Model Personalization

Background

Model Personalization is a toolset to take modeling accuracy to the next level. Every person's body is different and Model Personalization takes that into account by adjusting many key model parameters in order to more closely match the subject. It builds on top of OpenSim and adds on key capabilities to personalize the joints, muscle-tendon attributes, neural control pathways, and ground contact parameters.

The first tool is Joint Model Personalization (JMP), which edits the joint makeups in the model in order to allow it to more closely match the subject's experimental movement. Joint contact angles are fundamental in dynamics calculations throughout the motion, so the ability to adjust these can greatly improve model accuracy.

Next, the Data Preprocessing sub-tool can be used with the updated OpenSim model in order to collect and organize the data needed to input into the rest of the tools. Alternatively, the user may independently process their data as long as the final setup matches the output of the preprocessing tool.

The next tool in the pipeline is Muscle-Tendon Personalization (MTP), which edits various muscle-tendon properties in the model. It also has two built in sub-tools to aid this process, Muscle-Tendon Length Initialization (MTLI) and Synergy Extrapolation (SynX). MTLI utilizes passive movement data to set initial variables such as optimal fiber length, tendon slack length, and maximum isometric force of muscles. SynX is able to break down the EMG data into muscle synergies in order to fill in the gaps for adjacent muscles that don't have data. This is a much more accurate way to estimate muscle activations than other methods used in the past.

Following this, Neural Control Personalization (NCP) is able to calculate muscle synergies for the whole body, including both areas with and without EMG data. MTLI can also be used as a sub-tool with NCP. If SynX was run in the previous step, NCP will take that data in to refine and add to it.

Finally, Ground Contact Personalization (GCP) is a tool that can personalize the contact properties between the subject's foot and the ground. This allows for ground reaction force calculations in new movements, such as those being modeled in the last step of Treatment Optimization.

Significance

One of the biggest challenges facing the bioengineering world today is the fact that every person is different. There is no one size fits all solution. The Model Personalization toolset is a revolutionary step in being able to model and predict outcomes for people on an individual basis. The ability to tailor a model so specifically to one specific individual in integral in the simulations later on to determine the best treatment options.