Introduction to the NMSM Pipeline
The Neuromusculoskeletal Modeling (NMSM) Pipeline is a toolset to help model treatment options for people with movement impairments. It builds upon the OpenSim software in order to perform personalized treatment design for each subject.
There are two main toolsets built into the NMSM Pipeline. First, the Model Personalization toolset is used in order to modify model parameters such that they match a specific subject. Next, Treatment Optimization allows you to take that personalized model and input various treatment options in order to find the best option specific to them.
Model Personalization
Model Personalization is used to develop a subject-specific model that takes into account the unique joint parameters, muscle-tendon properties, neural control properties, and ground contact properties. Starting with a suitable OpenSim model, these properties can be refined from marker, EMG, and ground reaction data.
Main Tools
- Joint Model Personalization (JMP): Adjusts the joint parameters of a model in order to match the specific subject
- Muscle-Tendon Personalization (MTP): Adjusts the muscle parameters of a model in order to match the specific subject
- Neural Control Personalization (NCP): Calculates subject specific communication patterns between the brain and the muscles for particular movements
- Ground Contact Personalization (GCP): Establishes important physical properties regarding contact between the ground and the subjects feet or shoes
Sub Tools
- Data Preprocessing: Takes in raw data and OpenSim results and processes it into a useable form for the NMSM Pipeline
- Used after JMP
- Muscle-Tendon Length Initialization (MTLI): Presets particular muscle parameters such as optimal fiber length, tendon slack length, and maximum isometric force of muscles
- Used with MTP or NCP
- Synergy Extrapolation (SynX): Estimates EMG signals for muscles where data is not available
- Used with MTP
Treatment Optimization
Treatment Optimization uses a subject's personalized neuromusculoskeletal model to perform predictive modeling of various treatment options in order to provide personalized recommendations for the subject. There are three steps to this toolset: Tracking Optimization, Verification Optimization, and Design Optimization. Treatment Optimization can be used for any task that can be sufficiently described by an OpenSim model.
Main Tools
- Tracking Optimization (TO): Determines optimal controls settings for modeling in order to reproduce experimental data with minimal discrepancies between predicted and observed data
- Verification Optimization (VO): Confirms the results of treatment optimization in order to improve modeling reliability
- Design Optimization (DO): Framework to test different treatment options and determine which will lead to optimal results
Sub Tool
- Surrogate Model Creation: Fits polynomials to muscle and joint properties in order to more efficiently calculate data inputs
- Used automatically by each tool
For more information, go to User Guide