Recurring Meeting of Cleveland Clinic Core Team

Date: September 2, 2014

Time: 10:30 AM EST

Means: In person meeting


  1. Ahmet Erdemir
  2. Craig Bennetts
  3. Robb Colbrunn
  4. Tara Bonner


  1. Discuss immediate action items from the last meeting.
  2. Review of oks004 joint testing notes and data.
  3. Decide immediate action items for the next meeting.
  4. Other.

Immediate Action Items:


  1. Discuss immediate action items from the last meeting.
    • In the in-house Midas platform, Robb provided a file summarizing file types for joint mechanical testing data. This accommodates joint mechanics data of oks004. It may be useful to automatically copy this file in the root folder of the data during experimentation.
    • Robb recalibrated the pressure sensor used for the oks004. The sensor exhibits nonlinear behaviour, the rate of change in sensor output decreases as forces get higher. He provided a sanity check by comparing forces applied during the calibration against forces predicted by the sensor (with old and new calibration). Robb noted that the pressure sensor may have been taped during the first calibration. Jason and Katie may provide detailed information in that regard. Post-experiment calibration file needs to be uploaded to oks004 data folder in Midas. Samples of pressure distributions measured during patellofemoral joint will be evaluated by comparing sensor output using pre- and post-experiment calibration information.
    • Tara evaluated the joint coordinate system alignment, see relevant discussions in the following agenda item.
    • Separation of MR images and committing relevant code to Open Knee(s) repository got delayed. Craig will take care of these this week.
    • Snehal was out of town, her action items were moved to the following week.
    • Ahmet conducted various tests using the Biomomentum, Inc. system. The experiments included three stress-relaxation tests divided into separate ramp (move relative) and hold (wait) commands to ensure video data is collected in separate folders. A foam sample was compressed while collecting video data (at 10 Hz in raw or bmp format). An unloading step finalized the test. It was possible to separate video data for ramp and hold commands in different folders. This may help temporal downsampling of video, particularly for hold cycles. The video data appeared to be in sync with the force-displacement data. Mach-1 system was observed to wait for brief moments before and after the actual command was executed. For example, a 0.5 mm ramp at 1 mm/s resulted in data with a duration of 0.72 s, while the actual ramp appeared to be at an expected loading rate but with a 0.06 s wait before and after execution. It was also possible to downsample data as it was collected, i.e. a test enforcing data collection at every 1 second (rather than 100 Hz for force data and 10 Hz for video data) was successful. It is advised that stress-relaxation tests are conducted in this manner, separate ramp and hold cycles with data reduction during the hold cycles. Unfortunately, not all experiments completed successfully. The video collection seemed to crash the system, potentially due to buffering issues. This needs to be resolved to ensure seamless data collection. Robb will take a look at this issue and talk to Martin Garon from Biomomentum Inc to resolve.
  2. Review of oks004 joint testing notes and data.
    • Tara provided data to evaluate measured joint kinematics-kinetics response for oks004. Ideally, the forces/moments and the translations/rotations would be looked at the same coordinate system, e.g. tibia coordinate system. The kinematics was in optimized anatomical coordinate system, which deviated from the tibia coordinate system mostly for flexion. Forces/moments were in tibia coordinate system. The AP drawer test at 30 and 60 degrees flexion showed unexpected behaviour, i.e. increasing anterior load did not result in increased anterior drawer. Detailed inspection of data revealed that there were residual internal/external rotation moments indicating that the test was not necessarily a pure AP drawer test. For 0 degrees flexion for example, this was not an issue as internal/external rotation moments were negligible (assuming not to influence to joint kinematics). The team needs to be careful not to label tests with such issues as "AP drawer test", may be "AP dominant loading test". This will prevent confusion during physiological interpretation of the data. Robb and Tara will implement control scheme to further minimize off-axis loads during laxity testing to collect objective laxity test data, e.g. for AP drawer. Tara will annotate the oks004 wiki page and add laxity plots to explain this issue. The group will also check combined loading data sets.
    • The team reviewed joint testing protocols of oks004 and decided upon a few specifications changes:
      • Tibia and femur registration markers may need to be manufactured using 3D printing to ensure sphericity and concentricity.
      • Details on post-imaging activities, e.g. data organization, will be provided in the imaging specifications.
      • A post-experiment calibration will be performed on the pressure sensor used for patellofemoral joint testing. This information can be added in the specifications for specimen preparation or in joint testing specifications.
      • Fixation of fibula to tibia was found to be practical for oks004. The activity will be clarified and detailed in the specimen preparation specifications.
      • Tibiofemoral joint testing will be conducted with the tibia assembled in the robot pointing up. This resolves any issues related to gravity compensation of sensors. Joint testing specifications will be changed to reflect this.
    • During experimentation a few problems were noted. These can be accommodated by the BioRobotics Core during the tests:

      • During varus/valgus loading of the tibiofemoral joint at 90 degrees flexion, the control became unstable resulting in the robot halting.
      • During tibiofemoral joint testing at 90 degrees flexion, Optotrak markers on femur were blocked. Kinematics data from the robot can still be used to quantify tibiofemoral joint motion.
      • Assembly of the joint to accommodate quadriceps line of action for patellofemoral joint testing resulted in blockage of Optotrak markers. The tendon actuator may need to be repositioned to tackle this issue.
  3. Decide immediate action items for the next meeting.
    • Action items were discussed and agreed upon. See Immediate Action Items above.
  4. Other.
    • Tara got a bill from specimen supplier for a whole right leg. She will forward it to Snehal.
    • Ahmet has been working on an Orthopaedic Insights article about Open Knee(s). Orthopaedic Insights is a newsletter from the Department of Orthopaedic Surgery at the Cleveland Clinic. It is distributed nationwide.
    • Craig will try to convert Open Knee(s) - Generation 1 to FEBio version 2.0. He will use most of recent version of PreView for this purpose.

RecurringMeetings/2014-09-02 (last edited 2016-05-04 22:09:52 by localhost)