Recurring Meeting of the Advisory Board

Date(s): May 19-30, 2014

Due to scheduling conflicts, the group meeting with the Advisory Board members were scheduled on an individual basis with each member. Please refer to Meeting Minutes for more details.

Progress Report

Project Specific Aims

• To provide an open, freely available, and collaborative development, testing, simulation and dissemination platform for in silico exploration of the biomechanics of healthy and diseased knees.
• To develop in silico biomechanical models of healthy and diseased knee joints of different genders and ages, supported by specimen-specific joint and tissue level experimental mechanics.

Project Infrastructure

• https://simtk.org/home/openknee (project website)

• http://wiki.simtk.org/openknee (project wiki)

Project Contact

• Ahmet Erdemir, PhD, erdemira@ccf.org, https://simtk.org/users/aerdemir

Advisory Board

• Clinical. Jack Andrish, MD; Carl Winalski, MD; Morgan Jones, MD, MPH; Paul Saluan, MD (Cleveland Clinic)

• Engineering. Yasin Dhaher, PhD (Northwestern University); Trent Guess, PhD (University of Missouri), Rami Korhonen, PhD (University of Eastern Finland)

Progress Period

• October 28, 2013 - May 19, 2014

Development Team during Progress Period

• Cleveland Clinic Core Team. Craig Bennetts, Dylan Beckler, Tara Bonner, Snehal Chokhandre, Robb Colbrunn, Ahmet Erdemir, Jason Halloran

• Stanford University. Scott Delp, Joy Ku, Henry Kwong

• University of Utah. Ben Ellis, Jeff Weiss, Steve Maas

• Community. Elvis Danso, Shannon Donnola, Chris Flask, Cara Sullivan

See https://simtk.org/project/xml/team.xml?group_id=485 for full details on the project team.

Goals of Progress Period

• Complete a prototype of web based computation infrastructure (Stanford University)
• Complete implementation of high priority features in simulation software (University of Utah)
• Complete experimentation (specifications and data collection) for Specimen 001 (Cleveland Clinic)
• Disseminate data from Specimen 001 (Cleveland Clinic)

• Start development and review of modeling specifications (Cleveland Clinic & Community)

• Start development and review of simulation specifications (Cleveland Clinic & Community)

• Launch community projects program (reviews, internships, and collaborations) (Cleveland Clinic & Community)

• Publish previous work on Open Knee - Generation 1 (Cleveland Clinic)

Please also refer to previous meeting minutes of the Advisory Board at RecurringMeetings/2013-10-28. For more details, see the Roadmap and the proposed timeline of the proposal in GrantInformation/proposal.pdf.

Activities of Progress Period

• Cloud computing prototype

  • Cleveland Clinic and Stanford University teams continued to have monthly conference calls to discuss web based computation interface. Please refer to meeting minutes in Cleveland Clinic - Stanford University section of RecurringMeetings.

  • Detailed specifications on a prototype for web-based computing infrastructure was provided by the Cleveland Clinic team, see Revision as of May 18, 2014 or current version.

  • A prototype of a web-based interface for submission of simulations was developed on a staging server; this was summarized in the progress report submitted to National Institutes of Health along with a snapshot of the interface, see https://simtk.org/websvn/wsvn/openknee/doc/R01_rppr-y1-activities.pdf.

  • Upcoming steps include development of a results retrieval interface, model administration interface, and launching of cloud computing infrastructure.
• Simulation software features

  • Cleveland Clinic and University of Utah teams continued to have monthly conference calls to discuss enhancements in FEBio, the simulation software used in Open Knee(s). Please refer to meeting minutes in Cleveland Clinic - University of Utah section of RecurringMeetings.

  • The Cleveland Clinic team detailed the prioritization of desired features and provided various test problems, see Revision as of May 18, 2014 or current version.

  • Feature to accommodate in situ ligament strains was implemented; this was summarized in the relevant wiki page, see Revision as of May 18, 2014 or current version. This activity was also summarized in the progress report submitted to National Institutes of Health along with some preliminary results, see https://simtk.org/websvn/wsvn/openknee/doc/R01_rppr-y1-activities.pdf. Relevant to this work, a conference abstract was submitted to CMBBE 2014, 12th International Symposium on Computer Methods in Biomechanics and Biomedical Engineering. Additional work will be conducted for general purpose use of this feature. It is anticipated that FEBio version 2.0 will incorporate this implementation.

  • Upcoming activities for simulation software enhancements include work on set definitions and rigid body movements.
• Experimentation
  • Experimentation on Specimen 001 had to be postponed to accommodate maintenance and upgrades to the robotics testing system. The anticipated timeline for data collection, dissemination, and related modeling efforts was pushed back by approximately three months.
  • Three knee specimens were acquired and identified as the first set of Open Knee(s) - Generation 2 specimens:

    • oks001, see Revision as of May 18, 2014,

    • oks002, see Revision as of May 18, 2014, and

    • oks003, see Revision as of May 18, 2014.

  • The experimentation workflow was illustrated in the specifications wiki page; see experimentation workflow section in Revision as of May 18, 2014 or current version.

  • A large group of experimentation specifications, detailed description of protocols, are in a mature state (completely documented or close finalization):

    • For target specimens, see Revision as of May 18, 2014 or current version.

    • For detailed imaging protocols to quantify tibiofemoral and patellofemoral joint anatomy, see Revision as of May 18, 2014 or current version. Three specific protocols were decided upon to highlight different tissues. A specimen transportation device was built to accommodate neutral placement of the cadaver joints and to ease their transportation to the imaging facility. The team decided to conduct testing on the fully intact joint, with the surrounding tissues. This will help increase signal to noise ratio during imaging. The relevant specifications will be completed once the team receives detailed image acquisition parameters from the imaging facility.

    • For detailed joint testing protocols to quantify tibiofemoral and patellofemoral joint response, see Revision as of May 18, 2014 or current version. The team decided to conduct robotics testing on the fully intact joint, without removing surrounding tissues. Relevant to this work, a conference abstract was submitted to WCB 2014, World Congress of Biomechanics, to outline the estimated effect of surrounding tissue on joint response. A related specification, for joint pressure measurements, was also documented, see Revision as of May 18, 2014 or current version.

    • For alignment of coordinate systems of imaging and joint testing systems, see Revision as of May 18, 2014 or current version.

    • A data management approach was needed and devised by the team, particularly to accommodate large number of tissue samples, and for curation and dissemination of heterogeneous data, see Revision as of May 18, 2014 or current version.

  • The team has also worked on other experimentation steps to finalize the protocols and enable their achievement:

    • Specimen preparation was found to be a logistically challenging task given the intensity of experimentation that need to be conducted on a specimen. Various stages were agreed upon by the team and documented in the relevant wiki page, see Revision as of May 18, 2014 or current version. Remaining work on this specification (about joint level preparation) is merely editorial. Additional documentation is needed about tissue level preparation.

    • The Cleveland Clinic team purchased a materials testing machine dedicated for Open Knee(s) tissue testing needs, see section on characterization of tissue mechanical properties in Infrastructure/ExperimentationMechanics. Tissue testing on this setup has been performed; as pilot tests and as part of preliminary work on tissue material properties conducted with Elvis Danso & Rami Korhonen, which is relevant to Open Knee(s). Detailed tissue testing specifications needs attention, see Revision as of May 18, 2014 or current version.

• Modeling & simulation

  • Modeling efforts for Specimen 001 had to be postponed to accommodate delays in experimentation.

  • The modeling & simulation workflow was illustrated in the specifications wiki page; see modeling & simulation workflow section in Revision as of May 18, 2014 or current version.

  • Various modeling & simulation specifications, in particular input/output details, were outlined:

    • For image segmentation, see Revision as of May 18, 2014 or current version.

    • For geometry generation, see Revision as of May 18, 2014 or current version.

    • For meshing, see Revision as of May 18, 2014 or current version.

    • For constitutive modeling, see Revision as of May 18, 2014 or current version.

    • For whole tissue representation, see Revision as of May 18, 2014 or current version.

    • For mesh assembly, see Revision as of May 18, 2014 or current version.

  • The team has been iterating various procedures to accomplish the target outcome of these specifications. Discussions for streamlined model assembly, compartmental modeling, and whole joint modeling are ongoing.
• Dissemination

  • As of May 18, 2014, Open Knee(s) provided 1 release package, Open Knee - Generation 1 version g1-s1-v1.0.1.202, which was downloaded 567 times (426 unique downloads), see Open Knee(s) download page.

  • As of May 18, 2014, Open Knee(s) provided 4 developer, 2 user documentation in release form, see Open Knee(s) documents page.

  • With a few exceptions (as requested by a collaborating team), all of the project website; wiki site, e.g. specifications, meeting minutes; and source code repository are publicly accessible.

  • As of May 18, 2014, Open Knee(s) website received 144,359 page hits in the past 180 days with 24,624 unique visitors, see Open Knee(s) project activity.

  • As of May 18, 2014, Open Knee(s) enabled 6 studies by external research teams, see Open Knee(s) publications page.

  • Due to delays in experimentation, dissemination of Open Knee(s) - Specimen 001 data set had to be postponed.
• Community projects

  • Completed specifications (a total of 3) were promoted to get feedback from the community. These specifications were listed as documents for developers in Open Knee(s) documents page. A forum entry for each mature specification was created to obtain input from the community, see Open Knee(s) public forums. A public write discussion page was created for each of these specifications to curate feedback. Biomechanics community was asked to comment through a request in Biomch-L, a community of more than 17,000 worldwide.

    • Specimen characteristics, see Revision as of May 18, 2014 or current version.

    • Mechanical testing of knees, see Revision as of May 18, 2014 or current version.

    • Data management, see Revision as of May 18, 2014 or current version.

  • Few synergistic short duration side projects were launched:
    • Cara Sullivan has been working on a project through curation media and development of a promotional video. Her efforts may seed a promotion program for the Open Knee(s).
    • Elvis Danso, a PhD student from Rami Korhonen's laboratory, along with Open Knee(s) Cleveland Clinic team, has been working on a project to quantify relative tissue properties within the knee. This project can inform fine-tuning of Open Knee(s) tissue testing protocols.
  • Ahmet Erdemir had discussions with external investigators on potential collaborations based on the Open Knee(s):
    • Steve Klisch, Department of Mechanical Engineering, Cal Poly State University - San Luis Obispo, had an interest on expanding upon their work using Open Knee(s). His team was planning to recruit new graduate students, who may use Open Knee(s) for their research by developing extended versions of Open Knee - Generation 1. They may be able to contribute to Open Knee(s) - Generation 2 work depending on the alignment of priorities between groups.
    • Austin Ramme, Orthopaedic Surgery, NYU Langone Medical Center, showed an interest in knee modeling work with the Open Knee(s) team; by adapting the Open Knee - Generation 1 model or by building a problem-specific model. There may be an opportunity for him to contribute to Open Knee(s) - Generation 2 work.
• Publications
  • Open Knee - Generation 1 publication was delayed. At this moment, no estimate is available for a potential timeline.

  • Open Knee abstract, which was presented in ASME/FDA 2013 1st Annual Frontiers in Medical Devices: Applications of Computer Modeling and Simulation, September 11-13, 2013, Washington, DC, was published in Journal of Medical Devices, see doi: 10.1115/1.4025767.

  • An abstract on Open Knee - Generation 1, in particular to open development approach and scripting, was submitted to WCB 20014, World Congress of Biomechanics.
  • Two abstracts about Open Knee(s) related work were submitted to various conferences.
    • An abstract on the implementation of in situ strain feature was submitted to CMBBE 2014, 12th International Symposium on Computer Methods in Biomechanics and Biomedical Engineering. This was led by the University of Utah team in collaboration with the Cleveland Clinic group.

    • An abstract to clarify the role of intact surrounding tissue on knee joint response was submitted to WCB 2014, World Congress of Biomechanics.This was led by the BioRobotics Core of the Cleveland Clinic along with other members of the Open Knee(s) team at the Cleveland Clinic.

• Other

  • First year progress report was submitted to National Institutes of Health, see activities, products, resources.

  • The entry page for the wiki was organized for easy access to various areas of the wiki, see Revision as of May 18, 2014 or current version.

  • A story on Open Knee(s) was published in Crain's Cleveland Business February 24, 2014 issue. The team is also exploring the possibility to develop a promotional video. More details can be found in the visibility wiki page, see Revision as of May 18, 2014 or current version.

Details of activities can be found in past meeting minutes of various teams at RecurringMeetings. For a complete list of various tasks (finished or in progress), please refer to https://simtk.org/pm/?group_id=485.

Plans for Next Progress Period (May 19, 2014 - November 2014)

• Enhance prototype of web based computation infrastructure (Stanford University)
• Implement additional features in simulation software (University of Utah)
• Complete experimentation for Specimen 001 (Cleveland Clinic)
• Complete experimentation for Specimen 002 (Cleveland Clinic)
• Launch model for Specimen 001 (Cleveland Clinic)

• Complete development and review of modeling specifications (Cleveland Clinic & Community)

• Complete development and review of simulations specifications (Cleveland Clinic & Community)

• Start plans for 2015 summer internship program as part of community projects (reviews, internships, and collaborations) (Cleveland Clinic & Community)

• Publish previous work on Open Knee - Generation 1 (Cleveland Clinic)

For more details, see the Roadmap and the proposed timeline of the proposal in GrantInformation/proposal.pdf.

Meeting Minutes

Date, Time, Means:

• May 19, 2014, 2:00 PM EST, conference call (AE,YD,TG,RK)
• May 30, 2014, 8:00 AM EST, in person meeting (AE,JH,MJ,PS)

Attendees:

1. Ahmet Erdemir (Cleveland Clinic)
2. Jason Halloran (Cleveland Clinic)
3. Morgan Jones (Cleveland Clinic)
4. Paul Saluan (Cleveland Clinic)
5. Yasin Dhaher (Northwestern University)
6. Trent Guess (University of Missouri)
7. Rami Korhonen (University of Eastern Finland)

Agenda:

1. Progress update, see progress report at:

   • http://wiki.simtk.org/openknee/RecurringMeetings/2014-05-19

2. Recap of previous meeting minutes, see meeting minutes at:

   • http://wiki.simtk.org/openknee/RecurringMeetings/2013-10-28

3. Action items for following meeting.
4. Other.

Immediate Action Items:

See notes for details.

• All Advisory Board Members

  • Get an account at https://simtk.org for Ahmet Erdemir to add them as team members with write access to project site.

  • Get accustomed to the project site.
  • Comment on current progress and planned activities in case some issues were overlooked during the meeting.
  • Spread the word about Open Knee(s).

• Ahmet Erdemir

  • Confirm the required relaxation time for characterization of tissue stress-strain response; part of tissue testing specification.
  • Incorporate scheduling surgeons during knee testing to allow appropriate dissection of secondary tissue types for future testing; part of specimen preparation specification.
  • Confirm leg positioning and loading during anatomical landmark acquisition; part of specimen preparation specification.
  • Explore the possibility to develop a tissue bank for reuse of cadaver specimens.

Notes:

1. Due to scheduling conflicts, a series of meetings were conducted to inform the Advisory Board about Open Knee(s) activities and get feedback.
2. Progress update.
   • Ahmet provided a summary of progress by using individual items in the progress report as reference (see above). The Advisory Board members found the progress appropriate. Clinicians noted the strength of the bottom-up approach to build reliable knee joint models and appreciated the detailed processes involved in establishing and executing specifications.
   • Ahmet emphasized the full specimen-specificity aimed for Open Knee(s) models and the difficulties in logistics of related data collection. This approach requires collecting tibiofemoral and patellofemoral joint kinematics-kinetics data (with contact pressure for the patellofemoral joint); anatomical imaging of the same joint, along with specimen preparation to accommodate accurate registration between imaging and mechanics data; and the acquisition and testing of tissue samples from the same joint.
   • For tissue modeling, Yasin raised a concern about the in situ strain in ligaments. The knee joint data will be purely specimen-specific, from joint geometry and kinematics-kinetics to tissue stress-strain response. There will not be any explicit measurement of in situ strain, as this is a rather difficult metric to measure. Nonetheless, joint kinematics-kinetics data, particularly under laxity loading, may provide opportunities to calibrate ligament in situ strains with material properties of the ligament determined from tissue testing in a specimen-specific manner.
   • Rami noted that 45 to 60 minutes relaxation time during tissue testing, particularly for ligaments and tensile cartilage and meniscus specimens may not be enough. Ahmet acknowledged this limitation and indicated that tests may be extended to accommodate long relaxation times. This may also depend on measurement and experimentation resolution.
   • For tissue testing, Morgan and Paul noted the importance of various tissues of the knee other than the primary ones planned for further testing. These included MPFL, ligaments of the posterior-lateral corner, ALL, various zones of articular cartilage, the IT band, quadriceps tendon and hamstring. While the former tissues were noted for their potential influence on joint response, the latter two are commonly used for reconstructions, e.g. of the ACL. Therefore the knowledge of their relative properties may be important for specimen-specific interventions. Ahmet noted that testing of all these tissue may not be possible due to logistical constrained. Nonetheless, he acknowledged that these tissues should be dissected and kept for future testing, when and if the opportunity arise. Snehal, who performs dissections, previously noted the difficulty to locate and appropriately dissect these tissues, a difficulty shared by Ahmet and Jason. Morgan and Paul were asked to help tissue dissection, in particular for isolation of these tissues. Ahmet will inform them about scheduling when the time comes.
   • In regard to joint testing, Morgan wondered how the alignment of the joint, based on hip and ankle location, would be determined. Relative location of anatomical landmarks can provide this information. Yet, Morgan raised the concern about the importance of joint positioning and loading, i.e., standing posture with some compressive loading, when comparing tibia and femur axes. While this may be difficult to attain during specimen preparation, such information can be reconstructed from robotics data, which will be registered with anatomical landmark locations and magnetic resonance images of the joint.
3. Recap of previous meeting minutes.
   • A few discussions resurfaced from previous meeting minutes.
   • Trent reiterated that reproducibility of cartilage segmentation may be problematic. This issue was also discussed in the previous meeting and one way is to design a small study where different groups can work on the same knee to segment the cartilage and compare their results. At this moment, no resources are allocated within the Open Knee(s) team to conduct this. Nonetheless, the specifications on image segmentation will provide guidelines to minimize the subjectivity of segmentation.
4. Action items for following meeting.
   • See Immediate Action Items above.
5. Other.
   • Efforts to engage community were emphasized and the Advisory Board was asked to comment on the wiki about specifications. Ahmet also wondered if any of the board members had any ideas to engage the community. He summarized the plan for community internship program he had in his mind. Yasin proposed to reach out to individual laboratories and ask investigators if they have any student that can contribute during summer while they are hosted in their own institute. If a project that benefits both Open Knee(s) and the external investigators research program, this may be a good venue to pro-actively engage the community. Ahmet emphasized that these project should return something to the Open Knee(s) and if that is the case, the internship can be paid by the Open Knee(s). Ahmet also noted that there may be intern candidates who may not be associated with any laboratory or an established investigator but may still provide useful input and work for Open Knee(s). By having a mechanism to reach out such communities may be beneficial. Yasin also advised to have a virtual conference at the end of summer, where all the interns and their principal investigators may interact with each other. In another meeting, Paul mentioned the possibility to expose interns to clinical care of the knee joint. While this requires a physical visit to the Cleveland Clinic, it can be a unique opportunity allowing direct interactions with physicians, who are at Open Knee(s) Advisory Board.
   • Yasin wondered if Open Knee(s) data will provide cartilage thickness map in relation to anatomical landmarks. Such knowledge will be useful to morph cartilage geometries to different knees. While the project will not specifically provide this information, the image sets and registration approaches can facilitate generation of such processed data by others.
   • Morgan noted the possibility to build a specimen bank, using tissue dissected and stored during Open Knee(s) project. Previously, Paul kindly agreed to let the Open Knee(s) team utilize knee specimens used in his other studies, where various tissue types can be sampled and tested. All these tissue samples can be utilized to build a virtual tissue bank, providing their mechanical response. Nonetheless, if logistics allow, a physical tissue bank can be built and maintained.
   • Open Knee(s) specimens are hip-to-toe specimens. Paul asked about the possibility to use hips, ankles, and feet. Ahmet entertained this opportunity. Other investigators at the Cleveland Clinic may be interested in these, e.g. Jim Rosneck (hip), James Sferra (foot and ankle).