Mapping concussion-induced cerebellar injury: a personalized approach using diffusion tensor imaging (DTI).

OBJECTIVE: We investigated the potential of diffusion tensor imaging (DTI) in identifying alterations in cerebellar white matter (WM) in concussion patients. We hypothesized that DTI, combined with a Z-score analysis approach, could provide personalized assessment of cerebellar changes in these patients. MATERIALS AND METHODS: Fifty-one mTBI patients (25 females, age: 40.5 ± 11.7 years and 26 males, age: 45.0 ± 14.1 years) and 2700 healthy age/sex/vendor-matched controls were analyzed in this study. DTI data analysis focused on scalar metrics fractional anisotropy (FA) and mean diffusivity (MD). A Z-score approach was used to identify subject-specific deviations, while statistical models (linear regression, random forests, hierarchical clustering) compared the sensitivity of Z-scores and the Post-Concussion Symptom Scale (PCSS) in capturing individualized differences. RESULTS: Significant cerebellar DTI abnormalities were found in the concussion group. Z-score analysis showed higher sensitivity to cerebellar regions associated with balance and sleep. Statistical models revealed distinct patterns by age, sex, region, and symptoms, underscoring the value of personalized analysis and the subjectivity that remains in existing assessments like the PCSS. DISCUSSION: The combination of DTI and Z-score methods provides a promising framework for personalized detection of cerebellar changes in mTBI, potentially improving diagnostic accuracy and enabling more targeted interventions.