitk OverwriteSlicelmageFilter replaces a slice inside a 3D volume with a second slice which is a parameter to the filter These classes are used extensively by most of the tools to fulfill their task mitk OverwriteSlicelmageFilter cooperates with the interpolation classes to inform them of single slice modifications 5 Interpolation of unsegmented slices There are two classes involved in interpolation mitk SegmentationInterpolationController knows a mitk Image the segmentation and scans its Generated on Thu Nov 3 17 31 07 2011 for DI_App_docu by Doxygen 18 The Segmentation Module contents for slices with non zero pixels It keeps track of changes in the image and is always able to tell which neighbors of a slice in the three orthogonal slice directions contain se
73. tsu Filter Create polygon model Create smoothed polygon model Statistics Autocrop Figure 2 12 Context menu items for segmentations Create polygon model applies the marching cubes algorithms to the segmentation This polygon model can be used for visualization in 3D or other things such as stereolithography 3D printing e Create smoothed polygon model uses smoothing in addition to the marching cubes algorithms which creates models that do not follow the exact outlines of the segmentation but look smoother e Statistics goes through all the voxels in the patient image that are part of the segmentation and cal culates some statistical measures minumum maximum median histogram etc Note that the statistics are ALWAYS calculated for the parent element of the segmentation as shown in Data Man ager e Autocrop can save memory Manual segmentations have the same extent as the patient image even 1f the segmentation comprises only a small sub volume This invisible and meaningless margin is removed by autocropping 2 8 Surface Masking You can use the surface masking tool to create binary images from a surface which is used used as a mask on an image This task is demonstrated below Generated on Thu Nov 3 17 31 07 2011 for DI_App_docu by Doxygen 2 9 Technical Information for Developers 15 Data Manager vl a gt Surface Segmentation bi Segmentation x Reference image Segmentation gt s Conto
74. ure of interest within the each class Quantification MITK Diffusion Not for use in diagnosis or treatment of patients 1 14PM 2 Klaus Fritzsche 2 3 open Ga Save Project DX Close Project undo Ql Redo Image Navigator en Cale Data Manager i welcome 3 v E Diffusionimage_dti Circles WE 2 mip channeto Glyphs 500 Opacity 0 00 0 00 Min Max 2 scaling s Segmentation W PV Analysis Quantification Meis Image Diffusionimage_dti O O O Mask Circle1 Upsampling 25 Similar angles 90 display histogram 0 06 5 0014 T T E o o 02 03 04 OS 06 07 Green PV Red O Al Opacity Histoaram to Clinboard Position lt 20 00 6 74 52 18 gt mm Index lt 56 40 23 gt Time 0 00 ms Pixelvalue 0 00 2 36 GB 30 34 Figure 15 1 The Partial Volume Analysis View 15 1 Export All measures are automatically written to the clipboard once the estimation is updated The histogram export is provided by the button underneath the histogram The values can be pasted to excel or any text editor 15 2 Export Are not recommended for use yet 15 3 Export Diffusion tensor imaging in primary brain tumors reproducible quantitative analysis of corpus callosum infiltration and contralateral involvement using a probabilistic mixture model Stieltjes B Schl ter M Didinger B Weber MA Hahn HK Parzer P Rexilius J Konrad Verse O Peitgen HO Essig M Neuroim
75. uring Organs Lesions Surface Masking Select Surface For Image Masking Surface Create Segmentation From Surface Figure 2 13 Load an image and a surface Select the image and the surface in the corresponding drop down boxes both are selected automatically if there is just one image and one surface E Datamanager Display Data Manager v a Surface Y SegmentationSurface Reference image Segmentation saz Contouring Organs Lesions Surface Masking Select Surface For Image Masking Surface Create Segmentation From Surface Figure 2 14 After clicking Create segmentation from surface the newly created binary image is inserted in the DataManager and can be used for further processing 2 9 Technical Information for Developers For technical specifications see Technical design of QmitkSegmentation and for information on the exten sions of the tools system How to extend the Segmentation bundle with external tools Generated on Thu Nov 3 17 31 07 2011 for DI_App_docu by Doxygen 16 The Segmentation Module 2 10 Technical design of QmitkSegmentation e Introduction e Overview of tasks e Classes involved 2 10 1 Introduction QmitkSegmentation was designed for the liver resection planning project ReLiver The goal was a stable well documented extensible and testable re implementation of a functionality called ERIS which was used for manual segmentation
76. ver the important features of a diffusion weighted image like the number of gradient directions b value and the measurement frame Additionally it allows the extraction of the BO image and the generation of a binary brain mask The image volume can be modified by applying anew mesurement frame which is useful if the measurement frame is not set correctly in the image header or by averaging redundant gradient directions Generated on Thu Nov 3 17 31 07 2011 for DI_App_docu by Doxygen 44 MITK Diffusion Imaging MITK DI Tensors MX Q Balls ODF Details Info Number of Gradients b Value Reduce Size Multiple acquistions of one gradient direction can be averaged Due to rounding errors similar gradients often differ in the last decimal positions The Merge radius allows to average them anyway by taking into account all directions within a certain radius Merge radius 0 001000 a Y Average redundant gradients Non diffusion weighted image Average and extract all images that were acquired without diffusion weighting 2 Extract BO Brain Mask Y Estimate binary brain mask Measurment Frame Figure 6 2 Preprocessing 6 3 Tensor Reconstruction The tensor reconstruction view allows ITK based tensor reconstruction 3 The advanced settings for ITK reconstruction let you configure a manual threshold on the non diffusion weighted image All voxels below this threshold will not be reconstructed and left b
77. w View The Menu The Levelwindow e The System Load Indicator 5 3 Four Window View 5 3 1 Overview The four window view is the heart of the MITK image viewing The standard layout is three 2D windows and one 3D window with the transversal window in the top left quarter the sagittal window in the top right quarter the coronal window in the lower left quarter and the 3D window in the lower right quarter The different planes form a crosshair that can be seen in the 3D window Once you select a point within the picture informations about it are displayed at the bottom of the screen 5 3 2 Navigation Left click in any of the 2D windows centers the crosshair on that point Pressing the right mouse button and moving the mouse zooms in and out By scrolling with the mouse wheel you can navigate through the slices of the active window and pressing the mouse wheel while moving the mouse pans the image section Generated on Thu Nov 3 17 31 07 2011 for DI_App_docu by Doxygen 38 General MITK Manual In the 3D window you can rotate the object by pressing the left mouse button and moving the mouse zoom either with the right mouse button as in 2D or with the mouse wheel and pan the object by moving the mouse while the mouse wheel is pressed Placing the cursor within the 3D window and holding the F key allows free flight into the 3D view 5 3 3 Customizing By moving the cursor to the upper right corner of any window you can activa
78. with external tools 19 2 11 2 1 Tool classes A tool is basically any subclass of mitk Tool Tools are created at runtime through the ITK object factory so they inherit from itk Object Tools should handle the interaction part of a segmentation method Le create seed points draw contours etc in order to parameterize segmentation algorithms Simple algorithms can even be part of a tool A tools is identified by icon XPM format name short string and optionally a group name e g the group name for Segmentation is default There is a naming convention you should put a tool called mitk ExternalTool into files called mitkExternalTool h and mitkExternalTool cpp This is required if you use the conve nience macros described below because there need to be ITK factories which names are directly derived from the file names of the tools For the example of mitk ExternalTool there would be a factory called mitk ExternalToolFactory ina file named mitkExternalToolFactory cpp 2 11 2 2 GUI classes for tools Tools are non graphical classes that only implement interactions in renderwindows However some tools will need a means to allow the user to set some parameters a graphical user interface GUI In the Qt3 case tool GUIs inherit from QmitkToolGUI which is a mixture of QWidget and itk Object Tool GUIs are also created through the ITK object factory Tools inform their GUIs about state changes by messages Tool GU
79. wn problems All other problems Please report to the MITK mailing list See http www mitk org wiki Mailinglist on how to do this Generated on Thu Nov 3 17 31 07 2011 for DI_App_docu by Doxygen Chapter 11 The Measurement Module 68 The Measurement Module Figure 11 1 Icon of the Module 11 1 Features The bundle as it is depicted below offers the following features in the order of apperance on the image from DICOM MRI Image Draw Polygon Copy to Clipboard top to bottom The first information is the selected image s name here DICOM MRI Image followed by the measurement figures button line with the seven measurement figures From left to right the buttons are connected with the following functions 11 1 1 Draw Line Draws a line between two set points and returns the distance between these points 11 1 2 Draw Path Draws a path between several set points two and more and calculates the circumference that is all line s length summed up Add the final point by double left click Generated on Thu Nov 3 17 31 07 2011 for DI_App_docu by Doxygen 11 1 Features 69 11 1 3 Draw Angle Draws two lines from three set points connected in the second set point and returns the inner angle at the second point 11 1 4 Draw Four Point Angle Draws two lines that may but must not intersect from four set points The returned angle is the one depicted in the icon 11 1 5 Draw

here - MITK

Contents

Download Pdf Manuals

Related Search

Related Contents