Recurring Meeting of Cleveland Clinic Core Team

Date: August 11, 2015

Time: 10:30 AM EST

Means: In person meeting

Attendees:

1. Ahmet Erdemir
2. Craig Bennetts
3. Snehal Chokhandre
4. Robb Colbrunn
5. Tara Bonner
6. Omar Gad
7. Raghav Malik
8. Elizabeth Bond

Agenda:

1. Discuss immediate action items from the last meeting.
2. Overview of student contributions to source code repository and wiki.
3. Recap on tissue testing.
4. Decide immediate action items for the next meeting.
5. Other.

Immediate Action Items:

• Ahmet
  • Dissemination of joint mechanics testing data.
  • Prepare flowcharts to describe data flow for pre-processing of joint kinematics/kinetics data.
  • Complete August 4, 2015 meeting minutes.
• Craig
  • Continue learning IA-FEMesh to explore different meshing strategies.
  • Add stl files of Open Knee(s) - Generation 1 geometry to the repository.

• Craig & Robb

  • Meet to resolve issues related to processing of joint coordinate system transformation matrices.
• Snehal
  • Contact specimen supplier to follow up with acquisition of remaining specimens.
  • Contact additional specimen suppliers to facilitate acquisition of remaining specimens.
• Raghav
  • Quantify the repeatability of image based thickness measurement system.
  • Update code following based on the feedback from demonstration.
• Omar
  • Focus on segmentation of menisci and ligaments for oks001.
  • Meet with Ahmet to prepare for NSF REU Symposium.
  • Prepare a presentation on his segmentation efforts.

Ongoing Action Items:

• Ahmet & Craig

  • Incorporate model changes from summer students to Open Knee(s) - Generation 1 model for future release.
• Craig
  • Anatomical imaging manuscript.
  • Large-scale analysis of simulation parameters for convergence.
• Snehal
  • Continue interacting up with Biomomentum, Inc. about mechanical testing system tuning features.
  • Complete the additional structured testing plan for tuning the system for different sample sizes, strain levels and strain rates.
  • Pivot shift manuscript.

  • Cruciate ligament modeling & simulation.

  • Evaluate alternative tissue testing strategy.
• Robb
  • Patellofemoral joint experimentation manuscript.
• Tara
  • Tibiofemoral joint experimentation manuscript.

Notes:

1. Discuss immediate action items from the last meeting.
   • Ahmet wanted to disseminate joint mechanical testing data for oks001. He downloaded the folder as a zip file from MIDAS (in-house data management system). MIDAS automatically deletes empty subfolders when compressing the main folder. The file was 770 MB. Ahmet needs to decide to potentially break up mechanical testing data for dissemination. Robb noted that the index files can be deleted as these were generated by the operating system to facilitate search. This can half the size of the download package. Ahmet wants to keep the empty folders to provide the time history of joint testing. Robb recommended to put a readme.txt in empty folders as a place holder. Ahmet still needs to prepare the flowchart for joint mechanics data processing.

   • Craig described the meshing when using IA-FEMesh. He uploaded STL version of the Open Knee(s) - Generation 1 model geometries to the repository. He will resend Ahmet August 4, 2015 meeting notes as the previous version was blank. He will meet with Robb to resolve issues related to implementation of Grood & Suntay convention.

   • See agenda item on tissue testing for Snehal's updates. Snehal will contact other specimen providers to expedite acquisition of specimens from younger donors.

   • Robb uploaded the gauge R&R document on the wiki.

   • Raghav demonstrated the camera-based thickness measurement system. He discussed the use of RGB color ratios to perform boundary detection.
   • Omar finished segmentation of cartilage and bones for oks001. The group asked him to work on ligaments and menisci as well to see the performance of semi-automated approach. Omar presented a poster and gave a talk as part of the NSF REU program. He was nominated by the Department of Biomedical Engineering to represent Cleveland Clinic in the upcoming symposium of NSF REU students in DC area. Omar will discuss with Ahmet future directions to prepare for this meeting.
2. Overview of student contributions to source code repository and wiki.
   • The group went over repository areas where student work has been accumulated.
   • The scripts are in relevant utilities folders.
   • Segmentation results were stored under "segmentors" subfolder under oks folder, with each segmentor assigned to a numerical folder described in the readme.txt file: 01 (Raghav), 02 (Connor), 03 (Diana), 04 (Omar). Meshlab can be used to compare different segmentations of the same tissue.
   • Changes in Open Knee(s) - Generation 1 model are placed under _gen1/src folder with descriptive folder names indicating modifications done by the students.
3. Recap on tissue testing.

   • Snehal went over the recent tests she had conducted. A total of 20 tests were performed including all tissue and test types. Strain rates were 20%/s & 10%/s. When testing MCL, issues related to the tuning of the machine were raised indicating that anything above 20%/s strain rate may not work with current settings.

   • There were discrepancies in gauge length measurements of tensile specimens. To get the reference length, the clamps were aligned next to each other without the sample (zero position). This may introduce errors as it is a visual alignment albeit done using the camera. After placement of the sample, the displacement was measured using "find contact" feature of the testing system with 10 gram load. This value provides the gauge length. These two sources of error need to be quantified more elaborately. For the first one, the resolution need to be found. For the second, the role of force filtering on the repeatability of displacement at the same force needs to be assessed. Snehal will test out the use of force filtering to see its effect on determined gauge length.
   • There was a force overshoot during stress relaxation tests when displacement was constant. This seems to be a temporal misalignment between displacement and force data. If the force data were collected in a filtered form, a phase shift would be expected and it will explain this issue. If not, Snehal needs to contact Martin Garon (from Biomomentum Inc.) to resolve this.
   • Acceleration/deceleration phases seem abrupt. To reach desired velocities these phases need to be minimized. The average and instantaneous velocities during the ramp phase should be compared against the desired velocities. It should be noted that acceleration settings were default of those after control software was updated by Martin Garon.
4. Decide immediate action items for the next meeting.
   • See Immediate Action Items above.
5. Other.
   • Omar wanted to go through growth cut method of segmentation with the group. He will schedule a presentation within this week.
   • Omar's last day is this Friday (August, 14, 2015), Raghav's is next Tuesday (August 18, 2015).