Home

umi-uta-1162

Contents

1. 20 2 3 System TOP Vie eodeni ca aba bisa POLOS ca adds 20 Za TAP NIM Syster u acide itae bites 21 2 5 Resistance test Quantitative manual muscle test 22 3 1 Digital Crossbar Pin Assignments for the Serial Port and SP IANS ACES ut obras veo es nN 33 3 2 The Remote Sensor Module 1 RSM 1 52 3 3 Conceptual Block Diaeram for Acceleration and Rate SUBSYSTEMS ub asso vae 55 JA The HEMM Marn MOT aun 50 so The Generic Lest Ment nu an 60 2 0 Example 60 3 4 To ch Sensor Lay OE 69 2267 Three AXES Six Decrees ol unn aa EE 71 3 9 Flowchart part 1 for GTA 6 Hand Arm Steadiness Test 75 3 10 Flowchart part 2 for GTA 6 Hand Arm Steadiness Test 76 3 11 Flowchart part 3 for GTA 6 Hand Arm Steadiness Test 77 4 1 Finger Position Before LEDs Turn On in the Visual Iesponse peed 82 4 2 HPMM Position for the Grip Strength Test 82 4 3 Beginning Position in a Cycle During Upper EXttemity Coordinanon Tatu l pe e
2. 62 00 NE EK 3854 5973 2793 26942 4368 1476 4678 6575 1134 2373 4837 2134 6172 7357 4485 6079 6523 3243 7423 Lel Table B16 4 DOF Steadiness Test Session 2 N ON 95 4 900 1 NN NK 28 SU Steadinesss Unit 15 5 4 DOF Steadiness SUs 392 642 514 Non dominant Side 7153 3920 5271 57874 3447 920 761 4171 1007 636 7860 7815 1579 1849 8458 5918 3841 2797 1379 3992 2605 2544 13315 15368 21051 5710 5144 NO F 53 R 70 3666 3199 6677 11226 6502 APPENDIX C INSTITUTIONAL REVIEW BOARD DOCUMENTS 138 THE UNIVERSITY OF TEXAS AT ARLINGTON OFFICE OF RESEARCH COMPLIANCE Date July 11 2005 To Dr George Kondraske Electrical Engineering 19180 Re Protocol 05 395 Investigation of a Portable Performance Measurement System for Neurological Screening in Clinics Part 1 The Institutional Review Board IRB has reviewed and approved this research protocol under an expedited review in accordance with Title 45 CFR 46 110 Continuing review 1s scheduled for one year from the above approval date The Office for Human Research Protections OHRP requires you to submit annual and final reports for review and approval by the IRB The annual report must be on file with the IRB before the anniversary date of your initial approval O
3. 50 49 41 41 254 219 230 295 82 101 142 145_ m R 51 49 41 43 352 254 259 346 118 81 H1 51 27 57 49 49 47 296 205 357 338 121 95 128 5 eo 55 51 46 257 258 302 286 139 153 63 52 PB EN EN ss 52 Pp M M E M M M M E E LE EZ F EG M 5 NE MNT _ WM LT T eo 7 20050 1 28 _ R 25 R 59 52 24 R 58 55 ERES ee ee _ 26 R 4 gt OR L _ 24 R _5 R _2 R 24 R R EU Table B5 Upper Extremity Coordination NMCC Test Results Test Session 1 Number Sex Age Movement Speed cm s Accuracy 906 NMCC bits s F 24 M 28 M 27 BB 952 LR R 3 amp 1 4449 414 3657 3840 3200 3048 4li4 LR 5334 4522 4124 4419 M W 96 M Lp 2 R 4522 4826 3505 3200 830 7 M s 01 R a 3048 3657 85 8 Ea M 2 1 3962 3200 4724 900 __ M 26 2895 2590 3352 3352 10 M 2 R 6096 4267 4572 3962 F 24 5334 4572 4114 3962 _ 12 F 24 4114 3200 4267 900 13 F 26 R 350 4114 2743 2590 M 2 R 4267 3962 4114 3810 71 15 F
4. 60886 48973 60 886 63 909 68 787 23 R 76366 73 261 75093 76 366 79 045 81 181 23 R 70 954 68 787 70954 70400 66258 62 146 R 53638 40774 48 188 51200 45 975 41 912 25 R 6004 62 146 70400 65298 70400 64 365 25 R 40408 52 697 50341 53 638 50 910 48 709 26 R 77 018 68266 69 316 69 854 69 316 67 753 25 R 60074 66 749 67 753 61 300 60 886 62 577 24 R 68787 67247 67 753 72 089 66258 68 266 24 R 79045 77 018 65298 67247 69 316 70 954 26 R 69316 55283 50910 62 146 66 258 68 266 25 R 77 682 74472 73261 63459 54284 65 775 25 R 77 018 82 671 77018 77018 84216 81 920 24 R f 81920 85011 85 011 78 358 81481 88 345 25 R 80457 80457 82 671 80457 75724 80457 24 R 69 316 73 862 75 724 73 261 67 753 70 400 24 R 75093 70954 75093 63 909 70400 68 787 F 25 R 73261 68 787 66 749 73 862 82 671 76 366 s 79 eel Table B12 Hand Arm Steadiness Test Results Translational Steadiness Y Axis Test Session 2 Translational Acceleration Based Steadiness Y axis 1 g 1 24 R 88345 100 124 95 863 95 863 100 124 90 112 _ 1 345 487 866 866 863 345 866 4 1 1 9 1 6 AT 820 4 5 5 6
5. Figure 4 15 Comparison of Steadiness Values from present and previous HPMM versions 101 Translational Steadiness X Axis 1 0 Test Session 2 D Side Side 3 AU 50 70 50 Test Session 1 Figure 4 16 Comparison of sessions 1 and 2 for translational steadiness X axis Translational Steadiness Y axis 1 g c 2 Uu o E aD D Side N Side 80 Test Session 1 Figure 4 17 Comparison of sessions 1 and 2 for translational steadiness Y axis 102 As noted the rotational rate based steadiness measures are new in this version of the HPMM When mounted on the hand as was done in the current study the motions of wrist flexion extension rotation about the RSM x axis and forearm pronation supination rotation about the RSM axis are measured These measurements display higher test retest reliability than the translational steadiness measurements gt 0 8 These values are again surprisingly high given that we looking at only one DOF at a time and multiple DOFs are present suggesting that subjects tend to achieve a given amount of steadiness the same way upon repeat maximal performance testing That is subjects who have little rotational wrist flexion extension motion and high translational motion in a vertical direction tend to always exhibit this pattern While no data exists in the HPMM context for comparison of val
6. Mean 9 changel 33 9 11 5 17 0 20 4 16 1 17 4 9 3 17 1 18 4 26 5 39 7 93 7 Measure units Isometric Grip Strength N 319 0 329 6 108 9 Finger Tapping Performance 5 06 Speed taps s 5 06 35 66 7 l Finger Tapping Performance Duty Cycle Mean Finger Tapping Performance Duty Cycle SD 96 Upper Extremity Coordination 10 50 NMCC bits s Upper Extremity Coordination 128 Speed cm s Upper Extremity Coordination 82 90 Accuracy 96 Hand Arm Steadiness Translational Steadiness a 1 g ed Hand Arm Steadiness Translational Steadiness a y 1 g bus d Hand Arm Steadiness 1 2 192 Rotational Steadiness s deg Hand Arm Steadiness 32 Rotational Steadiness c s deg Hand Arm Steadiness 4 DOF Steadiness ax SUs IRA Primary measure Secondary measure Exploratory measure 11 50 2 69 91 92 59 14 76 m pes pepe 6655 9141 105 6 Eras 3 43 7 8 A A 0 8 16 13 6 9 EIN EE EN EN De EN EN E ES ES 4 4 Reliability and Validity of HPMM Measurements There are two aspects to the evaluation of measurements obtained by using the current HPMM prototype their reliability and their validity For each test these aspects of the corresponding measures are discussed in that order With regard to repeatability and the way in which it is classically measured it should be mentioned here that members of the
7. as mentioned in the previous section A large value of SC decreases power consumption but will result in a slower response For this reason the present version of the HPMM will use a low value for SC 2 Drift Compensation Signal drift occurs with changes in C and Vu electrode contaminations and aging effects Compensation is achieved by causing the signal reference level to track the raw signal while no detection is in effect This change should be slow and device slew rate limited Positive drift compensation PDC 0 255 should be set to a large number to compensate for the slow increase In C as an object comes close to the electrode Setting it too fast will cause compensation to occur even before the touch But NDC 0 255 must be set to a smaller value to compensate quickly for the removal of touch or an object or after a MOD recalibration discussed later This situation applies to the HPMM in tests such as the finger tapping performance and upper extremity coordination where the output od the sensor must not be active until the finger actually touches the electrode and the reference level must quickly be reached by the internal signal once the finger 1s removed 41 3 Threshold THR 0 255 Threshold is measured in counts of signal deviation from the reference level If the signal equals or exceeds the threshold value detection can occur The detection will end only when the signal goes below the hysteresis level or u
8. R 12 r 1 4 r 125 48 2 lt lt lt lt lt lt 20 Table B10 Hand Arm Steadiness Test Results Rotational Steadiness Z Axis Test Session 1 Sax Rotational Rate Based Steadiness Z axis s deg 3 R 06 06 on 9n R ow o5 10 06 96 x R or on 06 R 1 2 14 12 16 143 R 042 043 o3 05 056 05 R oso 050 034 032 043 3 R or 077 o7 083 R om 05 0 029 029 040 R 10 09 os os 033 R os 034 05 040 038 038 3 R 10 o7 0 05 083 036 R 0 ow or on os R 06 99 067 05 050 056 R 10 14 09 osi R 16 15 14 14 143 R on 06 06 056 063 R os 091 os 091 091 R o6 091 10 091 R on 06 o7 056 04 06 Cr 23 R oor em os 05 on 083 s 79 Table B11 Hand Arm Steadiness Test Results Translational Steadiness X Axis Test Session 2 Gay Translational Acceleration Based Steadiness X axis 1 g __24 R 71 517 75 093 68266 69 854 71 517 69 854 28 R 67 753 61300 67247 60 886 60 886 65 775 27 R 53959
9. This allows the HPMM to run in the all systems running condition for over 9 5 hours Considering that there 1s never a situation when all subsystems are simultaneously running it Is very likely that an 8 hour minimum operational time between charges can be achieved 3 2 Software Implementation 3 2 1 Operating System The HPMM operating system is intended to provide a basic set of functions that supports its operation These include user interface management menu generation and display monitoring of user inputs and port status calibration of sensors where applicable power management and other yet to be implemented maintenance functions 57 User Interface Management Two separate sections of subroutines called the LCD Module and the Touch Screen Module contain all routines required to manage these components The touch screen routines implement functionality such as using the SPI communications to initialize the touch controller acquire data from the touch screen controller scan for touch and map the coordinates of the touched region to the four button overlay esc enter up and down The LCD Module consists of routines that initialize the T6963C LCD controller and subroutines that build pages for each individual screen that is displayed to the user going from the menu screens to the test result pages The communication between the processor and LCD involves sending a set of specific code words that are organized as command an
10. without having to refer the patient to a specialized laboratory The current study seeks to evaluate if the instrument provides measurements that are comparable to those obtained with other laboratory based instruments PROCEDURES If I agree to participate in this study I will be involved in one testing session lasting approximately one hour Tests will be performed at the Human Performance Institute at the University of Texas at Arlington I will undergo a series of performance capacity tests in order to determine selected performance resource capacities These tests involve simple motions of the fingers and arms in response to visual and auditory signals Grip strength will also be tested The series of tests will be repeated twice during the session RISKS THAT MAY OCCUR DURING THE STUDY There are few potential risks involved in this study During the testing I may experience some fatigue or possibly some confusion The possible risks involved in this study are relatively slight The processes and protocols described in the proposal are the only way in which the necessary information can be dais 25 ner IF APPROVED BY THE UTA The IRB approval for this consent document will expire on l JUL 1 0 2006 Figure C2 Participation Explanation and Consent Form Page 1 140 BENEFITS FOR YOUR PARTICIPATION By participating in this study I will be assisting in the evaluation of a new tool
11. 10 2 99 112 Kondraske G V 1982 Design Construction and Evaluation of an Automated Computer Based System for Quantification of Neurologic Function Doctoral Dissertation Arlington TX The University of Texas at Arlington 144 Kondraske G V 1984 A capacitive displacement transducer for tremor measurement Proceedings Sixth Annual IEEE Engineering in Medicine and Biology Society Conference pp 432 435 Kondraske G V Potvin A R Tourtellotte W W amp Syndulko K 1984 A computer based system for automated quantification of neurologic function IEEE Trans Biomed Eng 31 5 401 414 Kondraske G V 1990 A PC based performance measurement laboratory system J of Clin Engr 15 6 467 478 Kondraske G V 1992 Palmtop Size Human Performance Multi Meter HPMM for Crew Monitoring NASA SBIR Proposal Kondraske G V 1995 A Working Model for Human System Task Interfaces In Bronzino J Ed The Biomedical Engineering Handbook pp 2147 2164 Boca Raton FL CRC Press Kondraske G V 1999 Determination of Fitts index of performance using constructs of General Systems Performance Theory and implications for motor control performance capacity measurement in Abstracts 17th Annual Houston Conference on Biomedical Engineering Research A M Sherwood Ed Houston pp 86 145 Kondraske G V 2000a Performance theory implications for performance measurement tas
12. 25 R 320 3200 3657 3200 95 16 F 24 R 5029 5638 4572 4419 F 25 R 3810 3505 3352 3048 96 95_ 18 F 24 5791 4267 3048 35 05 63 89 90 19 M 24 R 2895 2895 2895 35 05 94 89 78 20 F 25 R 3962 4267 3810 3352 92 82 76 Fes 74 78 60 LES 71160 69 89 86 68 gt Ne LCI Table B6 Upper Extremity Coordination NMCC Test Results Test Session 2 ue gt L r __ gt M 2 R 42 67 4419 32 00 2743 75 75 80 83 1091 1130 872 776 M 27 R 41 41 33 52 2895 3657 85 72 94 79 1192 823 927 962 4 M 2 R 45 72 4419 35 05 3048 83 93 86 90 12 93 1401 1027 935 __5 23 R 50 29 42 67 3657 2743 82 79 83 1440 1193 962 776 6 F 2 R 147241 4572 3810 3200 70 83 72 71 1127 12 93 935 774 __7 R 4419 4114 3810 2438 72 77 68 87 1085 10 80 883 723 M 2 R 154861 6248 4724 4419 91 85 77 86 17 02 1811 1240 1295_ 9 M 2 R 414 42 67 3657 3840 74 85 79 76 1038 1236 962 965 _0 M 2 R J4LI4 4724 44 19 3810 81 87 62 64 1136 1401 934 8 31 1 F 24 R 47241 4724 33 52 3810 87 93 81 92 1401 1475 925 1195_ 2 F 24 R 1411413810
13. 3352 3352 85 84 68 68 1192 1091 777 7 77 13 F 26 R 4267 25 90 3657 3657 78 94 70 83 1134 830 850 1034 14 M 25 R J4724 4876 39 62 3810 90 84 80 14 49 11 80 1134 10 39 15 25 R 4724 33 52 3657 33 52 80 95 62 95 1288 1086 773 1046 16 F 24 R 5456 5638 4419 3810 86 78 79 76 1608 14 99 1190 965 17 F 25 R 5505 3962 2895 3657 91 96 78 75 1087 1274 769 935 18 F 24 R 151811 5638 3657 4572 85 81 75 63 15 01 1557 935 982 19 M 24 R 2895 3048 25 90 2590 94 95 82 82 927 987 724 724 20 F 25 R 138 10 4114 35 05 39 62 88 85 86 76 1120 11 92 1027 10 26 Mai 5 03 Sci Table B7 Hand Arm Steadiness Test Results Translational Steadiness X Axis Test Session 1 Translational Acceleration Based Steadiness X axis 1 g 3 R 55970 56320 63459 60974 64365 R 69316 72670 70400 63 909 64365 x R 58514 48709 48709 51492 53007 R 7945 77018 80457 85820 3 R 59676 59676 61720 61220 64828 23 R 4050 41447 35617 37862 49774 R ana 48447 49241 54946 51200 R 40 960 46933 47479 51492 41 678 26 R 8501 siis 69316 67247 79045 R 58136
14. 88 345 91 022 94854 25 R 72670 70954 61 300 68 787 64 365 73 261 24 R 76366 72670 79045 79 745 79 745 75 093 _ __24 R 85 457 8l 26 R 95 863 23 345 25 __ R x 25 24 R 4 1 R x 25 4 528 5 2 954 5 6 670 5 5 5 R 88345 93 866 106 014 77 018 96 894 92 898 568 4 9 697 0 3 015 5 9 3 6 1 5 0 NENNEN ss L X 8 r 20 NO Table B9 Hand Arm Steadiness Test Results Rotational Steadiness X Axis Test Session 1 R 03 o 1 19 319 _ R 12 1 4 1 4 34 LM 2 R iu iu 09 10 19 _ M 23 R 1 4 125 167 20 29 29 5 M 3 R s an 125 35 rer om on o 99 L 2 re iz in is 16 125 16 3 M oo on 08 96 o 083 9 M 2 R 14 18 12 14 125 155 M 2 R 08 os oe os os 9n ou F a R t os L 99 Co F a R 10 s 1 4 14 L tu 3 Fr S Rr 10 oe 98 om e 9n R 90 on 15 os 10 1m R 1 2 18 18 125 15 1H Ca R 125 125 14 0 R ia 16 16 16 20 20 R 07 os on os 10 10 x R 1 4 14 18 LU 10 10 _
15. Database results e Suitable for office bedside field use Recommended for Clinic Use Software OPTIONAL Host based Mode HPMM driven by user specified test protocol i e predefined sequence of tests Tests now specific rather than generic e g shoulder abductor strength elbow flexor speed etc Results for a given subject session can be uploaded to Host PC for databasing reporting User Interface Charging LCD Prompts and Menus Unit e Pushbutton Navigation Test Test Subject Administrator HPMM User s Manual Figure 2 1 Major Features of the Overall HPMM System Concept 12 In the Generic Test Mode the unit is intended to be used as a stand alone general purpose device that is capable of performing different generic tests Much like a digital multi meter which can measure generic electrical quantities such as voltage and resistance and may specific voltages and currents can be measured in the Generic Test Mode the HPMM can be used to measure different generic performance capacities for example strength angular movement speed etc Similarly these could conceivably be applied to a number of different body subsystems for which the generic performance capacity is relevant For example strength capacity can be of interest for many different neuromuscular subsystems Each test can be selected from the menu of such tests executed and the results are displayed on the shown on t
16. HPMM will be disconnected and taken to the field site for operation Once the testing is done the unit can be brought back to upload the results to the host PC This thesis does not address any of the functionality associated with host based operation During development of versions 1 and 2 communication protocols were defined and preliminary versions of software were developed to explore the protocol driven mode These functions have been given lower priority in favor of a focus on the fundamental performance capacity measurement capabilities of the HPMM 2 3 HPMM Functionality and Architecture Throughout the course of HPMM development analyses were undertaken to determine the requirements for different subsystems or functional units that should be included in the HPMM architecture and how these subsystems should be included This is driven by consideration of the possible measurement functionality that could be included as well as the ability of certain functional units to support implementation of more than one performance capacity test A current result of this analysis and the status of each functional unit In the version 4 prototype 1s shown in table 2 1 As discussed in section 2 2 some of these basic functional units are located in the physically separate RSM unit 15 Table 2 1 Basic functional unit candidates identified for eventual incorporation into the HPMM and status in the version 4 platform Functional Units Requiremen
17. India in the year 2003 and his Master of Science degree in Electrical Engineering from the University of Texas at Arlington in the year 2005 149
18. a higher count for the detect integrator than for the end of detect integrator This will ensure that the start of detection 1s not caused by sporadic noise or an inadvertent touch or brush by the examiner which is necessary in all touch sensor based performance capacity tests A low setting for the end of detect integrator will allow a fast response during tests such as simple response speed where it 1s vital that there be no detection delay after the subject has removed their finger from the sensor Detection delays can be long with large burst spacing so the burst rate can be increased when the DIA or DIB counters are operating The bits DISA and DISB respectively enable this fast detection and the normal burst rate resumes after the DIB counter stops counting at the end of detection 6 Max On Duration MOD MOD 0 255 allows for automatic recalibration if the activation last longer than the designated timeout This can be caused by a stray object inadvertently coming in close proximity with the sense electrode If SC gt 0 the delay Tmoa 15 MOD 1 x l6x Ths 43 If SC 0 the delay 1s a function of the burst duration Tua Tmoa MOD 1 x 256 x The MOD function is disabled by setting MOD 255 so that the sensor recalibrates only when part is powered down and started up again This is suitable In case of the HPMM Primarily it is preferable that the output deactivate only when the finger 15 removed from the
19. at 81 7 272 4840 or the Office of Research Compliance at 817 272 3723 Sincerely Dr Jennifer Gray Assistant Professor IRB Chair BOX 19188 202 E BORDER SUITE 201 ARLINGTON TEXAS 76019 0188 T 817 272 3723 817 303 9187 Figure C1 Institutional Review Board Approval Letter 139 e Subject U TA IRB Protocol 7 S gt amp THE UNIVERSITY OF TEXAS 24 AT ARLINGTON PARTICIPATION EXPLANATION AND CONSENT FORM PROJECT TITLE Investigation of a Portable Performance Measurement System for Neurological Screening in Clinics Part INVESTIGATORS George V Kondraske Ph D Professor Electrical and Biomedical Engineering Rahul Mulukutla Graduate Student EE Department TELEPHONE NUMBER 817 272 3454 817 272 2535 BACKGROUND INFORMATION I have been asked to participate in a research study that will investigate the validity and repeatability of measures obtained using prototypes of components of a portable performance measurement instrument This tool is called the Human Performance Multi Meter HPMM and the investigation pertains to its latest version This instrument may ultimately combine several instruments that are currently only available in large laboratory settings into a single small unit It has the potential to allow medical practitioners e g neurologists and others to rapidly assess many aspects of human performance including strength speed of movements and tremor measurements
20. coordination resources This test 15 sometimes referred to as the finger tapping performance test GTA 5 Upper The Neuromotor Channel Capacity Test measures the Extremity availability of a performance resource that 1s related to the Neuromotor speed and accuracy of movement and their tradeoff It 1s Channel essentially a measure of coordination Capacity Test GTA 6 Steadiness The Steadiness Test 1s a generic test of the steadiness of a Test particular body segment This will most frequently be applied to a subject s hand or hand arm combination In the generic test administration sense this could also be performed on other body segments e g head or lower extremities 63 3 2 3 GTA I Isometric Strength Test A static or isometric strength test is a test to measure the capacity of muscles to produce force in such way that there 1s no substantial extension of the muscles and no movement occurs It measures the capacity of muscle groups to develop tension There are two different sub modes for testing grip and resistance but the same algorithm 1s employed The different modes pertain to different procedures in which the force sensor subsystem is employed in its interaction with the subject under test Upon selecting the Isometric Strength Test option from the menu the GTAISTART routine 1s called The appropriate configuration settings for the power management and data acquisition subsystems are a part of the initializat
21. for a selected subset of performance capacity tests that are emphasized in this thesis 11 2 2 System Concept In version 4 the HPMM system consists of the main HPMM unit the Remote Sensor Module RSM and the host PC Figure 2 1 The main unit contains the LCD and touch screen as user interface Menus can be navigated and options selected by the use of four touch sensitive buttons on the touch screen Selected basic sensors including the force sensor and touch sensors array are integrated into the main unit Rechargeable batteries power this main unit with an on board charging unit An RS 232 serial port supports communication between the HPMM and other devices such as a host computer The overall system concept includes two different test modes 1 the Generic Test Mode and 2 The Protocol Driven Mode Host PC with HPMM Windows based Remote Sensor Module HPMM Main Unit RSM xx e One or more sensors Supports Stand alone Menu Driven and Host based Supports one or more modes tests Battery powered rechargeable full day charge Optimized for Ergonomic Build test protocols Issues Test Subject Stand alone Mode Serial Download protocols Interface Suite of generic performance capacity tests of different la to HPMM Different RSMs types e g strength speed etc Data Link Upload test results envisioned Each test is short duration with results displayed on RS 232 Produce reports LCD
22. integrates the functionality of laboratory based performance capacity measurements instruments into a handheld portable model that 1s also compact accurate and relatively low cost Making the analogy in terms of concept portability and general purpose utility to the ubiquitous digital multi meter the Human Performance MultiMeter HPMM represents an effort in this direction The HPMM was first conceptualized in 1992 Kondraske 1992 Areas that would benefit from such a device include neurology emergency medicine 1 e status screening sports medicine rehabilitation battlefield medicine space medicine gerontology field sobriety testing toxicological screening and industrial medicine e g alcohol drug abuse A series of sequential developments and investigations were undertaken which are detailed in other documents Kondraske 2005 Early work involved the assessment of the types of measurements that were feasible and what their requirements were as well as the kind of trade offs that might be required given the system s compactness and portability Short term memory visual and auditory information processing speed neuromotor channel capacity in manual speed accuracy tradeoff tasks visuomotor coordination speech motor control isometric strength vibratory sensation steadiness and speech motor control fall within the measurement candidates considered for this instrument For a subset of this functionality individual subsystem
23. now provides the basis for careful scrutiny of the design production of supporting documentation and implementation of operating system and generic test algorithms on the new platform 3 Hardware Implementation A block diagram of version 4 was introduced in Chapter 2 figure 2 1 Major subsystems are considered here in more detail to provide a basis for understanding the implementation of test algorithms It is important to note that the HPMM 15 conceived of an expandable instrument It is not simply that a well defined set of functions exists and that hardware can be designed around these requirements Rather the goal is to 3l realize a powerful flexible hardware platform that allows for implementation of currently defined functionality as well as new functionality as it becomes defined In this context general features of the hardware platform are discussed as well as specific aspects pertaining to currently implemented functionality 3 1 1 Microcontroller Core The new HPMM 4 microcontroller core is the Cygnal now Silicon Laboratories C8051F020 This processor has an impressive collection of built in peripherals and is a high speed pipelined 8051 compatible microcontroller up to 25 MIPS with a clock speed of 25M HZ It contains a 12 bit IO0ksps 8 channel ADC with on chip Temperature sensor and an 8 bit 500ksps 8 channel ADC with analog multiplexer On chip memory includes 64kB of programmable FLASH memory and 4096 256 bytes
24. of on chip RAM Five general purpose 16 bit timers an on chip Watchdog Timer and Voltage Monitor and an on chip oscillator complete the package The Code Memory consists of the 64 kilobyte FLASH memory which can be rewritten repeatedly This can be accomplished through code using the MOVX command after certain Special Function Registers have been set and may also be programmed via a Joint Test Action Group JTAG IEEE 1149 1 interface The 100 pin processor package option used in HPMM version 4 contains four additional ports over the 64 pin package and was selected to meet system requirements To configure the processor I O lines it is necessary to use the processor s digital I O crossbar feature Thus the I O pins as defined within the HPMM context are valid only when the crossbar is properly configured and enabled Configuration is performed 32 immediately after power up as part of initialization and not altered during the course of system operation Po crossbar Register Bits 0 1 2 3 4 5 T TAO UARTOEN ABRO 2 RXD ____ mE Figure 3 1 Digital Crossbar Pin Assignments for the Serial Port and SPI Interfaces The two peripherals shown in the above table are automatically assigned to their port pins by the digital crossbar This 1s achieved by setting bits 1 and 2 in the XBRO register The remaining port pins on the microcontroller will fall under the category of general purpose I O pins GPIO
25. procedures and software algorithms for the new hardware platform new development for selected aspects of the system and rigorous experimental evaluation of a selected set of HPMM s performance capacity measures in human subjects 1 1 Human Performance Capacity Measurement Research Background Over 25 years of development effort encompassing 400 different measurements acquired through 20 different continuously evolving instruments has resulted in the Human Performance Capacity Measurement System HPCMS Kondraske 1990 Human Performance Measurement Inc of Arlington Texas provides modular instrument packages representing these instruments The philosophy associated with this technology is such that this laboratory based Human Performance Capacity Measurement System HPCMS may be viewed as a flexible set of items procedures modules et cetera that can be combined in various ways to realize a wide variety of different application specific human performance capacity measurement systems Human Performance Measurement Inc 2004 To provide perspective a brief review of this development history is warranted Potvin Tourtellotte Syndulko and colleagues dating from the late 1960s were pioneers of work that raised the need for quantification of what they termed neurologic function Potvin et al 1985 They investigated and established many basic methods and the first subset of devices for a neuro function laboratory and addressed key i
26. subject pool are all young and healthy resulting in the exercising of these measurements over a rather narrow segment of the range over which they are intended to measure However this population represents that which could be considered the most stable 1 repeat performance is not influenced by medication metabolism disease processes etc Nonetheless actual performance in maximal performance tests does vary even in this population Thus reliability measurements are performed under worst case conditions This issue has been discussed previously Sriwatapongse 2002 Mayer et al 1997 and is reiterated here due to its importance Because the subjects are all young and healthy the range of their performance for many of the measures is relatively small and the variation within a subject is a large fraction of the variation within a population This situation makes it more difficult to obtain high reliability coefficients In cases where the distribution of a measure covers a wide range such as grip strength it is generally easier to obtain higher reliability coefficients To investigate validity the agreement between these results and those obtained with earlier versions of the HPMM as well as those obtained with laboratory based 90 instruments must be checked In addition the presence of expected patterns 1s also illustrative of validity e g better performance on the dominant side when anticipated Perfect agreement is
27. subsystems as in gripping A wide range of instruments and approaches have been used for strength measurement Smith 2000 Beasley Beasley 1961 pioneered the measurement of isometric strength and provided numerous insights into reliability and validity of such tests The approach of interest is known as held held dynamometry Edwards and Hyde 1977 and involves incorporating a force sensor data acquisition and processing 25 algorithms into a paradigm commonly used by physical therapists called manual muscle testing In manual muscle testing the therapist uses his her hands to push against or resist motions that would be produced by an isolated muscle group They then must subjectively estimate the strength of that muscle group Hand held dynamometry attempts to make this process more objective and accurate with greater measurement resolution Kondraske and colleagues were among the first to computer automate the measurement process Kondraske et al 1984 and have more than 20 years of experience with the design of optimized sensors and processing algorithms within the context of the HPCMS A recent review of hand held dynamometry Kolber and Cleland 2005 concluded that this approach is generally reliable and a useful improvement to manual muscle testing Version 3 of the HPMM Sriwatanapongse 2002 utilized the grip strength sensor currently used in the HPCMS This is modified in version 4 to better integrate the sensor into the latest
28. that may benefit many others in the future It is expected to lay the groundwork for future work that may prove to be valuable for diagnosis and evaluation of the effectiveness of new drugs and therapies I will also learn about some aspects of human performance and performance measurement AVAILIBILITY OF COMPENSATION AND MEDICAL TREATMENT FOR PHYSICAL INJURY The investigators will make every effort to prevent physical injury that could result from this research If I am injured the research protocol does not require the payment of financial compensation to me from the investigator or the University of Texas at Arlington Medical treatment for physical injuries is not available from the researchers as part of the research protocol The University of Texas at Arlington Heath Services will provide medical treatment should an acute condition arise from my participation in this study I will be financially responsible for any emergency medical care I receive CONFIDENTIALITY I have the right to privacy and all information that is obtained in connection with this study and that can be identified with me will remain confidential as far as possible within state and federal law Everything the investigators learn about me in the study will be confidential The results of this study may be published in the medical literature or for teaching purposes no names will be used No photographs audio or videotapes will be used Records will be kept rega
29. this check 1s passed the algorithm begins to sample and store the sensor states at a rate of 100 Hz Any change in sensor state will cause a 3 second timer to start Should there be no further change before 3 seconds elapse the timer will timeout and a false start will be detected The test 1s then restarted The first sample that shows a change in sensor state determines which sensor was touched although it may so happen that during the touch which can last up to 50 60 ms the finger may also cover another sensor For example if the subject touches the error region in the first sample of the touch but eventually also covers the target region the finger Is on the border of the target error regions the result for that touch is still counted as an error A double beep signals the end of the test The equations for the speed and accuracy computations are the same as those presented in Sriwatanapongse 2002 One may note that the unit used for speed during the computation of neuromotor channel capacity is cycles s although the reported unit is cm s in the results section A slight modification has been made for the computation of NMCC The equation is NMCC Speed x Accuracy x logo A W 1 bits s The dimensions of the target width and movement amplitude target separation have not been changed in this version of the HPMM so that the same calculations for index of task difficulty also still apply The appropriate compromise between speed
30. touch screen to provide the UP DOWN ESC and ENTER buttons used for navigation and command entry Each button is 26mm x 20 mm The 35 upper portion of the screen thus remains available for use A single chip touch screen controller TSC2200 by Texas Instruments interfaces directly to the four wire touch screen and uses the SPI bus to communicate with the processor This controller includes advanced features that enable power saving and simplify coding 3 1 3 Optional Additional Memory Provision has been made in the HPMM main unit for the use of an external memory chip The chip will be a 256 KB 32K x 8 static RAM in a 28 pin surface mount package An example candidate is the M48Z35AV from ST microelectronics The processor can access this SRAM IC through the External Memory Interface EMIF which will be enabled on the lower four ports PA P7 using the digital crossbar see Table 3 1 It 1s proposed that due to the availability of the large number of I O pins no address data multiplexing will be required and so the device will be used in non multiplexed mode This external memory will work as an extra data memory for data intensive applications Currently code memory is also used to store digitized samples in cases where the number of samples is very large such as in the steadiness test see below 3 1 4 Programming and Updating HPMM Program Code There exist two options for programming or updating HPMM c
31. 12 bit ADC the measurement range for the A C coupled channel is 0 409 6 deg s For the D C coupled channel the overall sensitivity is 0 909 ADU deg s which gives a theoretical measurement range of 0 4 506 deg s Speech Measurement Subsystem The RSM 1 unit also contains a microphone and preamplifier for future use in a series of performance capacity tests that involve speech The output of the speech circuitry on the RSM which 1 in the form of a raw speech signal is connected the analog input channel on the HPMM microcontroller Thus this raw speech signal can be sampled by the ADCO system that was described earlier This subsystem is not presently used for any tests and therefore is not detailed in this thesis Specifications e g gain and frequency response have not been finalized RSM System Interconnections The arrangement of the functional blocks in the RSM and their interconnections is as shown in figure 3 3 Although the filters and amplifiers are shown as dual channel blocks for clarity there are in fact one amplifier and one filter per channel and are implemented using operational amplifiers The ADC communicates with the HPMM microcontroller through the Serial Peripheral Interface SPI This is a four wire 54 interface that enables efficient full duplex serial communication in a master slave configuration The SPI interface is enabled on port 0 of the HPMM microcontroller after proper configuration of the m
32. 2 4 18 Rotational Rate Based Steadiness X axis comparison socie u riae hians 104 IX 4 19 Rotational Rate Based Steadiness Z axis comparison between SeSSIons 4 20 4 DOF Steadiness comparison between the two sessions Table 1 1 Zal 2 2 Jal 3 2 3 3 3 4 3 5 3 6 3 7 4 1 4 2 LIST OF TABLES Page Summary of HPMM Development Mllestones 9 Basic functional unit candidates identified for eventual incorporation into the HPMM and status in the version 4 platform 16 Matrix showing the assignment of different HPMM functional units to different HPMM functions and those generic tests included da on einen 17 HPMM Port Pins and Descriptions 34 e UP uuu u nunus h nam dde ate 47 OT S20 Parameter SONES hump 48 Contribution of Various Subsystems to Power Consumption UES MEM 57 Description of Generic Test Aleortthms 63 Estimates of rotational rate based on displacement estimates 73 sobiware Memory NI cai sau 80 Summary of Experim
33. 2 Number Gender Age D Side Grip Strength N Grip Strength N D N D N _ __ F 24 R 21 218 217 188 2 M 28 R 287 287 330 25 _ 3 M 27 R 556 58 5155 50 pos p __5 M 2 R 47 364 386 35 243 __ 7 M 2 R 38 35 30 25 25 R 316 300 336 278 26 R 41 464 48 520 _ l M 25 R 31 337 457 dH ln F 2 339 36 34 338 24 R 23 26 204 189 26 R 262 242 332 30 14 M 25 R 469 590 55 56 15 F 25 R 230 218 268 255 16 F 2 R 192 160 21 20 25 R 230 20 267 229 18 F 24 R 259 236 227 X2 19 M 24 R 35 36 49 39 20 F 25 R 288 266 304 200 EE lt 92 lt 8 M 9 M 10 E u T 1 cel Number EE HN 8 j E s 20 N Table B2 Simple Visual Response Speed Test Results Gender Vcl Table B3 Finger Tapping Performance Test Results Test Session 1 Number sex Age D Side pepper pero er ss pas so aa es ss 104 128 147 00 mL 12 8 sss as ae 3u 93 nee R s3 54 44 46 249 315 269 71 148 100 Ls pw ar o 9
34. 4 10 Comparison of results from two sessions of the visual response speed test 4 4 4 Finger Tapping Performance The generic test involved here is termed rapid alternating movement performance which was applied here to focus on finger tapping with motion restricted to flexion and extension about the metacarpophalangeal joint of the index finger The primary measure is the tapping speed taps s which has been used extensively by many researchers Potvin et al 1985 Tap duty cycle mean 96 is a secondary measure the standard deviation of the tap duty cycle 96 1s an exploratory measure While more importance is currently attached to the primary measure the other two are also evaluated 95 The tapping speed results display good inter session repeatability r 2 0 57 and 0 83 respectively for dominant and non dominant sides Also there is a very good agreement between the mean values of tapping speed obtained by the present and previous versions of the HPMM in which similar populations were involved Tapping Speed taps s 7 o Test Session 1 Figure 4 11 Comparison of primary measure results from two sessions of the finger tapping performance test High test retest reliability is demonstrated by the tap duty cycle mean measurements but low r lt 0 3 values are shown by the tap duty cycle standard deviation One possible reason is that the subject pool consists only of health
35. 58436 52087 52087 55624 R 63459 62446 55 970 59284 65715 __ R 6257 62446 62146 67247 61300 _ 26 R 69316 60074 68266 65275 61753 R 75724 09 854 75093 58 896 68266 R sia 52697 59284 62146 63015 21 R 68266 772682 77682 74472 73261 R 43957 39350 42306 46211 46690 3 R 68787 5463 6257 70954 73362 R 64365 64828 80457 63909 55896 F 2 R 5090 7447 69316 66258 69316 rH 68 266 66 258 57 032 82 671 67 247 53 638 54 946 46 933 72 089 56 320 62 577 67 247 66 258 65 775 61 720 70 954 48 973 73 261 63 909 63 459 Table B8 Hand Arm Steadiness Test Results Translational Steadiness Y Axis Test Session 1 Translational Acceleration Based Steadiness Y axis 1 g eee 24 R j 67753 70400 82 671 78 358 84216 79 045 28 R 87487 92898 95 863 81 181 84216 88345 27 R 79745 64828 61300 61 720 77018 78 358 23 R 109892 89219 114 065 114 065 115 528 114 065 23 R 89219 80457 86 646 85 820 93 866 88 345 23 53 959 59 676 45975 53 007 52 087 72 670 25 R 57 032 58514 64365 71517 72670 71 517 25 R 68266 64828 59284 55283 62 146 58 136 26 R 111249 115 528 85 820
36. 7 s ms 39 8s 35 6 F 2 R 44 43 38 39 488 546 426 474 100 76 L 151 5 E 12 2 8 4 W N DOL oll 01 10 0 9 M 2 R 59 59 53 52 329 303 307 305 13 6 136 0 M 2 R 57 55 53 51 294 293 344 375 145 28 129 Ho F 2 R 59 56 54 150 394 448 512 393 121 85 84 a F R 52 54 as 382 49 427 122 90 152 13 F 26 R 45 47 44 42 330 345 294 529 85 125 131 13 14 M 25 R 56 553 58 54 396 353 399 343 160 88 153 133 _ 15 F 2 R 50 5 0 44 43 348 333 373 395 140 145 145 13 9 16 F 2 R 50 53 49 50 215 323 358 391 100 137 140 108 __17 F 2 R 47 43 47 43 341 369 425 410 150 22 117 140 18 F 2 R 51 52 47 48 246 312 278 300 58 123 63 148 19 R 56 56 53 50 367 433 441 407 106 89 72 103 20 F 2 R 50 51 47 47 386 405 314 423 106 104 158 c 776 2 R 67 67 62 58 377 368 347 425 79 122 N 95 ola Scl Table B4 Finger Tapping Performance Test Results Test Session 2 Tapping Speed Duty Cycle Mean Duty Cycle SD 4 LL __ Fl wm
37. 70 8 R 87487 66 258 68266 63 909 70400 84216 93 866 93 866 94 854 80457 69 316 86 646 022 065 249 024 4 8 L 28 R NE OBEN NE _ R E 98 RE 25 R 26 R _ 5 R R __ R 29 25 R __ R L5 24 R _ R 25 NENNEN ss L X 8 xr 20 NO vcl Table B13 Hand Arm Steadiness Test Results Rotational Steadiness X Axis Test Session 2 D R os 99 32 3 125 R 12 125 143 125 LU M 2 R 03 12 10 09 08 19 a M 23 FR 167 167 20 29 2350 239 on en os 3 M 5 Re nu 167 10 an Cs M FR os ee os on 083 M F 1 4 14 12 14 16 Co M 2 R 08 os 1 00 on os on Lu 32 10 T a pf F 2 F r s 125 10 083 F F 10 on 96 os Cm 23 R Hi or o 9n on R 12 14 1 4 1 4 14 125 Ca R 12 1 4 125 34 s 3 R 167 19 167 348 19 16 a nu au T 30 Ca R 12 125 125 39 ru in R 125 10 10 18 e zZ Z eee eem 20 Sel Tab
38. D ARM STEADINESS TEST INSTRUCTIONS TO SUBJECT e This test will measure the steadiness of your hand e will strap this small module to your hand Let me know if you find the strap too tight or loose e Extend your arm straight out in front of you and keep it as steady as possible e When start the test you will hear a single beep Keep your hand steady until you hear a double beep which will end the test The duration of each trial is 10 seconds e some reason you feel fatigue during the trial please inform me immediately INSTRUCTIONS TO EXAMINER e Plug the RSM connector into the main HPMM unit and ensure that it is connected properly e Make sure you strap the RSM securely to the subject s hand but do not strap it too tightly Ensure that the subject is comfortable e The subject must extend their forearm with palm facing downward and fingers extended and held together e Select READY on the HPMM touch screen only when you are sure the subject has attained correct position A single beep will signal the start of the measurement and a double beep will signal its end e Should the subject experience any fatigue stop the trial immediately and remove the RSM from the subject s hand Allow the subject to rest e Remember that the RSM is a very delicate device and handle it carefully 120 APPENDIX B HPMM VERSION 4 TEST RESULTS 121 CCl Table B1 Isometric Strength Test Results Test Session 1 Test Session
39. ESC ENTER UP or DOWN The core routine in the touch screen interface code 1 the SPI transfer routine that utilizes the SPI communication interface to send commands and obtain data from the TSC2200 touch screen controller The LCD subsection has a large set of subroutines To begin there are subroutines for building the start screens of each GTA and also the result screens of each GTA There are also routines that build screens for the menus i e the startup menu and the GTA menu Further there are system routines that set the cursor erase a page erase a row set the working page modify the attribute page and highlight a row and such other tasks At even lower levels there are routines which send a message byte or a command byte to the T6963C LCD controller There is also a routine to 78 convert from BCD to the LCD character map which is a very useful routine for storing results that will later be read by the LCD controller This subsection also contains a large set of data variables which are the contents of the menus and messages that go out to the screen Another set of routines are the serial communication subroutines These are used to connect the HPMM via the serial port to the host PC These routines are used to power up and enable the serial communication subsystem of the HPMM convert LCD characters to ASCII and at the core level to send a byte out of the serial port Finally a set of mathematical subroutines perform 24 bit di
40. Each sensor measures the angular rate along one axis and sensors are mounted so as to be able to sense angular rates about two orthogonal axes For example in case of the steadiness test one sensor measures the rate of flexion extension of the wrist which is about the x axis shown in figure 3 2 while the other measures the rate of pronation supination of the forearm which is about the y axis shown in figure 3 2 The nominal static sensitivity of these sensors is 1 11 mV deg s before signal conditioning and they have been found to produce fairly linear outputs for inputs up to 1000 deg s This brings to light the flexibility in the possible applications of these sensors While in the steadiness testing they are used to measure rates of a few deg s they can also be harnessed to measure quantities such as the rotation speed of the human arm about the shoulder joint which 1 of the order of hundreds of deg s The outputs of the angular rate sensors are filtered to remove the dc components and then amplified by a factor of 11 Again the raw outputs are also retained and both raw and conditioned outputs are digitized using two additional channels of the RSM 1 ADC For the angular rate measurement subsystem the overall system sensitivity for the A C coupled channel is computed as follows 53 Sensitivity Angular Rate Sensor Sensitivity x Gain x ADC Conversion factor 1 11 mV deg s x 11 x 4096 ADU 5V 0 ADU deg s Thus with the use of a
41. HRP does not allow for a grace period on the continual review requirement If the annual report is not received by the IRB by the anniversary date the IRB will be forced to terminate the approval of your protocol and if your study is federally funded the IRB will have to report the termination to OHRP as required by Title 45 CFR 46 103 b 5 If you require modifications to this proposal in the method of use of human subjects in this study change in the Principal Investigator PI or Co Investigator s or any change in the subject pool you are required to obtain prior approval from the IRB before implementing the modification as required by Title 45 CFR 46 103 b 411 The IRB approved consent form that is stamped by the IRB with the expiration date of the approval must be used for all informed consent procedures on al human subjects in this study The signed consent forms must be under lock and key on UTA campus for the duration of the study plus three years These consent forms are subject to inspection during this time period by the IRB Research Compliance staff and or federal agents All investigators listed in this protocol must have documented Human Subjects Involved in Research Tier II training on file with the Office of Research Compliance Please call the Office of Research Compliance if you have not taken the training course within the last year If you have any questions related to this research or to the IRB you may contact me
42. HUMAN PERFORMANCE MULTIMETER INVESTIGATION OF THE FOURTH GENERATION PROTOTYPE by RAHUL MULUKUTLA Presented to the Faculty of the Graduate School of The University of Texas at Arlington in Partial Fulfillment of the Requirements for the Degree of MASTER OF SCIENCE IN ELECTRICAL ENGINEERING THE UNIVERSITY OF TEXAS AT ARLINGTON December 2005 ACKNOWLEDGEMENTS The author wishes to express his heartfelt gratitude towards his supervising professor Dr George V Kondraske for his guidance patience and support The author would also like to thank Dr Wei Jen Lee and Dr William E Dillon for serving on the author s thesis committee and providing their advice when needed The author would like to express his thankfulness to Mr John Stevens for the help he rendered in constructing the system prototype Finally the author would like to express his appreciation for the support and encouragement given by his family and friends November 18 2005 11 ABSTRACT HUMAN PERFORMANCE MULTIMETER INVESTIGATION OF THE FOURTH GENERATION PROTOTYPE Publication No Rahul Mulukutla M S The University of Texas at Arlington 2005 supervising Professor Dr George V Kondraske The focus of this thesis is the development and investigation related to the newest prototype version of the Human Performance MultiMeter HPMM Version 4 is a technologically advanced compact portable and self contained unit that 1s the result of an
43. The four results are then converted into BCD format and finally mapped onto the LCD character map for display The following flowchart shows the structure of the software written to implement GTA 6 Although the later portion of it after the acquisition stage reflects only the path for translational steadiness the exact same path applies to rotational steadiness except the difference in the calibration offsets and the system static sensitivities 74 Load Channel 2 White tos PID A T Load GTA ID Erab Sensor Power Is 5 PIF Clear Code Memory Seti Issue beep Load C hannel 7 Switch code memory update data pointer Wteto5PIDAT Load C hannel 3 s 5 PIF Write to S PIDAT af I Read Is and copy to extemal memory Figure 3 9 Flowchart part 1 for 6 Hand Arm Steadiness Test 75 byte Swap nibbles s ave Read SPIDAT and copy to external memory Load C hannel Swap its rubbles and OE with White to SPIDAT swapped low byte save loa byte of yastitied sample Is s PIF 5 iret lugh b yte of sample Zem out low nibble and swap ribb les switch to external memory uplate data potter 5 ave high byte of yasttied sample All samples Fight pastitied to first acceleration s ample Y Figure 3 10 Flowchart part 2 for GT A 6 Hand Arm Steadiness Te
44. ance is multi dimensional In such cases the primary measure is derived as a combination or composite of two or more secondary measures For example in the coordination test also called the NMCC test speed and accuracy are secondary measures used to compute the primary measure termed neuromotor channel capacity Another type of secondary measure is one that characterizes some aspect of a test or a condition under which it was performed For example the tap duty cycle mean and tap duty cycle standard deviation are 86 characteristics of the finger tapping performance For duty cycle a larger or smaller duty cycle is not necessarily better A smaller variation of duty cycle however does correspond to a more consistent tapping performance This measure will be explored in the future for use In deriving a measure of consistency Exploratory measures are termed so because they are currently under scrutiny to solidify an understanding of how they can contribute to the characterization of the performance capacity of the human subsystem involved in relevant test task Each measure in tables 4 1 and 4 2 are marked to indicate how they are categorized Scatter plots are presented in the context of discussion of results for each test In the following section 87 88 Table 4 1 Summary of Experimental Results Separately for Dominant D and Non dominant N Sides Session 1 Session 2 Mean Pearson E CHEN echan
45. and accuracy will result in the best value of NMCC 68 Target Width W Figure 3 7 Touch Sensor Layout The major difference between the touch sensor arrangement for this and the previous HPMM versions is that the error regions in this version are separated into lateral and fore aft error regions as illustrated below This leads to the fact that while making a decision as to whether there was a hit or an error one must take into account the fact that that an error could be one of either lateral or fore aft nature This distinction is currently avoided by logically OR ing the sensor outputs for these two regions But in the future the errors will be distinguished in order to compute different lateral and fore aft accuracies 69 3 2 7 GTA 6 Hand Arm Steadiness Test The steadiness test has been given special emphasis in this thesis It has been expanded to include four measures instead of the two used previously by incorporating two angular rate measures in addition to translational accelerations The primary result in this test 1s one that has been formulated after application of the General Systems Performance Theory Kondraske 2000a and is a single number that represents the composite steadiness or 4 degree of freedom DOF steadiness of the body part under test Conceptual Background Steadiness can be measured with respect to any body part that 1s capable of motion To understand this generic test an outstr
46. apana ass 3l Sek Hardware 31 S Wall INDerocontrollep OFC uuu uu kaun u Q 32 Je User in or M 35 3 1 3 Optional Additional Memory 36 3 1 4 Programming and Updating HPMM Program Code 36 3 1 5 Main Unit Sensor and Stimulus Generator Subsystems 37 3 1 6 Remote Sensor Module 1 RSM l1 50 3 1 7 Power Management 32 55 55 59 3 toan se 55 23 2 SOlEWare ku 57 32 Operans T 57 22 2 Genero Test Al SOM IS a oe ch nad onset UR va deos n 62 3 2 3 GTA 1 Isometric Strength Test 64 3 2 4 2 Simple Visual Response Speed Test 65 3 2 5 GTA 4 Rapid Alternating Movement 66 3 2 6 GTA 5 Upper Extremity Coordination Test 67 3 2 7 GTA 6 Hand Arm Steadiness Test 70 3 9 S UDDOEL 78 2 4 Memory rion dak uu uter ion 80 4 EXPERIMENTAL EVALUATION 81 TIO v
47. as reduced the size and mass of inertial sensors and has also greatly increased their durability and cost effectiveness They are now considered attractive alternatives for tremor measurement in a portable device such as the HPMM In version 3 a dual axis accelerometer was employed The packaging of the sensor chip and signal conditioning was crude but results obtained were nonetheless encouraging Sriwatanapongse 2002 providing reasonable agreement to those obtained with the capacitive displacement sensor when appropriate conversions were applied A micromachined angular rate sensor was also incorporated into the version 3 architecture as part of the RSM The intent was to use this for a future generic test termed movement speed In the development of version 4 hardware two such angular rate sensors were included in the RSM design One commercial instrument for tremor measurement Motus Bioengineering Corporation Benicia CA incorporates a singe axis angular rate sensor Moore Ding and Bronte Stewart 2000 In version 4 the presence of four inertial sensors two accelerometers and two angular rate sensors will be exploited to more completely characterize the motions associated with tremor steadiness 20 2 6 Other Performance Requirements In order for the HPMM to function effectively as a portable device special consideration must be made for the power supply The batteries for the HPMM would have to be small but capable of providing
48. because their functions are not assigned to them by the digital crossbar Instead they are assigned by the HPMM system designer For example port O pins 6 and 7 have been assigned the WAKE and TS signal functions since they are connected to the DAV and PENIRQ signals on the touch screen controller Table 3 1 shows all the port I O pins used in the current HPMM system design their functions and how they are assigned The signals marked with an asterisk are assigned to the respective port pins using the digital crossbar For the signals marked with a signal assignments are context dependent for the external memory interface they are assigned using the digital crossbar for the LCD controller they are used as general purpose I O pins 33 Table 3 1 HPMM Port Pins and Descriptions SIGNAL PORT PIN NOs DESCRIPTION POO 62 Serial Communications Transmit ______ H fo Ree SCK PO2 60 JSPI Clck 0 0 0 WAKE UP PO6 56 signal from TSC2200 HOME touch sensor RD 1 P46 92 __ READ Signal ExternalRAM LCD 0 Signal on LCD Controller Address Line 0 on External RAM 1 2 Signal on LCD Controller 1 FS1 Signal on LCD Controller 1 RES Signal on LCD Controller A4 A7 P6 4 P6 7 76 73 Address Lines 4 7 on External RAM A8 A15 P5 0 P5 7 85 81 External RAM Address lines DO D7 P7 0 P7 7 72 65 Data Lines External RAM LCD 34 The F020
49. bled at all times Therefore there is no necessity for constant touch sensor health monitoring so that the heartbeat signal will be programmatically disabled 11 Sense direction POS NEG For positive sense direction POS the sensor is calibrated when no object 15 present and the OUT signal becomes active when object approaches the sensor For the negative sense direction the part is calibrated with the object present and the OUT signal becomes active in the object moves away from the 45 sensor No extensive analysis 1s required to show that positive detection 1s 1deal for the HPMM context 12 Detection Mode BG or OBJ BG background mode causes the calibration to occur at the baseline level as opposed to the signal level when the object 1s present The detection is made relative to this baseline reference level In OBJ mode the reference level becomes the signal level when the object 15 present Once again one can easily see that it is definitely more appropriate to use the BG mode in the HPMM context because the finger is more often away from the touch sensor than near or upon it The following tables provide the parameter settings for the QT310 and QT320 touch sensors in the present HPMM touch sensor subsystem The QT310 table also provides notes which summarize the reasons behind the choice of the settings and the same reasons also apply for the QT320 settings 46 Table 3 2 QT310 Parameter Settings Thres
50. crease in capacitance C will cause charge to 39 be transferred more rapidly into the sampling capacitor C This charge is then amplified by a charge amplifier which drives the input of a single slope switched capacitor ADC Additional signal processing is also implemented within the sensor IC to reject impulse noise To allow large values of Cx it is necessary to use a large C This increases the available resolution as well as the gain by decreasing the rise of differential voltage across C Longer burst lengths also increase gain and sensitivity but consume more power The following is a description of the various programmable parameters of the QT310 QT320 The QT320 is a two channel version of the QT310 It has a slower response time than the QT310 which however is sufficient for its applications in the finger tapping performance test and upper extremity coordination test In fact the touch sensor subsystem consists of four QT320 sensors and only one QT310 which drives the HOME electrode that is used in tests such as visual response speed where a higher sensor response speed is desired 1 Sleep Cycles SC SC is the number of intervals separating two consecutive bursts This influences the burst spacing parameter Tbs Tha SC x if SC gt 0 or Tos Tha 2 25 ms if SC 0 40 The parameter Tua 1 the burst duration and is a function of the number of pulses in a burst which again 1 a function of C and C
51. d data sets each command is followed by some data specific to that command The initialization consists of sending command data sets for specifications like text home address text area graphics home address graphics area mode setting address pointer setting etc Relevant reference information to support programming is supplied in the T6963C data sheet from Toshiba This LCD subsystem is a 5 V subsystem and the PWRE LCD bit in port 1 must be set in order to enable the REGIOI LDO which provides power to the LCD controller The POWERS5V bit must also be cleared to enable the L T1300 regulator thatprovides power all 5V subsystems A major component of the user interface are a series of hierarchical menus that permit selection of mode e g Generic Test Mode Protocol Driven Mode Maintenance 58 Mode etc and subsequently options associated with each mode Menu generation and selection management are coded as the MENUCTRL and ROUTECTRL subroutines respectively Only options associated with calibration and the Generic Test Mode are implemented at present Figure 3 4 The HPMM Main Menu In the Generic Test Mode the five generic tests that are currently 1mplemented can be selected Regardless of the test selected the operating system manages a common sequence of events Upon selection of a particular test the corresponding GTA routine is called which further call their own subroutines as required for the execution of the select
52. e that which would correspond to a 96 accuracy rate This 1s then used in modified version of Fitts relation to compute NMCC bits sec More recently Kondraske used General Systems Performance Theory to approach Fitts law from a different perspective Kondraske 1999 Kondraske 2000 It was found that a near perfect correlation between Fitts Index of Performance and the product of movement speed and accuracy in hitting fixed width targets An almost exact prediction was obtained by scaling the product using Fitts task difficulty index Version 3 of the HPMM incorporated touch sensors for NMCC testing and a preliminary evaluation of the NMCC in this context was carried out Sriwatanapongse 2002 The version 4 hardware platform incorporates new touch sensors and also divides the error regions surround the target regions into lateral and fore aft components 2 5 5 Steadiness Tremor Test Historically interest has been in the measurement of the pathologic state of tremor and measurement of hand tremor with an accelerometer has been most prevalent Potvin et al 1985 Takanokura and Sakamoto 2001 reduce the loading effect associated with mounting an accelerometer on a hand Kondraske developed a dual 28 axis non contacting capacitive displacement sensor for tremor measurement Kondraske 1986 A newer version of this device has utilized as part of the HPCMS for nearly 20 years In recent years micromachined sensor technology h
53. e strong agreement between these two generations would likely 91 contribute to validity of measures disagreement must be carefully evaluated as it would not be clear which version if not both represented the problem 4 4 1 General Observations Comparisons to data obtained from laboratory based instruments Potvin Tourtellotte et al 1985 shows good agreement for tests such as visual response speed isometric grip strength neuromotor channel capacity and finger tapping speed generally the GTA 15 called rapid alternating movement speed 4 4 2 Isometric Grip Strength The results for this test show good test retest reliability The Pearson correlation coefficients of 0 86 for dominant and 0 94 for non dominant sides are high and closely match those obtained using version 3 of the HPMM The grip strength scores also fall in an expected range for the subjects tested which has been established using laboratory based instruments as well as the previous version of the HPMM The scatter plot of Figure 4 8 shows points that are rather tightly distributed about the ideal line Comparison of mean grip strength shows that these measures are in good agreement with previously obtained values Figure 4 9 Overall these evaluation results are important The grip strength measurement performed with the current version incorporated the grip strength sensor into the handle of the HPMM This is therefore the first evaluation of this type of
54. e study procedures I will undergo have been answered to my satisfaction My signature indicates that I have made an informed decision to participate I have explained to the purpose of the experimental project the procedures required and the possible risks and benefits to the best of my ability Signature of investigator Date Signature of person obtaining consent Date has explained to me the purpose of the experimental project 1 have read or been read and understand this consent form I have been given an opportunity to ask questions regarding the experimental project and the study procedures I will undergo and I believe that I have sufficient information to give this informed consent Alternatives to my participation in the study have been discussed To the best of my knowledge I am not participating in any other medical research Therefore I agree to give my consent to participate as a subject in this research project Signature of subject Date JUL 11 2005 APPROVED BY THE UTA IRB The IRB approval fer this consent document wit expire on JUL 1 0 2006 3 Figure C4 Participation Explanation and Consent Form Page 3 142 REFERENCES Beasley W C 1961 Quantitative muscle testing Principles and applications to research and clinical services Archives of Physical Medicine and Rehabilitation 42 398 425 Behbehani K amp Kondraske G V 1986 Quantification of speed coordination and fatigue during a
55. ed packaging Administered the five performance tests to twenty healthy subjects Analyzed the large amount of data thus collected and performed statistical tests using specialized software Special emphasis was laid on test retest repeatability The validity of the measurements was investigated by comparison with previously obtained data through the previous HPMM prototype as well as laboratory based instruments Provided recommendations for future work 1n section 5 2 108 5 Conclusion After the experience of the experiments described in chapter 4 it is concluded that the performance of the aspects of the system tested is satisfactory The instrument was convenient to use in the Generic Test Mode More importantly very good test retest reliability has been found for all measures except for neuromotor channel capacity When dominant and non dominant side data were pooled reliability of all measures improved with neuromotor channel capacity test retest reliability approaching acceptable levels Validity of the results associated with the five generic tests studied was also supported by the fact that the values obtained for respective measures were in good agreement with the results obtained from laboratory based instruments and the previous prototype version of the HPMM It 1 thus concluded that the primary goal of this thesis has been achieved A set of important performance capacity tests have been implemented new hardware pla
56. ed test Prior to actual execution of a test the HPMM produces a screen with a prompt asking the operator to proceed when ready allowing the subject to be readied for a test by pressing enter When the selected test 1s completed results are displayed and the user 1s prompted to repeat the test or to return to the main menu 59 gt cod UM Pu ham Q _ et asf Figure 3 5 The Generic Test Mode Menu Fig 3 6 Example Result Screen A set of lower level subroutines were also developed for use by various generic test algorithms GTAs as need These are contained in the Utility Routines section of the code and include mathematical algorithms such as 24 bit division as well as 60 simple hardware management routines such as the BEEP routine which runs the piezoelectric beeper HPMM Resource Sharing The control of HPMM resources is switched between the OS routines and the GTA routines as the device is taken through a test session Upon startup the OS has control of the HPMM and this remains the case until the test administrator selects READY after choosing a particular generic test algorithm from the Generic Test Mode Menu At this time control 1s transferred to the GTA routines until the test 1s run completely and all results have been compiled Control is then regained by the OS routines which will run the user interface subroutines to display results Even in the case of repetition of a test the re
57. eds to be written in order to utilize its capabilities and hence implement the speech and audio performance tests Digital speech processing algorithms that are required for extra desired measures can be implemented on the fast HPMM microcontroller with an anticipation of good performance The touch screen currently serves as the user input device and generally performed well However this subsystem has numerous programmable parameters that affect its performance and these aspects should be studied to determine optimal responsiveness 1 the best balance between being sensitive enough yet not overly sensitive so as to generate false detections from glancing touches Appropriate software based timer based routines similar to those used in switch debouncing may be helpful Vigorous testing needs to be done to make sure that a smooth operation 1s achieved in situations like menu scrolling As noted previously the Protocol Driven Mode was not implemented on the version 4 platform Given that a basic operating system is now implemented and has 113 been reasonably exercised a reasonable goal would be to implement the Protocol Driven Mode see chapter 2 supporting protocols that consist of combinations of the five generic tests that have now been implemented 114 APPENDIX A HPMM VERSION 4 INSTRUCTIONS TO TEST SUBJECTS AND EXAMINERS 115 GTA 1 ISOMETRIC STRENGTH TEST INSTRUCTIONS TO SUBJECT e This test will measure the stre
58. edure for routine measurement of neuromuscular performance capacities Potvin et al 1985 Kondraske 1990 A commercially available device that is part of the HPCMS the Model BEP I Human Performance Measurement Inc Arlington TX measures central processing and upper extremity neuromotor control performance capacities including NMCC This device incorporates six high speed touch sensors along the front aspect of the unit for NMCC measurement It has been used and evaluated by others with a wide range of subjects Swaine and Sullivan 1992 Swaine and Sullivan 1993 Kauranen and Vanharanta 1996 An NMCC test is performed by asking the subject to use his her index finger and alternate between the narrower center sensors on left and right sides 40 6 cm separation as fast and as accurately as possible thereby stressing the involved neuromuscular systems along speed and accuracy dimensions simultaneously The first touch initiates a timed interval 10 sec during which sensor contacts are recorded and categorized as hits or errors Separate R to L L to R and average lateral reach 27 movement speeds and accuracy percent measures are computed Note that it is not required that the subject achieve 96 accuracy Rather a more broad range e g 60 98 is allowed which facilitates more efficient administration in clinical contexts Computations are performed that use the actual measured accuracy to determine an effective target width 1
59. ee 93 DOSIUO Lois e 93 a gt Final Position BEnd or One 93 4 6 Finger Position for Finger Tapping Performance 84 4 7 Hand Arm Posture Tor steadiness Test ett raro 84 4 8 Comparison of Grip Strength results from the two test sessions 93 4 9 Comparison of Grip Strength results from HPMM versions and 4 94 4 10 Comparison of results from two sessions of the visual response Speed u u eoe darte kau dened 95 4 11 Comparison of primary measure results from two sessions of the finger tappine performance test ooo uuu ette 96 4 12 Scatter plot for NMCC measurements between the two 98 4 13 Upper Extremity Coordination Speed scores for sessions 1 and 2 99 4 14 Upper Extremity Coordination Accuracy scores for the WOE SE SESSIONS 99 4 15 Comparison of Steadiness Values from present and Previous HEMM Versions u u u un ivi es te re edit 101 4 16 Comparison of sessions 1 and 2 for translational SIeddiness X 102 4 17 Comparison of sessions 1 and 2 for translational Steddiness es Y AXIS Sauce 10
60. enough power to operate the HPMM for a reasonable amount of time A power supply design study Hanson 2000 suggested that the batteries should provide at least eight hours of operation That should be sufficient for typical daytime operation in clinics The batteries can then be recharged during nighttime The current power supply design Hanson 2000 proposed the use of nickel cadmium NiCd rechargeable batteries for the HPMM A compact wall mounted A C to D C power module that is external to the HPMM is used to provide the power to the charging circuit when the batteries need to be recharged To optimize the battery power another issue to consider is power management Power management will be supported in hardware by providing the capability to turn each subsystem on or off by software This is done with the use of low dropout regulators LDO Power management implemented in software will control the power to each subsystem Certain subsystems can be turned off when not being used More detailed discussion and simulation model for the power supply can be found in Hanson 2000 30 CHAPTER 3 IMPLEMENTATION As noted the version 4 HPMM hardware platform was designed under Dr Kondraske s guidance as part of a one semester senior capstone design course This was an aggressive effort leaving little time for testing of even the most basic hardware functions Nonetheless a relatively complete version of the main unit was fabricated and
61. ensions of performance while time series data is collected This data is processed in real time to produce single number results representing availability of the isolated performance resource e g visual information processor speed This type of protocol is generally known as a maximal capacity test Descriptions for the generic tests listed in table 2 2 and which are included as the subject of study in this thesis are provided in chapter 3 2 4 Version 4 Hardware Platform Overview As noted previously a version 4 hardware platform for the HPMM main unit was designed as part of senior capstone design course A block diagram indicating key features is shown in figure 2 3 and photos of the device are shown in figure 2 4 and 2 5 This platform serves as the basis for further work pursued as part of this thesis Additional details regarding the hardware are provided in chapter 3 19 Block Diagram Main Processor Board PCBUT Touch Sensor Board PCBOZ rule Serial Port Llc Cygnal C8051F020 Processor Sensor Circuitry Touch Screen Controller B pin MiniDIN Handle Remote Sensor Module A bl ssembly Touch Screen and 4 button overlay Figure 2 3 HPMM System Top View 20 Figure 2 4 HPMM System Bottom View The features that may be observed from the top view are the handle assembly that doubles up as the force sensor for isometric strength tests the previously described LCD and touch
62. ental Results Separately for Dominant D and Non dominah N sides 98 Results of Analysis with Pooled Dominant and Nop dommant Sde atau u muu eR 89 Xl CHAPTER 1 INTRODUCTION The Human Performance Institute has been the seat of research in the area of the measurement of a wide range of performance capacities These include human sensory information processing neuromuscular and cognitive systems For these measurements a modular set of laboratory instruments have been developed Collectively known as the Human Performance Capacity Measurement System HPCMS this can be viewed as a set of items that can be combined in a variety of ways to realize various different application specific human performance capacity measurement systems Sriwatanapongse 2002 In 1992 the challenge to develop what was termed a Human Performance Multimeter HPMM was identified The HPMM was conceived as a portable compact version of the HPCMS that would exploit emerging microprocessor sensor and other technologies Several preliminary versions of the HPMM were designed and evaluated with each subsequent design converging to an optimization of functionality packaging and measurement performance This thesis represents the latest step 1n the evolution of the HPMM It focuses on verification of the functionality of a fourth generation hardware platform modification of previously developed test
63. erformance capacity measurements The result of this effort was the preliminary version 4 HPMM hardware platform Operating system and test algorithms from the version 3 effort Sriwatanapongse 2002 were not implemented and only very basic functional testing of the platform and its key subsystems was completed as part of this effort Table 1 1 Summary of HPMM Development Milestones 1992 First Conceptualization of HPMM small Business Innovative Research Grant Proposal to NASA G V Kondraske principal investigator 1996 V1 0 Design and Prototype Based on Senior capstone design course in 1992 proposal Definition of key Electrical Engineering spring operational modes partial functionality semester limited implementation of specific tests bench top realization no packaging issues addressed 2000 V2 0 Design and Prototype Senior capstone design course in Electrical Engineering spring semester 2002 V3 0 Design Prototype and Human EE Masters Thesis W Subject Testing first formal human Sriwatanapongse subject tests for five generic performance capacity tests 2002 V4 0 Design and Preliminary Prototype Senior capstone design course in More powerful processor low power Electrical Engineering fall increased display capacity touch screen semester enhanced sensors near final portable packaging 1 3 Objectives The main goal of this thesis 1s to develop software components and define
64. erview ODISCUVES yy u uu u a 81 22 IN CUM OOS S Be sio 86 4 4 Reliability and Validity of HPMM 90 bil General ODSeryvatlODs u u yuy tU sua asa 92 ZS OMe 02 4 4 5 Visual Response Speed E 94 4 4 4 Finger Tapping Performance 95 4 4 5 Upper Extremity Coordination 97 2 4 0 Hahnd rtm E 100 5 CONCLUSIONS AND FUTURE RESEARCH 107 S 109 5 2 Recommendations for Future Research 110 Appendix A HPMM VERSION 4 INSTRUCTIONS TO TEST SUBJECTS AND EXAMINERS 218 2 5 tec ali us 115 B HPMM VBRSION 4 TEST uzanan 121 C INSTITUTIONAL REVIEW BOARD DOCUMENTS 138 REFERENCE C 143 BIOGRAPHICAL INFORMATION 149 Vil LIST OF ILLUSTRATIONS Figure Page 2 1 Major Features of the Overall HPMM System Concept 12 2 2 HPMM System Block Diaeram
65. ery good degree of test retest reliability Since each of the constituent components appears to have a reasonable degree of validity 1 e no anomalies were observed in the numerical values of the two translational and two rotational constituent components it is reasonable to attribute at least a basic level of validity to the composite 4 DOF steadiness measure Rotational Steadiness X axis s deg C 2 D 9 D Side a N Side 1 5 Test Session 1 Figure 4 18 Rotational Rate Based Steadiness X axis comparison between sessions 104 Rotational Steadiness Z axis s deg Test Session 2 D Side a N Side 1 5 Test Session 1 Figure 4 19 Rotational Rate Based Steadiness Z axis comparison between sessions 4 DOF Steadiness in D 2 db D Side N Side 10000 20000 30000 40000 50000 80000 Test Session 1 Figure 4 20 4 DOF Steadiness comparison between the two sessions 105 One may notice from figures 4 18 and 4 19 that there is a fairly tight distribution of points around the ideal line except for one outlier This person with high rotational rate steadiness has influenced the 4 DOF steadiness result by causing an outlying point in the scatter plot for this measurement as well which is reflected in figure 4 20 106 CHAPTER 5 CONCLUSIONS AND FUTURE RESEARCH The work in this thesis represents the mos
66. etched hand arm combination is employed Of the six degrees of freedom required to completely describe motion three are translational and three are rotational The three axes involved in the quantification of the motion along these degrees of freedom can be described as vertical lateral and longitudinal for the hand arm combination with reference to a sensing point on the dorsal surface of the hand The hand arm combination or any object of interest can thus either translate along or rotate about either of these axes In the case of the hand arm combination we expect negligible translation along the longitudinal axis forward backward and negligible rotation about the vertical axis We thus restrict ourselves to only four degrees of freedom 70 LONGITUDINAL Z LATERAL X VERTICAL Y Figure 3 8 Three Axes Six Degrees of Freedom It may be noted also that rotational measurements are new to the HPMM design and have been incorporated beginning this current version Also while rotational measurements are rate based deg s translational measurements are acceleration based m s or g And in keeping with the principles of the General Systems Performance Theory the average values of these quantities are inverted to obtain translational acceleration based steadiness and rotational rate based steadiness where more steadiness is indicated by a larger numerical value Finally these four secondary measures are multiplied toge
67. for the study The set consisted of 10 males and 10 females The males averaged 25 1 years In age and the females averaged 24 4 years with the age ranges being 23 26 years for females and 23 28 years for males Subjects were recruited from the staff and student community at the University of Texas at Arlington The study was reviewed and approved by the UT Arlington Institutional Review Board A signed informed consent document was obtained from each subject 81 Subjects were tested in two sessions on the same day separated by a 10 minute break A given test consisted of a predetermined number of trials for each of the five GTAs Final results were determined using rules employed in earlier laboratory based instruments and earlier versions of the HPMM Detailed test administration procedures are provided in Appendix A The pictures below summarize the set up of the HPMM relative to the test subject for each of the tests evaluated Figure 4 2 HPMM Position for the Grip Strength Test 82 Figure 4 5 Final Position End of One Cycle 83 1 2 3 4 Figure 4 7 Hand Arm Posture for Steadiness Test simple Response Speed Test Final result 1s the average of three best trials out of the five performed Rapid Alternating Movement Test Final result is the average of two trials Neuromotor Channel Capacity Test Final result 1s the average of two trials Note previously the rule used was the better of two tr
68. ge r Snead NI Eee eee SD b 5 en 9 Finger Tapping Performance HEN EE HEC CEN QM NE M Speed taps s Finger Tapping Performance FP 3e es omo za 2s 089 Da ENE AM WW Et IE TR EE ONE W Finger Tapping Performance 10 75 2 00 18 6 11 36 2 48 21 7 22 0 0 24 mg FN s 25 ues 206 104 935 Upper Extremity Coordination IN NS MENT NMCC bits s Upper Extremity Coordination Lp s 7 sr 70 369 38s 98 Spee Ins RENE EN EE ME IE MN NM EN M Upper Extremity Coordination D 81 75 9 42 11 5 84 05 5 58 10 1 0 19 Acura D Fw mus e mus ex 187 995 Hand Arm Steadiness EXE RN EN EE M Wm mM M Translational Steadiness 1 g Hand Arm Steadiness Lp s me p seas 5 209 08 wa iy Hand Arm Steadiness 1 19 0 30 25 4 1 20 0 30 24 6 12 8 0 79 w pr os 1 oe Hand Arm Steadiness _ os ow Rotational Steadiness s deg Measure units Hand Arm Steadiness p s aes n ww em s 08 4 DOF Steadiness SUs Primary measure Secondary measure Exploratory measure 68 Table 4 2 Results of Analysis with Pooled Dominant and Non dominant Side Data
69. gher level tasks in daily life A modular measurement system architecture which facilitated expandability was introduced based on factors such as 1 The complexity of the human system and the recognition that there are more measures that will be required 2 Each patient or subject is unique It is thus likely that for optimal characterization a unique subset of the tests and measurements would be used in a given situation for a given patient or patients of different types This also allows us to view the system as a flexible measurement toolbox 3 It 1s highly desirable to integrate results acquired from several modules to facilitate clinical interpretation Largely with funding from the National Institute for Disability and Rehabilitation Research second and third generation laboratory based Human Performance Capacity Measurement Systems were developed The application independent philosophy and basic architecture facilitated expansion of the basic system to include modules which meet broader needs within rehabilitation 2nd generation Common denominator measurement issues across these diverse disciplines that were often hidden or confounded by different terminologies and traditions were identified by simultaneous involvement of different professionals which make up rehabilitation teams not only neurologists but also orthopedists physical and occupational therapists and others The name given to describe the system conseq
70. has two on chip analog to digital converters ADCs One is a 12 bit 100 KHz ADC and the other one an 8 bit 500 KHz ADC The HPMM is designed to use the internal 2 4 volt reference is used for these on chips ADCs The 9 channel 12 616100 KHz ADCO subsystem is used by the current HPMM prototype with the 2 4 volt internal reference The ADCO also has a programmable gain feature and eight out of nine channels are available for analog inputs while the ninth 1 internally connected to a temperature sensor ou put 3 1 2 User Interface The user interface includes a Liquid Crystal Display LCD module Model AGM2412C AZ Displays Inc with a 240 x128 dot screen and both character and graphics capabilities The module includes a Toshiba T6963C controller with 8k of memory It is connected to the microcontroller through a set of I O port pins see table 3 1 At 10mm thick this module 1s much thinner than screens used earlier versions A touch screen Model 95644 DYNAPRO has been incorporated into the design of version 4 that eliminates the push buttons used in earlier versions for user input and provides new flexibility for user inputs associated with yet to be defined performance capacity tests The touch screen is 117 2 mm x 88 4 mm x 1 3 mm and has a 0 5 mm horizontal resolution and 0 35 mm vertical resolution with the use of an 8 bit ADC in the controller A 4 button overlay 120 mm x 30 mm 15 applied over the lower portion of the
71. he screen for immediate observation and use The Protocol Driven Mode is considerably more sophisticated A protocol is defined as a predefined series of performance capacity tests In this mode a generic test now becomes a specific test For example isometric strength will now become strength 1 e a particular body part Is identified Moreover each result is labeled with its specific name and means must be provided for databasing of results Several different functions must be implemented to support the protocol driven mode For some of these the HPMM main unit 1s linked to a host PC Users can define one or more protocols e g Parkinson Disease screening Multiple Sclerosis screening etc using host based HPMM software and then download these protocols to the HPMM main unit The HPMM can be used its normal portable fashion and the user can select from available protocols and then will be prompted through this series of tests AII 13 results must now be labeled properly and recorded for subsequent upload to the host Prior to version 3 all hardware was contained in main unit Beginning with version 3 Sriwatanapongse 2002 it was decided to include the capability to connect a remote module to the main unit which is called a Remote Sensor Module The concept of RSM facilitates expandability of the HPMM Additional tests which required extra sensors and hardware can be added to the system by
72. his test is exactly the same as in version 3 for a more detailed discussion see Sriwatanapongse 2002 The procedure begins with the subject placing the index finger of the dominant side hand on the HOME touch sensor tips of digits 3 and 4 may also be included When this touch 1 detected it is checked for again after a one second delay An error is signaled if the finger is not present on the sensor by blinking the LEDs and the test is restarted This permits discrimination of a true attempt to start a test from an inadvertent touch of the home sensor by the examiner in the process of position the HPMM for this test If this check is passed a beep is signaled to start the test After a random delay of to 3 seconds both LEDs are turned on simultaneously A timer is started that will measure the time from the point the LEDs are lit to the point the subject takes the finger off the sensor with a quick motion of the fore arm The 65 reciprocal of this response time is computed and the result is the response speed with units of responses per second resp s This result 1 reported on the LCD 3 2 5 GTA 4 Rapid Alternating Movement Test The rapid alternating movement test is used to stress selected neuromotor coordination resources to assess how fast a person can reciprocally move about an isolated body joint A specific manner in which generic test is used is to characterize finger tapping performance which is used here as a basis to p
73. hold THR 6 Low for increased sensitivity 50 of the threshold ensures Hysteresis HYS 3 Higher count for ensurin Detection Inteerator DISA 10 We Ru valid detection Lower count for fast end of detection response Quicker compensation for removal of finger Slower compensation to ensure no false detections End Detection Integrator Negative Drift Compensation NDC Positive Drift Compensation PDC Burst Length Ty4 ms BL SC HB BG 2 10 25 Results In fast response time for HOME sensor Also results in faster response time Active low ensures low power consumption Disabled unnecessary for HPMM context Suitable because finger is more absent than present on the sensor Sleep Cycles Output Polarity OUTP Heartbeat Disable K BG Mode Infinite required for long touch durations 1 0 5 1 1 Table 3 3 QT320 Parameter Settings Common to Both Channels Peor mesos 08 enne Dri NDC 2 Pose Compensation PDC Busta Ty __ s 2 Sace fe 48 Force sensor subsystem The force sensor subsystem for the HPMM consists of a low profile top hat model load cell LFH 71 from Honeywell 250 Ib maximum load rating The load cell is a sub miniature force transducer that utilizes foil strain gages to measure compression loads Thus the sensor circuit is primarily composed of strain gages in a bridge configuration followed by a fixed gain instrumentation amplifier The stat
74. ials Steadiness Test The result 1s the average of the best two of three trials for each of four steadiness values namely two that are based on translational 84 accelerations a and ay and two that are based on rotational rate measurements and These results are then combined to construct the 4 DOF steadiness composite measure which is defined using performance theory concepts Kondraske 2000 as the mathematical product of the four constituent measures 5 Grip Strength Test Three trials are performed The result is the average of the two best trials After computing final measures descriptive statistics mean standard deviation and coefficient of variation were computed for each measure keeping dominant and non dominant side separately In addition the absolute value of the difference between Session 1 and Session 2 was computed for each subject and expressed as a percentage of the Session 1 measurement value These values were averaged across subjects to provide a single number indicator of repeatability e g mean of the absolute value of percent change Formal reliability measures were also computed between Session 1 and Session 2 Intraclass correlation coefficients ICC 3 1 were computed using SPSS version 12 Pearson product moment correlation coefficients were computed using Microsoft Excel 2003 The results were virtually identical for both methods thus only the Pearson correlation coefficient r 1s
75. ic sensitivity of the overall system is 25mV lb 5 618 mV N which provides a measurement range of 0 to 889 N and a corresponding output voltage range of 0 25V to 2 4V The sensor is calibrated to include a 0 25V offset with zero force applied so that small drifts keep the signal in an operable range The subsystem is packaged into the HPMM handle This allows for greater portability of the overall system because there is no separate cable or grip assembly as in the earlier versions of the HPMM The output of the force sensor 15 connected to analog input channel associated with one of the two ADCs in the HPMM microcontroller As mentioned previously this is an ADC which can sample at up to 100 KHz and which works in 12 bit mode There is an internal programmable gain amplifier feature gain options of 0 5 1 2 4 8 and 16 which 1 set to unity for the force sensing subsystem This ADC uses an internally generated reference voltage of 2 4 V and therefore the output of the sensor subsystem 15 constrained to remain within 2 4 V but appropriate choice of its static sensitivity The power to the force sensor subsystem 45V 1s provided by the power subsystem on board the HPMM through a separate connection 49 Visual Stimulus Generator To generate the visual stimuli that are required for different performance capacity tests such as simple visual response speed the HPMM is equipped with two high intensity light emitting di
76. ich is really a test type can be selected The user will make the generic test a specific test by choosing for example which neuromuscular subsystem to test This 1s like deciding which voltage to measure in a circuit The procedure will be carried out and the result will be shown on the screen If desired the neurologist could record the results manually 22 A wide array of generic tests is envisioned for the HPMM At present this thesis focuses on five generic tests These are all designed to be consistent with concepts of General Systems Performance Theory and the Elemental Resource Model for human performance Kondraske 2002 Furthermore each of these 1s derived from tests incorporated into the laboratory based human performance capacity measurement system Kondraske 1990 Each has also been implemented and evaluated to some degree in previous HPMM versions Kondraske 2005a Sriwatanapongse 2002 Substantial changes have been incorporated into the hardware design that affects each of these tests none have been implemented and tested on the current hardware platform A review of the relevant background for each of these performance capacity tests follows 2 5 1 Isometric Strength Test This is perhaps one of the most generic of all the HPMM tests It basically involves measuring the maximum force that a subject can generate or resist with a given neuromuscular subsystem such as a shoulder abductor or combination of such
77. icrocontroller s digital crossbar Power to the RSM 5V 1s supplied via a separate LDO on the main HPMM board X AXIS ACCELEROMETER RETER AMPLIFIER Band pass G 11 1 Hz to 22Hz Y AXIS ANGULAR RATE SENSOR 1 X AXIS FILTER AMPLIFIER High Pass G 11 fc 1 Hz ANGULAR RATE SENSOR 2 Z AXIS Figure 3 3 Conceptual Block diagram for Acceleration and Angular Rate Subsystems 3 1 7 Power Management The power management subsystem of the HPMM consists of two dc dc converters and five low dropout regulators LDOs all of which are housed on the HPMM main board The dc dc converters LT1300 of Linear Technology convert the 3 6V input voltage from the battery to 3 3V for the processor and touch screen controller and 5 V for the remaining subsystems respectively The second 5V dc dc converter has its active low Vsypn pin connected to a port pin on the microcontroller so 55 that the entire 5V subsystem set can be disabled with a single instruction This is especially useful when the HPMM needs to be placed in power saving or sleep mode The five low dropout regulators REGIOI of Texas Instruments all draw power from the second LT1300 described above They control power to the LCD the RSM the touch screen the serial communication subsystem and the accessories such as the LEDs and the beeper Each of these has an enable pin which 1s active high logic The enable pins are connected to unique port pins on the microc
78. integrated circuits Of these one is a QT310 while the rest are QT320 chips from Quantum Research Group An illustration of the touch sensor array 1s provided in Figure 3 7 under the software section The touch sensor regions can be divided into four groups namely the HOME sensor the LEFT group the RIGHT group and the extreme group The HOME sensor is a single region driven by the QT310 touch sensor IC The QT310 a single channel touch sensor has been chosen for the HOME electrode because of its faster response time as fast as 1 ms as opposed to the QT320 which has a best response time of 5ms This ensures accuracy comparable to lab based instruments for response speed tests The remaining regions are driven by QT320 ICs which are dual channel sensors Thus one QT320 can drive two regions The LEFT and RIGHT sensor regions are divided into target and error regions and the error regions themselves are divided into lateral and fore aft regions This division represents an improvement over previous designs of the touch sensor subsystem The software section of this chapter discusses the use of these sensor groups in specific generic test algorithms The extreme sensor regions are the two small regions located at the left and right bottom corners on the touch sensor board They were incorporated in the design to be used as keys on the backside of the HPMM main unit facilitating additional options in the design of the user interface during variou
79. integrated circuits have become available with apparently suitable characteristics for human performance measurement applications The current hardware design incorporates new touch sensor subsystems 2 5 4 Upper Extremity Neuromotor Channel Capacity Test In 1954 Fitts introduced a mathematical relationship between speed accuracy amplitude of movement and target size for upper extremity tasks This relationship derived using basic information theory constructs of Shannon has become widely known as Fitts law Fitts 1954 The mathematical statement of Fitts law was defined originally only for translational motion in one dimension In Fitts experiment which was not intended to be a measurement protocol but rather an attempt to understand human motion subjects held a stylus in their hand and were asked to move alternately between targets Performance was controlled to achieve 26 96 accuracy i e indicating that the system Isolated e g the upper extremity was being maximally stressed Movement time tm was measured Target width W and movement amplitude A were varied across a series of experimental trials with different subjects He found that data fit the relationship that is now known as Fitts law IP bits sec 1 tm logo W 2A IP dubbed by Fitts as the Index of Performance was shown to be relatively constant across a range of W and A values Kondraske and colleagues have adapted Fitts experimental proc
80. ion The ADC that is on the microcontroller chip ADCO is used for this test A beep signals the beginning of the test When the subject begins squeezing the grip sensor assembly the microcontroller constantly monitors the ADC output samples and waits for the applied force to cross a preset threshold 30 N Once the threshold is crossed the microcontroller samples the grip sensor output for a period of 3 seconds The calibrated sensor offset value is subtracted from each incoming sample and a variable that stores the sample corresponding to the maximum applied force is continuously updated After the 3 second period the processing ceases but sampling continues until the applied force falls below the threshold value The acquisition is then stopped and the highest applied force is the result of the test This is then converted from A D units henceforth referred to as ADUs to newtons and reported on the LCD 64 The system static sensitivity is calculated as follows Force Sensor Sensitivity 25 mV lb or 5 618 mV N Voltage Divider 0 5 V V ADC 4096 ADU 2 4 V 1706 67 ADU V Multiplying these system functions together we obtain an overall system sensitivity of 4 794 ADU N The calibration offset is subtracted from each sample as it is acquired and finally the sample corresponding to the maximum exerted force is converted from ADU to force units Newtons and reported on the LCD 3 2 4 GTA 2 Simple Visual Response Speed Test T
81. it 200 ADC TLV2553 By Texas Instruments for digitizing various signal conditioned outputs from sensor channels Thus digital data is passed to the main unit from RSM 1 for superior noise performance An input to the ADC on the main processor of the main unit is incorporated into the RSM interface This is used on RSM 1 for direct high speed digitization of the signal conditioned raw speech channel Acceleration measurement subsystem The design of the acceleration measurement subsystem is driven primarily by the desire to accurately measure the oscillatory movements produced in a subject s body part such as the hand arm combination However other possible future applications were also considered and this has impacted the final design utilized The present design includes an ADXL210 dual axis accelerometer with a sensitivity of I100mV g and specified input range of 0 to 10g The outputs of this accelerometer are of two kinds duty cycle modulated digital outputs and analog outputs Only the analog outputs are used in the design For steadiness tremor measurement these outputs one for each axis are first filtered with simple high pass and low pass filters to restrict their frequency content to lie between 1 Hz and 22 Hz and then amplified by a gain of 11 However for each axis the unamplified raw output signal that includes the dc component is also retained and connected to separate channels of the RSM 1 ADC These signals ca
82. ity were observed 1V TABLE OF CONTENTS ACKNOWLEDGEMENTS eese tt pet E ER PIER EHE PLN LISFOFILCUSTRA TION iaces Quote Mai com eund IST Chapter INTRODUCTION t escas db issue ti oa oe Uu decus 1 1 Human Performance Capacity Measurement Research Background 1 2 The Human Performance Multimeter LODEVE RIPE PER 2 ACKGROUND dade ba diente a evens 2 1 HPMM System Concept and Design PSY ONC EDL i 2 3 HPMM Functionality and Archttecture 2 4 Version 4 Hardware Platform OvervIew 25 Generic Performance Capacity 2 5D SOME MIC TES Qa DRM E NL GAS 2 5 2 Simple Response Speed Tests 2 5 3 Rapid Alternating Movement Performance Test 11 11 12 2 5 4 Upper Extremity Neuromotor Channel Capacity Test 26 2 3 0 Sead ness Tre mot Test uuu tr a a 28 2 0 Other Performance s o GE IS aa 30 35 IMPLEMENTA TION us cetus era E oad
83. k analysis and performance prediction CD ROM Proceedings of the World Congress on Medical Physics and Biomedical Engineering July 23 28 4 pgs Kondraske G V 2000b A working model for human system task interfaces In J Bronzino Ed The Biomedical Engineering Handbook 2nd Edition pp 147 1 147 17 Boca Raton CRC Press Kondraske G V 2001 Appendix A Design and Implementation Support Materials for Specific HPMM Performance Capacity Tests Human Performance Multimeter HPMM Conceptualization and Design Evolution HPI Technical Report TR2000 005R Version 3 1 Arlington TX The University of Texas at Arlington Kondraske G V 2005a Appendix A Design and Implementation Support Materials for Specific HPMM Performance Capacity Tests Human Performance Multimeter Conceptualization and Design Evolution HPI Technical Report TR2000 005R Version 3 2 Arlington TX The University of Texas at Arlington Kondraske G V 2005b Personal communication August 11 2005 Lachman R Lachman J L and Butterfield E C 1979 Cognitive Psychology and Information Processing An Introduction Lawrence Erlbaum Assoc Hillsdale NJ 146 Mayer T Kondraske G V Brady Beals S amp Gatchel R J 1997 Spinal range of motion accuracy and sources of error with inclinometric measurement Spine 22 17 1976 1984 Moore G P Ding L and Bronte Stewart H M 2000 Concurrent Parkinson trem
84. l help to make the subsystems more accurate as the overall system design will be able to account for sensor manufacturing variability In the version 4 hardware platform the touch sensor design was changed from version 3 to version 4 n version 3 a central target was surrounded by an error ring In version 4 this ring was divided into two lateral regions that are electrically connected to each other and a fore and an aft region 1 e above and below the target which are also electrically connected The ability to distinguish the type of errors made was not exploited in the current version In the future versions of the NMCC test 111 algorithm it is proposed that a distinction be made and the errors be classified as lateral or fore aft errors The only change would be to include at least one additional secondary measure lateral to fore aft error ratio There is currently no basis for knowing what value of this quantity would be considered normal or better although it would seem that a random distribution would be found in healthy subjects and the thus a ratio of approximately 1 0 might be expected Another issue has been noted in relation to the upper extremity coordination test It appears that stopping the test exactly after ten seconds results in having some subjects lose a cycle if at the end of the tenth second the finger is In mid flight en route to the touch sensor It is proposed that the test be stopped after the com
85. lated for each and every tap For the computation of the results the following equations are used Tapping Speed Number of Taps Total Acquisition Time Tap duty cycle mean Sum of Tap Duty Cycles Number of tap duty cycles Tap Duty Cycle SD X Tap Duty Cycle Duty Cycle Mean p Number of Tap Duty Cycles 3 2 6 GTA 5 Upper Extremity Coordination Test The coordination or neuromotor channel capacity NMCC of the upper extremity is measured by this GTA This performance resource 1s related to the speed and accuracy of movement and these two measurements form the secondary measures of the test while NMCC is the primary measurement A detailed description of the algorithm is provided in Sriwatanapongse 2002 The procedure remains the same for this version of the HPMM A single beep signals the beginning of the test The data acquisition is in the same fashion as for the previous GTA e g continuous 10 ms interval samples of all the touch sensors are saved in memory and includes the three sample check and the three second timer for eliminating false starts The subject is advised to alternately touch the left and right target regions which are illustrated below 67 while trying to be as accurate and as fast as possible A single beep signals the beginning of the test The sensors are checked to ensure that the finger is present on any one sensor left or right for at least three consecutive samples When
86. le B14 Hand Arm Steadiness Test Results Rotational Steadiness Z Axis Test Session 2 N x R 05 on 08 08 99 99 R o o on 96 06 M 2 R 05 os oe 067 05 08 _ 4 M 23 R 1 2 is 167 29 235 3 Rr j 09 05 o9 95 Ff Rr 08 063 067 032 030 e oo on en 067 3 M 09 03 07 032 o 92 M 26 R 10 oor om 08 9n to M oss os 0 94 0 os au F 24 R 0 oe 10 08 99 Co F a R on 06 en es en 96 a Fr 39 x er os o9 os on 98 Cm s R r s os 05 06 R 125 1 4 19 1n 1 4 16 Ca R r 10 on H 19 3 R 18 19 15 18 Ca R 0 H 09 125 dr 99 x R on or os os on om _ R on os 05 on os on e periere z lt lt lt lt lt lt lt ped 20 961 Table B15 4 DOF Steadiness Test Session 1 Sex Aoi 4 DOF Steadiness SUs Non dominant Side 944 585 629 623 388 585 624 508 792 561 9809 7000 3961 47906 08 NO 5303 7101 3961 4796 3354 F 25 R 2983 68 98 58 6197 SU Steadinesss Unit 15 ON
87. merely creating new RSMs The RSM also increases portability ease of operation and measurement quality from an ergonomic perspective The RSM can be placed on or attached to the selected part of the subject s body that needs to be tested while the test administrator controls the tests and views test results from the main HPMM unit The RSM may be customized to include certain sensor systems that specifically suit certain applications which may lead to the use of the HPMM in broader application fields The host PC contains HPMM software that enables communication with the HPMM main unit and also performs higher level administrative functions associated with performance capacity measurement It is only relevant for the Protocol Driven Mode Communication is designed to be accomplished via a RS 232 serial link As noted earlier in describing the Protocol Driven Mode the host software lets the user create test protocols and download the test protocol to the main unit Test results can also be uploaded from the HPMM a report can be printed and results can be stored in the PC Host software will ultimately manage the results database and provide search and results analysis capability The use of HPMM along with the host PC is suitable for 14 applications that need to perform standard test protocols and store a large amount of test results database such as in clinical use In such settings test protocols will be downloaded to the HPMM unit and the
88. n be useful for other performance tests and also for extending the 51 effective measurement range for very large and high frequency abnormal motions Both the conditioned and raw outputs are thus available for digitization using the ADC The accelerometer chip is mounted perpendicularly to the main RSM printed circuit board 1 along the y axis shown in figure 3 2 below to measure accelerations along the x and y axes Figure 3 2 The Remote Sensor Module 1 RSM 1 The overall system sensitivity for the A C coupled channels is Sensitivity Accelerometer Sensitivity x Amplifier Gain x ADC Conversion Factor 100 mV g x 11 x 4096 ADU 5V 901 12 ADU g where ADU Analog to Digital converter units For the 12 bit A D used the measurement range is thus 2 to 4 5 2 The overall sensitivity for the D C coupled channels is 81 92 ADU g This provides a theoretical measurement range of 0 50 g which 15 more than the sensor s output range 0 10 Thus the effective range 15 0 10 52 Angular Rate Measurement Subsystem This subsystem consists of two integrated circuit MEMS based angular rate sensors ENCO5 EA of Murata The basis for the operation of these sensors is the Coriolis Effect An out of plane bending force called the Coriolis force 1s caused by the momentum stored in a vibrating element when a rotation is applied to it The sensor demodulates this force and accurately depicts the rotational rate
89. n index finger tapping test Proceedings Eighth Annual EEE Engineering in Medicine and Biology Society Conference pp 605 606 Callaghan J T Cerimele B J Kassahun K Nyhart E H Hoyes Beehler P J Kondraske G V 1997 Olanzapine interaction study with imipramine J Clinical Pharmacology October 971 978 Edwards R H T amp Hyde S 1977 Methods of measuring muscle strength and fatigue Physiotherapy 63 2 51 55 Fitts P M The information capacity of the human motor system in controlling the amplitude of movement J Exp Psychol vol 47 pp 381 391 1954 Halstead W C Brain and Intelligence Chicago University of Chicago Press 1947 143 Hanson C 2000 HPMM Electronic Hardware Design Document Power Supply and Management EHWDD 06 V25 Arlington TX Human Performance Institute The University of Texas at Arlington Human Peformance Measurement Inc 2004 Human Performance Capacity Measurement System The Manual Human Performance Measurement Inc Arlington IX Hyman R 1953 Stimulus information as a determinant of reaction time J of Experimental Psychology 45 423 32 Kauranen K and Vanharanta H Influences of aging gender and handedness on motor performance of upper and lower extremities Perceptual and Motor Skills vol 82 pp 515 525 1996 Kolber M J and Cleland J A 2005 Strength testing using hand held dynamometry Physical Therapy Reviews
90. ng that the calibration routine has been executed the test administrator straps the RSM to the hand of the subject The test administrator can select the steadiness test from the GTA menu and commence the test by pressing START The power subsystem associated with the RSM is activated and a small delay is provided to 73 allow the sensors to stabilize A single beep signals the test start The ADC routine is called which samples the four channels coming from the four sensor outputs successively Thus one set of four samples is obtained every 1015 While the accelerometer channels samples are stored in on chip external memory the angular rate sensors samples are stored in code memory After a period of 10s the sampling is stopped and analysis begins In the analysis section the accelerometer samples are processed first one axis after the other Firstly the left justified samples are bitwise shifted to right justify them For each sample the calibrated sensor offset 1s subtracted and the absolute value of this difference is computed Then for each axis the average of the absolute differences is computed This average in ADUs is then divided into the static sensitivity of the system 901 12 ADU g to obtain in units of 1 g the required translational acceleration based steadiness value The same procedure is repeated for the rotational rate based steadiness measurement after moving the acquired samples from code memory into on chip memory
91. ngth of the group of your muscles that is used in producing a grip with your hand e Siton this chair facing me e will hold the device like this and ask you to squeeze the handle examiner grasp the long edges of the HPMM with one edge in each of your hands and the force sensor handle vertical Hold the device at a level above the floor so that the subject s forearm is approximately parallel to the floor e Try to squeeze the device now so you may get a feel for it During the test you will need to squeeze as hard as you can for about three seconds Do not push toward me or pull toward you just squeeze e now that you are familiar with the procedure we will run the test Press the touch screen button to start the test and then hold the instrument in front of the subject e When you hear a beep you may start squeezing as hard as possible for about three seconds then relax I will verbally urge you to squeeze as hard as you can during the test Upon completion you will hear a double beep e We will do this trial twice INSTRUCTIONS TO THE EXAMINER e Configure the HPMM so that the force sensor handle is above the touch sensor panel and locked in this position e Select Isometric Strength Test from the menu and proceed up to the point where the next touch screen entry will start the test e Give the subject the opportunity to become familiar with the device and procedure before initiating the actual test see above e Then
92. not expected in these evaluations since data from different subjects is being compared Some of the measurements hitherto reported using comparable laboratory instruments are also slightly different sometimes in terms of the approach to the measurement itself and at other times in trivial areas such as the unit used For example isometric grip strength was sometimes reported with units of pounds and since the introduction of the HPMM is measured in newtons The earlier versions of visual response tests measured and analyzed data in response times but we now measure this quantity as response speed In such cases results may easily be converted from their old units to newer ones for comparison with present results In cases such as that of the steadiness test where measurement was earlier made using a non contacting capacitive displacement transducer a mathematical transformation needed to be performed because previous versions of HPMM used a measure of acceleration as the basis for computing steadiness performance In other words displacement based steadiness results had to be converted to acceleration based steadiness Another measurement the rotational rate based steadiness did not exist prior to the current fourth generation HPMM prototype However in all measurements except rotational steadiness a reasonable direct comparison of results from third and fourth generation HPMM systems can be made Finally 1t should be noted that whil
93. ode The first involves circuitry that can connect the HPMM via the JTAG interface header to a serial to JTAG adapter The serial to JTAG adapter is connected to the Host PC via the serial 36 port A proprietary integrated development environment IDE running on the host PC then allows code to be compiled linked and downloaded onto the HPMM via this serial to JT AG adapter This approach is good for initial programming and testing but since both the adapter and the IDE are necessary field programming is not feasible in this fashion For field programming a simpler approach is adopted A firmware updater has been developed and initially downloaded into the HPMM microcontroller flash memory using the JTAG approach Subsequently a direct connection between the HPMM and the host PC via the included HPMM serial port is used to download the HPMM code in Intel HEX file format Simple custom software running on the host will manage handshaking and control of this process 3 1 5 Main Unit Sensor and Stimulus Generator Subsystems The main HPMM unit integrates several sensor and stimulus generator subsystems They include the touch sensor subsystem the force sensor subsystem and the LED based visual stimulus generator subsystem The following discussion describes each of these in detail o Touch sensor Subsystem The touch sensor subsystem consists of thirteen touch sensor regions that are driven by five high speed capacitive touch sensor
94. odes LEDs on the underside of the main unit They are capable of emitting red diffused light of intensity greater than 2000 mcd and are connected to the HPMM microcontroller through an LED driver circuit that consists of two independent power MOSFET switches IRFD9110 of International Rectifier one for each LED These LEDs can be independently controlled because each MOSFET 1 connected to a separate port pin on the HPMM microcontroller RIGHT port 2 pin 7 and LEFT port 3 pin 5 and the corresponding port registers are bit addressable Thus by simply driving the port pin high or low these LEDs can be turned off or on as required 3 1 6 Remote Sensor Module 1 RSM 1 As indicated previously the HPMM is equipped with a general interface to support a Remote Sensor Module RSM This interface includes several signals including 5V power and ground connections to five digital port pins and a connection to an analog input channel on the microcontroller The definition of the digital port pins will be application specific Currently one RSM module has been defined and it is denoted RSM 1 It consists of three subsystems acceleration measurement angular rate measurement and speech processing These are all integrated into one small housing that measures 49mm X 34mm X 19mm and weighs 32g It is connected to the HPMM 50 main board through an eight wire flexible cable and an 8 MiniDIN connector RSM 1 contains an 11 channel 12 b
95. on going effort in human performance measurement covering a 25 year history at the Human Performance Institute involving the conception development and evaluation of over four hundred different measures Such measurements are applicable in areas ranging from medical diagnosis and rehabilitation to ergonomics and athletic proficiency iii The work included verification of the functionality of a new hardware platform modification of previously developed test procedures and software algorithms for the new hardware and new development for selected aspects of the system Steadiness measurement has been expanded to include components based on two axes of rotational rate and the formulation of a four degree of freedom composite steadiness measure Five generic tests incorporated in this version of the HPMM were used in an experimental study with 20 healthy adult volunteers to evaluate reliability and validity Selected performance capacities of specific body subsystems isometric grip strength visual hand response speed index finger tapping speed upper extremity neuromotor channel capacity and hand arm steadiness were measured in a test retest design Test retest reliability was found to be very good r 0 75 for most measures Results were compared to those from HPMM v3 0 and to results from pre established validated data acquired over the years from laboratory based instruments Good agreement was noted and expected patterns supporting valid
96. ontroller Thus power to each subsystem can be individually enabled or disabled when needed by using software instructions for power management Bench measurements were made with a digital multimeter to determine the power consumption impact from the perspective of the rechargeable battery for each of the major HPMM subsystems Table 3 4 At the input to the power supply subsystems 1 the point at which the battery would be attached the multimeter was inserted in series with the positive supply lead to an adjustable voltage bench power supply set for an output of 3 4V Current was measured with all subsystems disabled 1 only the microcontroller and basic support subsystems powered and then with each subsystem enabled one at a time The change in current over the microcontroller baseline was computed for each subsystem This is multiplied by 3 4V to determine the impact on power consumption at the battery 56 Table 3 4 Contribution of Various Subsystems to Power Consumption at Source Contribution to Change in Power Subsystem Current AC AN mA mW Remote Sensor Module 111 86 Touch Sensor Subsystem 77 86 Microcontroller and support 606 20604 The table shows that the total power consumed in a condition where all subsystems are enabled is 598 06 mW Given that the proposed power source for the HPMM is a battery pack that provides 1600 mAh at 3 6 V the power rating of the battery pack 1s 5760 mWh
97. or of the instrumentation Nonetheless as figures 4 16 and 4 17 illustrate the system is able to discriminate differences in the steadiness of healthy subjects with good repeatability 1 e for the most part those who are least stable during Session 1 are the least stable during Session 2 and vice versa 100 Along the y axis the average translational steadiness of the subjects 1s higher than that along the x axis figure 4 16 This is somewhat unexpected because more movement results in the y direction when the RSM is mounted on a subject s hand according to the procedure for this test see Appendix A Upon further investigation this anomaly is attributed to the fact that the noise levels on the x axis output of the sensor are slightly higher than those on the y axis output Chapter 5 discusses a possible remedy for this situation The test retest repeatability of the x axis translational steadiness is comparable to that of the y axis translational steadiness as indicated by their Pearson correlation coefficient values in Table 4 1 The coefficients themselves are somewhat higher than those obtained with version 3 They are surprisingly high given that the number of degrees of freedom in motion providing a wide variety of ways in which a subject could conceivably exhibit the same amount of overall steadiness upon repeat testing Translational Steadiness Acceleration Based ws L1 E
98. ors Journal of Physiology 529 1 273 281 Potvin A R Tourtellotte W W Potvin J H Kondraske G V and Syndulko K The Quantitative Examination of Neurologic Function Boca Raton FL CRC Press 1985 Prigatano G P Wong J L Speed of finger tapping and goal attainment after unilateral cerebral vascular accident Arch Phys Med Rehabil 78 847 852 1997 Smith S S 2000 Measurement of Neuromuscular Performance Capacities In J Bronzino Ed The Biomedical Engineering Handbook 2nd Edition pp 148 1 148 12 Boca Raton CRC Press Sriwatanapongse W 2002 Human Performance Multimeter Investigation of the Third Generation Prototype Master s Thesis University of Texas at Arlington 147 Swaine and Sullivan S J Relation between clinical and instrumented measures of motor coordination in traumatically brain injured persons Arch Phys Med Rehabil vol 73 1 pp 55 59 1992 Swaine B R and Sullivan S J Reliability of the scores for the finger to nose test in adults with traumatic brain injury Physical Therapy vol 73 2 pp 71 78 1993 Takanokura M and Sakamoto K 2001 Physiologic tremor of the upper limb segments European Journal of Applied Physiology 85 214 225 148 BIOGRAPHICAL INFORMATION The author was born in Pune India in the year 1981 He received his Bachelor of Engineering degree in Telecommunications Engineering from Visveswaraiah Technological University
99. packaging which reflects final design packaging In addition the test instructions see Appendix A have been modified to 92 improve grip strength measurement in extremely weak subjects and also to facilitate test administrator interaction with the HPMM user interface The good reliability and the finding of measured performance in the expected range for this population are encouraging in light of the incorporation of these new features Isometric Grip Strength N CX un 9 gt D Side a N Side 300 400 5n Test Session 1 Figure 4 8 Comparison of Grip Strength results from the two test sessions 03 Isometric Grip Strength c di Figure 4 9 Comparison of Grip Strength results from HPMM versions 3 and 4 4 4 5 Visual Response Speed Since this test focuses on central processing it is not considered to be side dependent and is thus performed using only the dominant hand as a participant but not the system of focus in the test The results of these measurements show clear test retest reliability with a high Pearson coefficient of 0 83 Measurements also fall within the expected range and compare well with previously obtained values Figure 4 10 shows a comparison of results from the two test sessions 94 Visual Response Speed resp s C e un ul a 9 TN ul 3 4 Test Session 1 Figure
100. packaging concept and to allow for testing of the strength of muscle groups other than those associated with hand grip 2 5 2 Simple Response Speed Tests This test is representative of a class of tests commonly referred to as reaction time tests Kondraske and Vasta 2002 These tests involve responding as quickly as possible to some type of stimulus e g visual auditory etc in some specified manner e g moving a body segment The stimulus is required to have a low information load and thus requires minimal cognitive processing This gives rise to the 24 characterization of the test as simple which also distinguishes it from other tests of information processing speed This class of tests has long been used to characterize subjects with neurologic diseases such as multiple sclerosis and Parkinson s disease Potvin et al 1985 and individuals who have sustained traumatic injuries such as concussions or other head injuries It is also useful in detecting and characterizing neurologic side effects of drugs Callaghan et al 1997 It is therefore desirable to incorporate this type of test into the HPMM it has been 2 5 3 Rapid Alternating Movement Performance Test Halstead introduced a maximal performance finger tapping speed test as one component of a basis for discriminating the intelligence of individuals Halstead 1947 This test has been incorporated into a popular neuropsychologic test battery known as the Hal
101. pletion of the flight cycle in progress when the ten second test duration elapses It is also recommended that three trials be taken for side dominant non dominant during a test session and the average of the best two trials be considered the result for that session These changes should improve the reliability of the coordination measurements With regard to the finger tapping performance test consideration should be given to sampling touch sensors at a much faster rate about 1 KHz so that time samples are 1 ms apart This will allow for increased accuracy in measurements such as the duty cycle standard deviation which in the current situation has been shown to exhibit low reliability Use of the optional 256 KB external memory chip will allow the storage of the much larger number of touch sensor samples that will result from the increased rate 112 Power management issues have been dealt with in an efficient manner with a well designed power subsystem that includes dc dc converters and low drop out LDO linear regulators A battery charger subsystem is present on the version 4 main unit but it was not exercised as part of this thesis work Another facility that will be planned upon is the implementation of a sleep wake up power management mode Inactivity lasting longer than fixed time duration will cause the HPMM to go into this power save mode A speech processing subsystem has been designed and the hardware implemented Software ne
102. pon a MOD recalibration In the context of the HPMM a low threshold is required because it translates to high sensitivity 4 Hysteresis HYS 0 255 Hysteresis is measured in terms of counts of signal deviation relative to the threshold level The output becomes inactive when the C level falls below the level corresponding to THR SYS A zero value represents no hysteresis but HYS should not be set greater than THR this may cause a malfunction For best results it must be set between 10 and 40 of the threshold value Hysterisis affects detection stability An optimal value of hysterisis is required for the signal to remain stable during detection If hysterisis is set too low even very small changes in the detected signal will cause the output to switch which is not desirable especially in case of the finger tapping performance test where measures such as the tap duration and duty cycle are vital to the assessment of the performance resource 5 Detect Integrators Detect integrators serve to filter out sporadic electrical noise The detection integrator DIA is a counter that increments as soon as the signal crosses 42 the threshold level until it reaches the terminal count DIAT 1 255 when the OUT signal is activated DIB is the end of detection counter which counts up as soon as the signal falls below the hysteresis level and the OUT signal is deactivated when DIBT 1 255 is reached The HPMM touch sensor settings will include
103. press ENTER on the touch screen to initiate the actual test and carefully hold the HPMM in front of the subject as described above e beep will sound ask the subject to squeeze while applying maximum possible force for 3 to 5 seconds e Encourage the subject while squeezing by saying Squeeze Squeeze Squeeze in about 1 sec intervals Ok Stop Be sure that the subject only squeezes and does not push toward you or pull away from you during the test Use your own control of the instrument to balance out push or pull forces while the subject squeezes e When finished ask subject to relax Press REPEAT on the HPMM screen to repeat the test 116 GTA 2 SIMPLE VISUAL RESPONSE SPEED TEST INSTRUCTIONS TO SUBJECT Place your finger on this region here direct subject to home sensor region You will hear a beep which indicates the start of the test In any time between one and three seconds both these red lights will light up As soon as this happens you must lift your finger off of the region When you lift your finger off the region make sure you move your entire arm and not just your finger or wrist Do not try to guess as to when the lights will come on Once you are done lifting the finger off the region place it back there for the next trial INSTRUCTIONS TO EXAMINER Flip the HPMM over and show the home sensor to the subject Make sure the subject is seated with their arm in line with the home sensor Show the subjec
104. rding my participation in the study and will be made available for review only as required by the Food and Drug Administration REQUEST FOR MORE INFORMATION If I have any questions about the study I should contact George Kondraske 817 272 3454 If 1 need to report any adverse event or problem concerning my participation in the study 1 should contact George Kondraske 817 272 3454 If I have any questions about my rights as a research subject I should contact UTA Office of Research Compliance 817 272 3723 REFUSAL OR WITHDRAWL OF PARTICIPATION My participation is voluntary and I may refuse to participate or may withdraw consent and discontinue participation in the study at any time without prejudice to my present enrolment or employment at the University of Texas at Arlington Either George Kondraske or Rahul Mulukutla may terminate my participation in this study after explaining the reasons for doing so JUL 1 1 2005 APPROVED BY THE UTA IRB The IRB approval for this consent document wil expire on JUL 1 0 2006 I will be given a copy of this form to keep Figure C3 Participation Explanation and Consent Form Page 2 141 CONSENT TO PARTICIPATE I am making a decision whether or not to participate in this study 1 should not sign this consent form until I have read or have been read and understand the information presented in the previous pages and until all my questions about the experimental project and th
105. reased acceleration and indicate less steadiness for more motion or acceleration Rotational Rate based Steadiness For obtaining a model for this kind type of steadiness measurement we begin with the rotation of the hand about an axis The rotation rate equation is rad s It 1s estimated the range of displacements of interest on the back of the hand due only to angular motion at the wrist or forearm would range from approximately 0 1 mm to 72 10mm Given an estimate of the moment arm r the size of angle can be estimated Incorporating an estimate of the frequency associated with the angular motion f 5 Hz an estimate of the range of angular rates can be determined Table 3 2 lists these estimates for three different radii The first column represents the radius which in other words is the distance of the rotational rate sensor from the wrist The second and third columns are mathematically estimated angles and rotation rates These values were used to determine overall gains and measurement range in the hardware design Table 3 6 Estimates of rotational rate based on displacement estimates wis Angle of Rotation Rate of Rotation E deg deg s Mn Max Min _ b The results of these measurements in deg s are converted as mentioned earlier to s deg in order to obtain values of steadiness for two axes that is consistent with performance theory Measurement Algorithm Assumi
106. red touch sensor areas Have the subject go through the required motions for a few seconds and select the most comfortable position Usually this is with the forearm resting on the examination table the HPMM shifted toward the side of the involved arm and placed further back on the table and with either one of the left or the right touch sensor regions under the subject s finger Encourage the subject so that the best effort is made continuously Be sure that the subject keeps his or her finger stiff that flexion occurs only at the knuckle and that the finger is definitely lifted not wiggled around each tap cycle Some subjects may have difficulty obtaining distinct lifts but it will be apparent that additional speed is not being gained Others will try to gain a speed advantage by staying very close to the plate Interrupt the test give the subject a short rest and start over if necessary to obtain compliance 118 GTA 5 UPPER EXTREMITY COORDINATION TEST INSTRUCTIONS TO SUBJECT e The purpose of this test is to measure your ability to make coordinated movements with your hand and your arm e We will use these two regions of the HPMM Each of these left and right has a target and an error region show regions e Should your finger land on the target region it will count as a hit A landing on the error region will count for a miss e Use the pad of your index finger to alternately tap between these left and right regions as fa
107. rength simple visual response speed rapid alternating movement upper extremity coordination and hand arm steadiness 6 Conduct a performance evaluation of the current version HPMM prototype in a complete final package form 1 not a bench set up and more specifically then evaluate the reliability and validity of the five generic tests noted above in a population of healthy young adults T Provide concluding thoughts and recommendations for future improvements and future experimental work 10 CHAPTER 2 BACKGROUND 2 HPMM System Concept and Design The concept of the Human Performance MultiMeter HPMM is analogous to that of a digital multimeter used in a laboratory Its purpose is to integrate as much functionality of laboratory based human performance measurement instruments as possible into one single portable unit The platform design for this device is structured similar to that of a personal digital assistant PDA Apart from the integration of various sensor units that support performance measurements the platform must present a graphical user interface that allows navigation of menus and accepts commands that will cause the system to perform the desired performance tests Complete descriptions of the system design evolution design process and current status are provided elsewhere Kondraske 2005a This chapter presents a review and summary of the overall system concept and current design as well as background
108. reported here Typically there are differences in dominant and non dominant side performance with dominant side values generally higher than non dominant side especially for more complex tasks To increase the measurement range over which these measures are exercised dominant and non dominant side data for a given measure were pooled and all analyses listed above were repeated for these data sets 85 Scatter plots were also generated Session 2 vs Session 1 for each measure These provide a detailed overview of all data points in the context of test retest repeatability 4 3 Results Results from the analyses described are summarized in tables 4 1 and 4 2 Table 4 reflects results for dominant and non dominant side data separately whereas table 4 2 reflects the analysis results for the pooled data sets Note that the five GTAs which were studied produce collectively 13 different measures These are categorized as primary secondary and exploratory Primary measures represent the result that Is of most interest and possesses the characteristics in the context of the respective test of a true performance measurement 1 a larger numerical value would reflect a better test result In tasks used as test tasks that are unidimensional with respect to performance there is usually only one measure and it 1s the primary measure Secondary measures generally reflect a particular aspect of performance in a test task where perform
109. rithm GTA The fundamental design and operation of algorithms and the associated test administration procedures have been rigorously analyzed and tested in a previous study Sriwatanapongse 2002 These GTAs are largely based on the corresponding tests In the laboratory based Human Performance Capacity Measurement System Kondraske 1990 The specifications for algorithm constituting each GTA are described In more detail in an Human Performance Institute technical report Kondraske 2002 62 Table 3 5 Description of Generic Test Algorithms GTA ID GTA Name GTA Description GTA 1 Isometric Measurement of the force production capacity of muscles Strength in an isometric test Measures the capacity of a muscle or Test muscle groups to develop tension Two sub modes grip and resistance and both use the same algorithm GTA 2 Simple Visual This test measures the speed at which a response to a Response simple visual stimulus 1 low information content can Speed be generated centrally to produce a simple upper extremity Test motor action The focus is on speed indirectly measured as the time required by the subject to detect the visual stimulus and generate the response command using a selected upper extremity GTA 4 Rapid The Rapid Alternating Movement test measures the Alternating capacity for speed of movement in a simple reciprocal Movement motion task The alternating nature of the task stresses Test neuromotor
110. rovide a description of this GTA and associated procedures The test procedure begins with a single beep The subject then begins tapping as fast and consistently as possible at any one touch sensor group either left or right The tapping must be with the index finger only with motion restricted to the metacarpophalangeal joint knuckle No wrist or elbow movement is allowed The acquisition from the touch sensor group does not begin till the finger remains on the sensor s for three consecutive 10 ms samples reflective of the first true tap When this three sample check is satisfied the remaining acquired samples at 10 ms intervals are stored in external memory This sampling occurs for a period of 11 seconds During this period whenever there is a change in the sensor status such as when the finger is placed on or lifted off the sensors a 3 second timer is started or restarted If there is no change in the sensor states before the 3 seconds have elapsed a false test start 1s detected 1 e continuous tapping 15 not taking place and the test is restarted The acquired samples are then analyzed to compute the number of taps the tap duty cycle for each tap and then the mean and standard deviation of the tap duty cycles 66 Within a single finger tap if the finger is on the sensor for n samples and off the sensor for z samples then the duty cycle for that tap is Tap Duty Cycle n n z This duty cycle value is calcu
111. s operational scenarios They are not currently used but Chapter 5 discusses possible uses for them 38 Touch Sensor Performance Requirements The HPMM version 4 0 touch sensor subsystem will be utilized for conducting performance capacity tests such as visual response speed finger tapping performance and upper extremity coordination These applications and others that are currently being researched upon require the touch sensors satisfy certain performance characteristics These can be stated as follows e The ability to detect a touch of a human finger and to function as an on off switch in response to this touch e Fast response time preferably less than 1 ms e Automatic calibration capability e Good immunity to electromagnetic interference and cross talk in a multi sensor arrangement e The ability to work effectively even with small footprint sense electrodes Touch Sensor Parameters The QT310 touch sensor IC is a charge transfer based capacitive touch sensor IC It employs bursts of charge transfer cycles to acquire its signal The capacitor C which is connected between the two SENSE pins is a sampling capacitor that forms a floating store of accumulated charge which is switched between the SENSE pins The electrode which is connected to one of the sense pins is thus used to periodically project a sense field and hence the capacitance associated with the electrode will increase when a finger is brought close to it This in
112. s lie close to the noise level and cannot therefore be measured with the same accuracy It is recommended that an auto ranging capability be explored for these signals when used in the context of steadiness measurement The version 4 hardware architecture allows for this type of function which can be implemented by digitizing 110 both the DC coupled low gain and AC coupled high gain signal paths in the acceleration and angular rate sensor subsystems during the course of a timed steadiness test Thus the gain for the AC coupled path could be increased over its present value to improve resolution of small signals and the threat of saturation can be avoided Processing can use the low gain DC coupled channel for large signals 1 e whenever saturation is approached on the high gain AC coupled channel The fact that the DC coupled channels are not high pass filtered 1s not as critical for large signals as DC errors contributed would be relatively small when expressed as percentage of typical signal amplitudes It 15 necessary to preserve DC coupled channels from each of these sensors in order to support other tests planned for HPMM incorporation Another possible improvement with regard to the acceleration and angular rate measurement subsystems could be made with regard to calibration While these sensors are presently being calibrated only for offset it is recommended that the calibrations be performed for sensitivity gain as well This wil
113. s were designed prototyped and bench evaluated Two successive total system designs have been fleshed out partially implemented and tested by students of the Electrical Engineering Senior Capstone Design classes taught by Dr Kondraske at the University of Texas at Arlington in 2000 This resulted in a preliminary realization of what was termed version 3 of the HPMM hardware platform As part of a subsequent thesis Sriwatanapongse 2002 a set of software algorithms covering a subset of the desired measurement functionality were developed and tested with version 3 of the HPMM A set of five so called generic test algorithms were implemented and tested experimentally with 18 subjects for reliability and validity Generally good reliability was obtained Moreover most results compared favorably with lab based instruments This work included considerations and development of a basic HPMM operating system as well Version 3 was still very much a bench top realization it was not yet packaged in the form of a portable multi meter In another Electrical Engineering Capstone Design class taught by Kondraske during the fall of 2002 student teams developed a next generation hardware platform for the HPMM This included a more advanced processor important changes to packaging and architecture A display with greater capacity and the incorporation of touch screen technology and many other changes aimed at improving the basic platform to support p
114. screen with the button overlay and the special RSM port The bottom view shows the two high intensity LEDs and the touch sensors which are capacitive and have a much faster response than the touch screen buttons The HPMM will also have a special rubber pad located on the bottom of the unit This pad will be placed on the body part for force resistance test quantitative manual muscle test as shown in figure 2 5 The purpose of this pad is to alleviate discomfort to the subject s body during testing In this test the administrator applies force through the handle positioned on the top of the HPMM The subject will try to maximally resist the force applied while the administrator steadily increases the force until the subject can no longer resist 21 The force sensor in the handle measures this resistance to the applied force and the processor will determine the maximum force resistance as the strength capacity result RESISTANCE Figure 2 5 Resistance test Quantitative manual muscle test 2 5 Generic Performance Capacity Tests In the General Purpose Test Mode the unit is intended to be used as a stand alone general purpose device that 1 capable of performing sets of generic tests that are accessed at random For example a neurologist could use the device while making rounds and decide on the spot that a particular measurement is of interest One of the generic tests e g isometric strength capacity wh
115. sion 1 Figure 4 14 Upper Extremity Coordination Accuracy scores for two test sessions 99 4 4 6 Hand Arm Steadiness The results of this test include four constituent components and one composite measure as discussed in section 4 2 The four constituent components represent 4 degrees of freedom with regard to the motion of the body part involved The final one which is obtained by multiplying the first four together is called the 4 degree of freedom 4 DOF steadiness Before the work done in this thesis the HPMM incorporated only translational acceleration as the basis for steadiness measurement Hence no direct retrospective comparisons can be made Proceeding in this fashion There is a good agreement between the translational steadiness values obtained by the present and previous versions of the HPMM figure 4 15 However the laboratory based measurements were primarily displacement oriented detailed treatment on the conversion of the acceleration based results to a displacement based form has been provided in Sriwatanapongse 2002 The author there found close agreement between the results obtained using the two types of measures Hence one can state by inference that the present version of the HPMM has provided results that are also in close agreement with the classic laboratory based instruments results It is noted that these measurements involve very small motions in healthy subjects that approach the current noise flo
116. sources are intermittently returned to the OS routines and then transferred back to the GTA for repetition This mechanism provides a clear resource allocation protocol and avoids resource conflicts at any given time Calibration This function involves calibration of sensors such as the accelerometer rotational rate sensors and the force sensor Calibration is performed once at startup and all three sensor subsystems are calibrated only for offset not for gain sensitivity The resulting sensor offsets are stored in system ROM for later use during the algorithm execution While the calibration procedure is executed no user inputs from the touch screen are accepted The calibration lasts for approximately 5 seconds The sensor 61 subsystems that are being calibrated are individually enabled during this procedure and disabled immediately after This ensures that no power is wasted and provides for longer battery life It may be noted that the remote sensor module RSM must be plugged in before startup and placed steady on a flat surface 1n order that the sensors contained therein might be calibrated correctly 3 2 2 Generic Test Algorithms A major objective of this thesis is to review enhance as necessary implement and evaluate the same set of generic tests that were implemented in version 3 0 of the HPMM The software routine that clearly and precisely defines the implementation of a given test Is referred to as a generic test algo
117. ssues of measurement quality such as reliability validity age and gender effects and subject motivation Their laboratory was applied exclusively as a research tool primarily in clinical trials of new drugs aimed at progressive neurologic diseases such as Parkinson s disease and Multiple Sclerosis At the University of Texas at Arlington a first generation computer based system was developed by Kondraske as a basis of his dissertation research Kondraske 1982 A new set of specially designed instruments were incorporated which were capable of implementing modified versions of test items in the neuro function laboratory as well as new items which added to the scope of the system The characterization of individual subjects in contrast to group study research applications was emphasized as the primary need upon which research and design were focused Attention was given to items which could be viewed as being application independent or those items which reflected more intrinsic characteristics e g strength speed et cetera of human subsystems This is in contrast with approaches that focus on performance of the individual in relatively complex higher level tasks such as gait or activities of daily living Kondraske has argued Kondraske 1990 Kondraske 2000b that there is a finite albeit large set of the more intrinsic characteristics associated with a fairly well defined set of subsystems as opposed to the infinite variety of the hi
118. st 76 Poirt to fist X Find averaze of all X axis axis sample absolute differences Compute 31 1200 ave And convert to BCD Conte ample EX and store Convert BCD result to LCD characters and stow Find averaze of all Y axi absolute differences all A axis Samples Compute 501 1200 avg And convert ta BCD Point to frst Y axis sample Convert result to LCD characters and stow Compite sample Y of ef Dis play results on LCD a Are all Y axis S amples Figure 3 11 Flowchart part 3 for 6 Hand Arm Steadiness Test 11 3 3 Support Subroutines The support subroutines are those that are called by every GTA They include user interface drivers communication routines mathematical subroutines and delay routines The user interface drivers deal with three components of the user interface namely the touch screen the LCD and the beeper The simplest of the three 1s of course the beeper and there is only one routine that 1 associated with it One may take note of the fact the beeper circuit gets its input from the microcontroller DAC which 1s currently a square wave Thus the business of the beeper routine is to use the DAC to produce this square wave The touch screen has an elaborate initialization routine followed by a routine to scan for a touch and routines that convert the coordinates to button numbers either
119. st as you can while making no errors That is the goal for this test e You will hear a beep when its okay to start tapping and the test will begin with your first tap It will last 10 seconds and a double beep will signal the end of the test Remember that both your speed and your accuracy will go into calculating your score So do your best INSTRUCTIONS TO EXAMINER e Flip the HPMM over and show the subject the target and error regions e Position the subject s chair and test module so that the shoulder of the arm involved is centered on the module Thus when the right arm is used the module center will be slightly to the right of body center and vice versa e After selecting READY a single beep will occur and you may ask the subject to start tapping between A long beep will ensue if for some reason the subject stops tapping for longer than 3 seconds The test will automatically restart e Since this test involves a trade off between speed and accuracy the best score is obtained when the subject is going fast enough to make a few errors Therefore the subject s final score should show somewhere between a 5 25 error factor 75 9596 accuracy e Some impaired subjects may not be able to obtain error rates this low You should watch the subject and scores from early trials and try to achieve the best score by encouraging a slow subject to speed up and by reminding a more careless subject to be more accurate 119 GTA 6 HAN
120. sted Reitan battery Vega and Parson 1967 In neurology the finger tapping test is representative of a more general class of tests that have come to be known as rapid alternating movement tests Kondraske and colleagues Kondraske 1990 have adopted the Halstead paradigm for evaluation of performance involving other body segments isolating reciprocal motions about the wrist elbow shoulder and ankle The original finger tapping test has been shown to be a sensitive indicator in Parkinson s disease Potvin et al 1985 which is characterized in part by bradykinesia slowness of movement stoke Prigatano and Wong 1997 and many other clinical situations 25 The subject under test 1s instructed to tap their index finger as fast as possible for a prescribed time e g 10 seconds Historically perhaps because of limitations in instrumentation and computing only speed in taps sec has been measured Behbehani and Kondraske 1986 explored extracting additional parameters such as time up and time down from data acquired with electronic touch sensors The most recent version of the HPMM that involved formal human studies Sriwatanapongse 2002 incorporated a preliminary version of the rapid alternating movement test that measured speed duty cycle and consistency variability of duty cycle That prototype utilized touch sensor designs from lab based systems These were rather complex custom designed circuits More recently special
121. t by using capacitive touch sensors In addition to the fundamental hardware platform an initial set of generic tests to be incorporated into the HPMM and their specifications has been identified Sriwatanapongse 2002 These specifications are routinely reviewed and updated Kondraske 20053 Each of these generic tests draws upon different subsystems and is supported by a corresponding Generic Test Algorithm GTA In version 3 five different generic tests were implemented and evaluated Sriwatanapongse 2002 The same generic tests with selected enhancements are now of interest 1n the context of the version 4 hardware platform These are summarized in Table 2 2 along with the identification of the HPMM functional units included in version 4 hardware that are engaged in the implementation of each generic test 18 Procedures characterizing how the basic functional units are to be employed to achieve the desired performance measurement capability have been closely modeled according to those used for existing well established laboratory based tests These are described in a separate Human Performance Institute technical document Kondraske 2005a Briefly each test consists of a test task which is designed to isolate to the degree possible a given system at a given hierarchical level 1 e basic element or generic intermediate level of the Elemental Resource Model Kondraske 1995 and maximally stress that system along one or more dim
122. t how to take their finger off the home sensor using their entire arm The motion must be about the elbow and not the knuckle or wrist Cancel any trial in which the subject fails to do this properly Allow the subject to practice the movement and start the test only when you are confident of their correctness When the subject is positioned properly select READY A beep will sound only when the HPMM detects the subject s finger on the sensor If the subject tries to guess the turn on time of the lights and wrongly lifts the finger an anticipation error is detected and the lights will flash repeatedly Ask the subject to replace the finger on the home sensor region to redo the trial 117 GTA 4 RAPID ALTERNATING MOVEMENT INSTRUCTIONS TO SUBJECT This test will examine how fast you can move your index finger about your knuckle Place your hand here and first make a fist then extend only your index finger like this Using the pad of your index finger you must tap within this square region point region outline to subject Your finger may fall anywhere within the region Lift your finger about 1 inch from the plate during each tap A single beep will begin the test but the timing will begin after your first tap A double beep will signal the end of the test Ensure that your movement is about your knuckle only Do not curl your fingers or flex your wrist INSTRUCTIONS TO EXAMINER First flip the HPMM to show the subject the requi
123. t recent phase in the development and continued evaluation of a portable packaged and functional fourth generation prototype of the HPMM Selected performance capacity tests implemented on this platform to gauge reliability and validity of each The following objectives of the thesis which were presented in chapter 1 have been achieved 1 Studied the previous version of the HPMM and relevant laboratory based instruments from architectural and functional perspectives Studied the documents that described the present version of the HPMM and made improvements to them 2 Modified the earlier version of the HPMM operating system software to create a new version that runs on the current HPMM hardware platform and included routines that drive the touch screen and LCD 3 Verified the performance of new sensor subsystems such as accelerometer angular rate sensors and touch sensors The new force sensor subsystem was packaged into the HPMM handle 107 4 5 6 T Proposed a method of combining the four secondary results of the steadiness test into one primary result called the four degree of freedom 4 DOF steadiness Reviewed and implemented the five GTAs namely isometric grip strength simple visual response speed rapid alternating movement upper extremity coordination and hand arm steadiness in strict conformance to the procedures dictated by the HPI internal documentation Evaluated the performance of the HPMM in its intend
124. test and measurement protocols that will yield a prototype of the HPMM that can be deployed in clinical applications Specifically beginning with the earlier work Kondraske 2001 Kondraske 2002 on the HPMM the area that will be addressed 1s the upgrade of hardware and software software being the majority area onto a new system that yields faster results is more portable and conforms to newer technology standards The objectives of this thesis are to 1 Review the architecture and functionality of previously designed instruments that were used in the same or similar areas Analyze and enhance documentation of the design of the present version of the HPMM strengthening the characterization of its various subsystems 2 Adapt the current basic HPMM operating system software to the new hardware platform incorporating a touch screen user interface and verify operation to support use in its Generic Test Mode 3 Test and verifying performance of new hardware subsystems including high speed capacitive touch sensors multi axis inertial sensors and the integration of a force sensor into the HPMM main unit packaging 4 Propose a preliminary approach for the integration of dual axis accelerometer measurements and dual axis angular rate measurements to form a composite steadiness measure 5 Review revise as necessary and implement test algorithms on the version 4 platform f for a selected subset of five HPMM generic tests isometric grip st
125. tform in final packaged portable functional prototype of the HPMM and with minor exceptions the current set of measurements have been demonstrated to exhibit good reliability and encouraging validity 109 5 2 Recommendations for Future Research There are a good number of positive results that have been obtained for this realization of the HPMM However further improvements may be considered The following is a discussion of those issues which emphasize the system in general and issues that pertain to the five generic tests studied Discussion of completely different generic tests is beyond the current scope The gain of the accelerometer and angular rate subsystems requires further investigation in order realize optimal performance with the widest possible range of subjects When these are used as components of steadiness measurement an increased gain would permit small accelerations and angular rates to be better resolved The present gains were optimized to avoid saturation for very large pathologic tremors while still attempting to provide a reasonable ability to resolve small motions such as those present normally in healthy subjects In addition the use of these sensor channels for other tests which impose different gain requirements was also considered Currently signals from pathologic subjects such as those who exhibit symptoms like tremor lie far above the system noise floor However small inputs typical of healthy subject
126. ther to obtain the single number result called 4 DOF steadiness Averaging the four results 1s not a conceptually sound solution because they do not all have the same units and dimensions This approach also allows the dimensionality of all four quantities to be preserved the composite score Translational Acceleration based Steadiness The translational motion 1n the hand is measured in terms of acceleration We begin with displacement The motion is modeled as a sinusoidal one that can be described mathematically as 71 s t A sin 2z t m where imeters 15 the amplitude of the displacement with frequency f In Hz The corresponding acceleration is obtained by differentiating this equation twice a t 4x sin 2nft m s Further the units of acceleration are converted to g s For example a displacement of 1 cm peak leads to an acceleration of roughly 1g and a displacement of 0 1 mm corresponds to an acceleration of 0 01 g While it is difficult to accurately estimate what the values are of the smallest and largest accelerations that can be produced in the human hand the range described above 1 0 1 to 10 mm is assumed to be an adequate estimate These values were used to determine overall gains and measurement range in the hardware design Inverting these limits gives 1 1 g and 100 1 g and these are the units one by g that will go into the steadiness measurement model These numbers decrease with inc
127. touch sensor This becomes an important issue when dealing with pathologic subjects who may have very slow responses to visual stimuli Also given the procedures of conduction of the various tests one may safely assume that the examiner will not allow any stray object to present itself in close proximity to the touch sensors 7 Polarity OUTP The polarity of the output can be set to active high but is active low by default The main advantage of the choice of an active low output polarity is that it complements the use of pull up resistors at the sensor outputs The current drain at the output only occurs when detection occurs so that the power consumption 1 much lower In the active high case where the output would have to be pulled down during all such times when there was no detection much higher current drain can be expected 44 8 Toggle Mode TOG This mode gives the output a touch on touch off flip flop action and hence the output changes state on detection 9 Toggle Latch Mode TOGL OUT becomes active on a detection but will become inactive only when a logic low pulse is applied on the CAL_CLR pin 10 Heartbeat Output HB Heartbeat indicator pulses are superimposed on the output to indicate the health of the IC They can be removed by using a capacitor at the output pin or if SC 0 by setting HB 1 Since the process of touch detection in the HPMM is not a continuous one the touch sensors are not ena
128. trix showing the assignment of different HPMM functional units to different HPMM functions and those generic tests included in version 4 HPMM General Isometric Strength Visual Response Rapid Alternating Neuromotor Steadiness Functional Unit User GTA 1 Speed Movement Channel Capacity GTA 6 Interface GTA 2 Prompts Prompts Results Results Test Progression Test Progression EX j 2 0 PX IX X X X Visual Display Graphics Visual Display LEDs conditioning Touch Sensor Array Simple Sound Generator Prompting Prompting Prompting Prompting Prompting LI Force Sensor o The user interface consists of a low power LCD screen for displaying menus and test results and a touch screen consisting of four buttons for sending commands to the HPMM Various stimulus units are also included The LEDs vibration generator and sound generator provide visual vibration and audio stimulus respectively Accelerometer and angular rate sensor are used to measure various parameters of human movement see table 2 2 These sensor subsystems will be placed on the tested body part of the subject This requires the sensors to be as small and light as possible to have neglect effects on the test results The touch sensors are used for sensing finger contact These sensors must be highly sensitive for accurate measurement of finger movement on the sensors The speed requirement is me
129. ts and Other Specifications Visual Display Graphic Included LCD low power medium resolution For user interface and warning visual feedback mechanisms as part of selected tests Visual Display LEDs Included High intensity high speed 2000 mcd red 0 75 in diameter Two units that can be independently controlled Force Sensor Included 0 1120 Newtons 1 N resolution tension and compression integrate into main unit housing design Microphone with signal Included High sensitivity directional conditioning RSM 1 Included At least 2 translational degrees of freedom desired RSM 1 Measurement range uncertain Angular Rate Sensor s Included Inertially based low drift micro miniature RSM 1 Initially only one degree of freedom was incorporated Expanded to two degrees of freedom for version 4 to support steadiness measurement Vibration Generator Not Piezoelectric or electromagnetic included Touch Sensor Array Included High speed capacitive Five touch sensitive regions up to and including version 3 Expanded to nine independent sensor regions to improve measurement capability and also enhance user interface Sensor included Piezoelectric with ability to control pitch to provide different feedback beeps to subject and examiner Acoustic Stimulus Not Support auditory screening and cognitive Generator Included performance tests high degree of calibration accuracy 3dB 16 Table 2 2 Ma
130. uently changed from neuro function laboratory to sensori motor performance laboratory and ultimately to human performance New issues were now to be explored given that the basic toolbox had now grown in power The Big Picture context was analyzed leading to questions such as what these measurements meant in larger scenarios such as an individual s ability to memorize an identification number or live alone independently The issue that was now discovered was that despite efforts to quantitatively characterize capacities there was still a lack of a conceptual model in this field General Systems Performance Theory and the Elemental Resource Model Kondraske 1995 Kondraske 2000b were introduced to address this and a number of related issues in human performance The Elemental Resource Model is a hierarchically organized model based on a small but robust set of systems performance constructs was introduced to address the need for a broad unifying understanding of the human system and its relationship to tasks These conceptual developments led to subtle but 1mportant transformation of several major classes of measures in the HPCMS as well as a clearer definition of performance measures and the protocols under which they were acquired Sriwatanapongse 2002 1 2 The Human Performance Multimeter Recent advancements in sensor low powered electronic microprocessor and battery technology motivated the vision of a device that
131. ues obtained there 1s a commercial device Motus tremor measurement system that does utilize an angular rate sensor for similar measurements This Motus unit only incorporates a single degree of freedom sensor whereas the HPMM version 4 senses two angular DOFs as described above Thus two separate tests must be done with the Motus sensor mounted differently each time to evaluate both DOFs of interest The opportunity to access one of these systems provided the basis for a preliminary crude evaluation of the range of values that were obtained in the present study Data from the Motus system for the dominant hand of a single healthy male subject showed Kondraske 2005 the equivalent rotational steadiness for wrist flexion extension of 1 27 s deg compared to 1 19 for the HPMM and for forearm pronation supination of 0 93 s deg compared to 0 83 s deg For both the HPMM and Motus 103 data steadiness was better for wrist flexion extension than for forearm pronation supination motion Finally it is desirable to have one single number that can represent the steadiness of a subject rather than four numbers This necessity motivated the development of 4 DOF steadiness according to the rules of the General Systems Performance Theory where the dimensions of the four constituent quantities are preserved This is accomplished by multiplying the four quantities together The resulting 4 DOF steadiness numbers have been observed to possess a v
132. ulty ensures that the contribution of change in instrument dimensions is adequately factored in so that the subject s score will be primarily dependent on their speed and accuracy The mean scores for both dominant and non dominant sides are about 10 15 higher than those obtained with HPMM version 3 Sriwatanapongse 2002 The touch sensor subsystem has been completely 977 redesigned in version 4 see Chapter 3 and is more likely to provide more accurate registration of tapping Also as noted previously the expression used to compute NMCC from speed and accuracy measurements is also slightly different Scatter plots are also included for the secondary measures of speed figure 4 13 and accuracy figure 4 14 However since speed and accuracy may trade off it is neither necessarily expected nor desirable that these values remain the same on retest Upper Extremity Coordination NMCC bits s C 9 TN 10 15 Test Session 1 Figure 4 12 Scatter plot for NMCC measurements between the two test sessions 96 Upper Extremity Coordination Speed cmis Test Session 2 gt D Side Side 0 00 0 00 10 00 20 00 30 00 40 00 50 00 50 00 70 00 Test Session 1 Figure 4 13 Upper Extremity Coordination Speed scores for sessions 1 and 2 Upper Extremity Coordination Accuracy 70 C 7 D 9 Ell in D D Side a N Side 40 0 50 0 Test Ses
133. vision 16 bit by 24 bit multiplications and conversion from binary to packed BCD The last set of support subroutines consists of various small useful delay routines and one very important timer O interrupt service routine which is in fact located at OOOBH in code memory 79 3 4 Memory Utilization Plan As mentioned earlier the HPMM microcontroller has 64 KB of code memory and 4KB of on chip external memory These resources have been utilized to store the HPMM software code and data elements and an organizational description is best provided by a memory map The following table shows the memory map for the current version of the HPMM software Table 3 7 Software Memory Map E E m Address Address Contents Ge e Mathematical and Timer Routines External On Chip Memory Volatile Test results buffer read by LCD code 80 CHAPTER 4 EXPERIMENTAL EVALUATION 4 1 Overview and Objectives Reported in this chapter are the results of experiments conducted to investigate the performance of the current HPMM prototype To investigate the reliability of the set of measures implemented a test retest repeated measures design with two sessions was used This protocol also provides data that is useful for gaining insight into validity of measures The methods are similar to those incorporated in studies of earlier laboratory based instruments and earlier versions of the HPMM 4 2 Methods Twenty normal adults volunteered
134. y individuals and the inter subject variability for this measure is quite low relative to the range over which this parameter can vary across all subject types of interest 96 4 4 5 Upper Extremity Coordination Repeatability for the primary measure in this test 1 neuromotor channel capacity NMCC was the weakest of all measures studied This finding was also similar to that found during version 3 evaluations version 3 vs version 4 r 2 0 36 vs r 0 34 for the dominant side and 0 6 vs r 0 39 for the non dominant side This again may be due to narrow distribution of this performance in the subject pool as discussed elsewhere Supporting this the reliability increased r 2 0 6 when dominant and non dominant side data was pooled to create a larger measurement range In a clinical context where pathologic patients will be involved the ability to discriminate healthy and poor performance is thus likely to be rather good There is also one outlier data point produced by a subject who exhibited an excessively high score for NMCC during Session 2 This score is currently not explainable and may be due to a subtle data dependent software error A higher Pearson coefficient r 0 5 was obtained when the analysis was performed excluding this subject s reading Other issues related to NMCC reliability are discussed in Chapter 5 The NMCC values are in good agreement with those obtained from previous studies The index of task diffic

Download Pdf Manuals

Related Search

umi uta 1162

Related Contents

inspiro Classic Gebrauchsanweisung  SOFTWARE USER MANUAL  取扱説明書（PDF - Bluebanana99  高性能ファンコイルユニット    Handleiding  VITA vPad easy  

Copyright © All rights reserved. Failed to retrieve file