Predicting recovery from concussion during military cadet training using multimodal MRI data and machine learning

In the military, concussions are common and many occur while non-deployed, including during cadet training exercises. For the majority of those with concussions, symptoms resolve on their own but for a miserable minority symptoms persist beyond the typical 3-month recovery period, impacting quality of life. Most concussion research produces group level inferences which cannot be used to make individual predictions. We propose a supervised machine learning approach to build a model to predict symptom recovery from multiple MRI brain measures. The ability to identify those in the acute phase likely to have poor symptom recovery at 6 months post injury is incredibly useful for clinical decision making, concussion management, optimized treatment and personalized medicine. This project will contribute to bridging the gap between research and clinical use, by adapting and validating machine learning applications in neuroimaging. TO BE CONT’D

Faculty Supervisor:

Douglas Cook

Student:

Ashley Ptinis

Partner:

Synaptive Medical Inc.

Discipline:

Medicine

Sector:

Medical devices

University:

Program: