Combining MRI and cognitive evaluation to classify concussion in university athletes

Area of Science:

• Neuroscience
• Medical Imaging
• Sports Medicine

Background:

• Current concussion assessment tools lack objectivity and reliability.
• Neurological injury detection requires advanced diagnostic methods.
• University athletes are a key population for concussion research.

Purpose of the Study:

• To develop and evaluate algorithms for objective concussion detection.
• To integrate neuroimaging and cognitive measures for improved accuracy.
• To assess the reliability of multi-modal data in diagnosing acute concussion.

Main Methods:

• Multi-site study involving diffusion tensor imaging and resting state functional MRI.
• Cognitive assessments including concentration and delayed memory tests.
• Machine learning models (logistic regression, support vector machines) trained on athlete data (29 concussed, 48 controls).

Main Results:

• Concussed athletes showed greater symptoms, poorer concentration, and delayed memory.
• Lower functional connectivity was observed in frontoparietal and visual networks in concussed athletes.
• Combined MRI (mean diffusivity) and cognitive data achieved 74% accuracy and 64% sensitivity.

Conclusions:

• Algorithms integrating multi-modal data reliably detect concussion-related neurobiological changes.
• This approach offers an objective and reliable tool for concussion assessment and diagnosis.
• Combining neuroimaging with cognitive performance enhances concussion detection capabilities.