Back to 2019 MSM Agenda
Session Description: This session will present the results of the 2018 model credibility plan review organized by the Committee on Credible Practice of Modeling & Simulation in Healthcare (CPMS) and highlight some of the best practices identified by that review.
2017 MSM Model Credibility Session
2018 MSM Model Credibility Session
11:00-11:20 Results of the 2018 Model Credibility Plan Review - Jerry Myers, NASA - GRC & CPMS Co-Chair - 2018-2019 Mid-Term Credibility Plan Review
This presentation summarizes the IMAG-MSM's Committee on Credible Practice in Modeling and Simulation in Healthcare's (CPMS) mid-term review of each MSM U01 project's progress toward fulfilling their Model Credibility plan. The reviewers, consisting of the members of the CPMS Executive Committee, focused on the documentation and communication of each PI's credibility efforts and did not perform an assessment of the implemented credible practice. The CPMS members sought insight on how PIs view the importance of communicating credibility, as well as how PIs interpret establishing evidence for credibility within the scope of the CPMS Tens Simple Rules. The model credibility review also sought to evaluate several aspects of the reporting and update process with respect to communicating model credibility status to the uninformed user with respect to the intended use of each U01 model, as well as communicate where the more interested user can go to find detailed information. Reviewers and U01 participants identified several important recommendations for the IMAG-MSM community that should improve the continued efforts to establish acceptable credibility practices and community communications.
11:20-11:35 Best Practices Talk 1 -- Bill Lytton, SUNY Downstate -- The conundrum of ``appropriate data'' in biomedical modeling
In traditional mechanical engineering applications, the concept of appropriate data is relatively clear -- if a material is to be used in a model jet engine, its properties at vastly different temperatures will be measured and considered as properties in the engine model. In biomedical applications, however, we are obliged to take our parameters where we can get them -often from in vitro preparations, cell culture, or isolated proteins, very far from the complex modulatory environment encountered for the in vivo context that is of greatest interest. For this reason, biomedical modeling is an ongoing work-in-progress where estimated parameters must often be re-evaluated and modified.
11:35-11:50 Best Practices Talk 2 -- Ellen Kuhl, Stanford -- Credible enough? Perspectives from cardiac growth and remodeling
Our models are continuously growing in complexity and it is increasingly difficult to ensure that our implementations are correct. Using contrasting modeling and simulation execution strategies can provide insight into the agreement and discrepancies of different implementation strategies. Here we illustrate how we can increase model credibility through testing competing implementations in cardiac growth and remodeling. We show how competing implementations can increase the confidence in our models and, ultimately, help translate our models into clinical practice.
11:50-12:00 Group Discussion
Interactive Discussion (please put you name before your comments):
DK- did you examine the report results by gender?