Recurring Meeting of Cleveland Clinic - University of Utah

Date: November 11, 2015

Time: 2:00 PM EST

Means: Conference call

Attendees:

1. Ahmet Erdemir (Cleveland Clinic)
2. Jeff Weiss (University of Utah)
3. Ben Ellis (University of Utah)
4. Steve Maas (University of Utah)

Agenda:

1. In situ strain feature - new directions.

   • http://wiki.simtk.org/openknee/Febio/InSituStrain

2. Open Knee(s) related FEBio features. What did we do? What is next?

   • http://wiki.simtk.org/openknee/Specifications/FebioFeatures

3. Decide action items for next meeting.
4. Other.

Immediate Action Items:

• Cleveland Clinic (Ahmet)
  • Setup a conference call between University of Utah and Stanford University teams.
• University of Utah (Jeff)
  • Explore manuscripts for membrane/shell element formulations suitable for implicit analysis.
• University of Utah (Steve)
  • Send Cleveland Clinic team the Open Knee(s) - Generation 1 model (with in situ strains) used in the in situ strain feature manuscript.
• University of Utah (All)
  • Prioritize upcoming feature requests for implementation in FEBio.

Notes:

1. In situ strain feature - new directions.
   • The group discussed new research directions related to in situ strain feature. In particular, the prescribing of in situ strain of multiple ligaments when joint level equilibrium (or joint reaction forces) are sought after. Jeff mentioned data on medial collateral ligament where ligament forces were calculated with a model after application of specimen-specific experimental in situ strains. A possible research direction may include exploration of uncertainties in ligament forces as a function of variations in situ strain profiles. Nonetheless, data with measured joint forces may serve better as the gold standard. Open Knee(s) specimens have specimen-specific joint kinematics-kinetics supported by specimen-specific anatomical imaging and prospectively, tissue properties. These data may be used to identify in situ strains for ligaments for specimen-specific calibration. The group will continue brainstorming based on these discussions.
   • Ahmet asked for the Open Knee(s) - Generation 1 model file with in situ strains implemented by Steve. This file can inform future releases of Open Knee(s) - Generation 1. Steve will provide the relevant model input files. 1. Open Knee(s) related FEBio features. What did we do? What is next?
2. Open Knee(s) related FEBio features. What did we do? What is next?
   • Ahmet went over the FEBio feature requests wiki page to summarize those implemented for Open Knee(s) and start discussion for additional features that may be useful.
   • Ahmet asked for penetration based contact, i.e., a contact formulation based solely on penetration depth and a contact stiffness. This may be useful to simplify models, in particular when joint level mechanics is sought after. Steve noted that FEBio essentially has this feature. The augmentation during contact can be turned off.
   • Ahmet also asked for the possibility of springs with wrapping. This feature may use contact formulation between nodes of 1D elements (springs) and a surface. Steve can implement this. Jeff noted that some usability features may need to be implemented, particularly when a user wants to prescribe the total stiffness of a ligament where a series of springs may need to be implemented to represent the ligament.
   • The group also discussed implementation of an auto-contact feature, which may simplify model definition potentially at the expense of computational cost.
   • Ahmet asked if a parts feature can be implemented in FEBio. This can provide an additional layer on top of node, element, and set descriptions that can be children of a part. The feature will allow swapping tissues, without the requirement of renumbering the elements and nodes.
   • Another topic of interest was related to modeling of thin layers. For ligaments, tendons, and capsule, e.g. the MCL, one may want to use a membrane representation. For cartilage, one may want to utilize a shell representation or a thin layer representation on a foundation. The University of Utah team has some experience using other solvers with such formulations. Jeff will look at some literature on membrane formulations, particularly from the perspective of numerical solution stability.
   • In summary, the group recognized four features for prospective implementation: part based numbering nodes/elements, autocontact, membrane/shell formulations (implicit based), and springs with wrapping. Ahmet is primarily interested in the former. The University of Utah team will try to prioritize as well.
3. Decide action items for next meeting.
   • See Immediate Action Items above.
4. Other.
   • The University of Utah team mentioned the progress on quadratic tetrahedral element manuscript. These types of elements may be useful for Open Knee(s) - Generation 2 models.
   • Ahmet noted that Open Knee(s) has a summer internship program that may be beneficial for University of Utah students to consider.
   • Jeff asked about progress in Stanford University team. Ahmet recommended to setup a conference call in between groups. Jeff is interested in how cloud computing will be used. Ahmet described the workflow and mentioned that cloud computing will be available for specific models not for general purpose simulations using FEBio. Jeff is interested in launching a hub for FEBio. Ahmet recommended him to chat with Joy Ku from Stanford University.