While it is relatively easy to collect a bunch of data, it is significantly harder to interpret the meaning behind the numbers in a way that is validated, intuitive, and useful. As a result, we have spent decades developing and validating sophisticated computational models to assist us with this challenging task. These models are the "glue that hold our logic together", and are the foundation for connecting the dots between the complex system of parameters that are important for understanding causal pathways. While our models share many things in common, they can be tuned and customized for different applications depending on the level of complexity required to answer the question at hand and the types of data that are available. Below are some examples of the different types of models we build.
Relationships between the outcomes we are interested in and the things we can measure are often complex and difficult for human brains to identify. Using traditional statistics and machine learning techniques, however, we are often able to map these connections so that readily available measurements can be used as reliable and validated predictors of the outcomes that really matter.
Biomedical Imaging Models
Traditional biomedical imaging modalities such as MRI and CT produce a series of two dimensional images that can be hard to use or interpret. By stitching together multiple images from a single or even multiple modalities, we are able to create 3D reconstructions that are both accurate and intuitive for a variety of applications.
Rigid Body Dynamics Models
Our rigid body dynamics models leverage inputs from various sensors that track motion, muscle activity, and external forces in a sophisticated physics engine that acounts for gravity, inertia, and other important factors. The outputs of these models are the dynamic joint forces inside the body, which can be used as predictors of injury risk during nearly any occupational or clinical activity.
Advanced FEM Models
Finite-Element-Method (FEM) models study the mechanical behavior of a complex structure by simplifying it into many small elements that can be evaluted independently and then pieced back together. We use this technique to understand the complex behavior of different tissues inside of the body under mechanical forces.
Patient Specific Models
Our patient-specific models combine many of the modeling techinques described above to quantify the complex tissue forces in a patient's spine. The information gathered from these models can be used to perform advanced diagnostics, predict surgical outcomes, or educate patients on their condition.