Recurring Meeting of Cleveland Clinic Core Team

Date: January 26, 2016

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. Jason Halloran (Cleveland State University)
7. Will Zaylor (Cleveland State University)

Agenda:

1. Discuss immediate action items from the last meeting.
2. Continue evaluation of drafts for data manuscripts.
3. Discuss tissue mechanical testing protocol evaluation.
4. Decide immediate action items for the next meeting.
5. Other.

Immediate Action Items:

• Ahmet
  • Work on segmentation of MCL of oks001.
• Craig
  • Document manual segmentation in the relevant specifications page.
  • Expand the documentation on output of segmented volumes as surfaces in the segmentation specifications page to emphasize raw (non-smoothed) surface output.
  • Continue working on recommendations discussed in the meeting for the imaging manuscript.
  • Explore and document geometry generation workflow.
• Snehal
  • Update tissue testing analysis script to identify zero force displacement (next two weeks).
  • Write a Python script for manual identification of tissue thickness from tissue images (next two weeks).
• Jason
  • Provide the background for the patellofemoral joint testing manuscript.
• Robb
  • Complete the first draft of the patellofemoral joint testing manuscript.
  • Work with Craig to get patellofemoral joint contact pressures analyzed.
• Tara
  • Continue reviewing sections of the patellofemoral joint testing manuscript.
  • Complete tibiofemoral joint testing manuscript.

• Robb, Tara, & Craig

  • Meet to work on overlapping content of data manuscripts.

• Tara & Ahmet

  • Meet to work on AP laxity repeatability data analysis.
• Will
  • Put tdmsParser in Open Knee(s) source code repository.

Ongoing Action Items:

DEADLINE FOR SECOND DRAFT DATA PAPERS IS APRIL 1, 2016

• Ahmet
  • Prepare flowcharts to describe data flow for pre-processing of joint kinematics/kinetics data.
• Craig
  • Anatomical imaging manuscript.
  • Large-scale analysis of simulation parameters for convergence.
• Snehal
  • Add citations and summaries to tissue data papers to the wiki for potential use to model tissue properties.
  • Pivot shift manuscript.

  • Cruciate ligament modeling & simulation.

• Robb
  • Patellofemoral joint experimentation manuscript.
• Tara
  • Tibiofemoral joint experimentation manuscript.

Notes:

1. Discuss immediate action items from the last meeting.
   • Ahmet did not work on the segmentation of MCL of oks001.
   • Craig did not document the manual segmentation. Ahmet mentioned that a few sentences will be sufficient to direct the user to the appropriate tools. Craig mentioned that the segmentation specifications has a section on output of segmented volumes as surfaces. Ahmet noted that explicit information related to raw STL output, i.e. without decimation and smoothing, would be helpful.
   • Craig has been working on geometry generation paths, e.g. smoothing, resampling. In this regard, he has identified strategies to compare surfaces, i.e., raw vs smooth to understand the influence of smoothing and front vs back surfaces of cartilage to map tissue thickness. He also gained experience using Blender for selecting/splitting surfaces. Craig mentioned that using Slicer for exporting the segmented volume as a smoothed surface may provide better outcome as the software may utilize the segmentation information to ensure volume and boundary preservation. In this regard, he will compare smoothing of surfaces with Slicer and with Meshlab.
   • For discussion on immediate action items relate to tissue testing, refer to relevant agenda item below.
   • For discussion on immediate action items relate to data manuscripts, refer to relevant agenda item below.
2. Continue evaluation of drafts for data manuscripts.
   • Snehal heard back from Scientific Data asking about the progress of the manuscripts. Scientific Data is fine with the submission of three manuscripts for joint level data. They recommended that each submission should include the other two as supporting material.
   • Craig has been working on the recommendations on the imaging manuscript. He has not completed this work yet. He also noted that he has not received any feedback from Carl Winalski to provide him Open Knee(s) specimen MRIs for grading.
   • Jason has collated the information to write the background for the patellofemoral joint testing manuscript.
   • Robb provided feedback on the anatomical imaging manuscript. He believes that he can contribute to become a co-author. His work on patellofemoral joint testing manuscript is ongoing. Robb and Craig need to meet about comparing pressure sensor analysis for quantification of patellofemoral joint contact.
   • Craig, Robb, and Tara need to meet to appropriately cross reference the manuscripts.
   • Tara went over various sections of the patellofemoral joint testing manuscript.
   • For the tibiofemoral joint manuscript, the group had discussions on technical validation:
     • Tara noted that she calculated RMS error between desired and actual loads for the laxity tests. Off-axis RMS errors, if close to zero, will indicate that the tests are indeed laxity tests. The group discussed how to present this. One option is to have a table for each specimen - each laxity testing condition - each flexion angle, where a row of maximum errors are reported for forces and moments.
     • Kinematics response of specimens during passive flexion may serve for technical validation. This can be done by having a plot of each degree of freedom as a function of flexion angle, where responses of all specimens are plotted in the same figure.
     • Tara also provided the range of anterior-posterior movement during repeated anterior-posterior laxity tests conducted at 30 degrees flexion. Ahmet noted that while such a presentation provides a quick and accessible summary of the repeatability, it may be deceiving. First, the RMS errors between actual loading profiles of repeated tests should be calculated and provided. This will ensure that the robot applied the same loads and therefore one can expect the same movements. Second, after the first condition is met, the RMS errors between kinematics of repeated tests should be calculated and provided. This will confirm that the knee is not damaged and moves in a similar way under similar loading. Ahmet noted that Omar Gad wrote scripts to for the compare kinematics and kinetics of repeated trials - available in the source code repository under utl/SimVitro folder. Tara will work with Ahmet in this regard.
     • Tara also asked about using registration marker data for technical validation. First, fits to the measurements during joint mechanics testing should result in marker dimensions, e.g. a technical validation done similarly as in the anatomical imaging manuscript. Second, the distances between centers of the markers of a same bone should be similar in the MRI and as measured during joint mechanics testing. This comparison tests if the markers or bone moved/deformed between testing sessions.
3. Discuss tissue mechanical testing protocol evaluation.

   • Snehal updated tissue testing script to obtain an estimate of average modulus from the slopes of 3 force-displacement measurements. She still needs to incorporate zero force displacement extraction into her tissue testing analysis script. In response to Robb, Ahmet noted that the average modulus will solely be used for rough comparisons. For modeling & simulation, a more elaborate constitutive model will be fit to data. Snehal also mentioned that the script provides a comparison of desired and actual loading rates as another level of quality check.

   • Snehal continued working on the Python script for manual identification of tissue thickness. She would like the interface to be similar to the contact based tissue thickness measurement system.
   • Snehal change protocols for tissue testing to separate zero load displacement estimation using "find contact". She expanded the documentation related to tissue testing protocol evaluations. Ahmet recommended that she includes quantitative information in the conclusions in a summary form to support her judgements.
   • Snehal started learning how to use the cryostat. Cryostat works on frozen/hard tissue. For the study, it is used to trim samples to uniform thickness, in particular the ligament and cartilage. It does not allow cutting thick slices (max 150 um). Previously, samples were cut by an employee of the Histology Core. Due to the employee leaving his position, Snehal needed to learn how to do this herself.
   • Snehal heard back from Martin Garon at Biomomentum, Inc. about software freeze during calibration of 150g load cell. Martin has recommended configuring the load cell to factory settings to see if that fixes the issue.
4. Decide immediate action items for the next meeting.
   • See Immediate Action Items above.
5. Other.

   • Will developed a Python suite, tdmsParser, to work with TDMS and configuration files generated by SimVitro. He is using Sphinx for automated documentation. Ahmet noted that Omar Gad and Connor Lough wrote some code for similar reasons. It may be a good idea to incorporate these into Will's code. Will will create a tdmsParser folder in utl/SimVitro folder of the source code repository to put his scripts.

   • Will demonstrated visualization of ligament length changes with rhombodial insertions to define cruciate ligament fibers. He has been using oks001 data for this purpose.
   • Ahmet mentioned that Connor Lough will be returning to the laboratory by the end of May for another summer internship.