Structural brain changes in subacute spinal cord injury: an analysis of diffusion kurtosis imaging and diffusion tensor imaging metrics with clinical correlation

Ernst Christiaanse^1,2,3*Jothini Sritharan^2,4Patrik O Wyss¹Anke Scheel-Sailer^2,5Alexander Leemans³Rajeev K Verma^1,2Alberto De Luca^3,6Giuseppe A Zito^2,4

• ¹Swiss Paraplegic Center, Nottwil, Switzerland
• ²Schweizer Paraplegiker-Forschung, Nottwil, Switzerland
• ³Image Sciences Institute, Center for Image Sciences, University Medical Center Utrecht, Utrecht, Netherlands
• ⁴Faculty of Health Science and Medicine, Universitat Luzern, Lucerne, Switzerland
• ⁵Centre for Rehabilitation & Sports Medicine, Inselspital and Berner Reha Zentrum, Inselspital Universitatsspital Bern, Bern, Switzerland
• ⁶Neurology Department, UMC Utrecht Brain Center, Universitair Medisch Centrum Utrecht, Utrecht, Netherlands

Introduction: Diffusion tensor imaging (DTI) and diffusion kurtosis imaging (DKI) can quantify indices related to brain structure and their change in pathology. However, only few studies have applied these techniques to spinal cord injury (SCI), and subtle microstructural changes in the brain of SCI individuals are not well understood. Our goal was to investigate structural changes in the brain using DTI (fractional anisotropy, FA; mean diffusivity, MD) and DKI parameters (kurtosis anisotropy, KA; mean kurtosis, MK) in subacute SCI and to study whether these changes were associated with clinical outcomes. Methods: Twenty-eight individuals with SCI underwent brain MRI three months post-injury, alongside 20 healthy controls. Imaging included a multi-shell diffusion protocol, from which DTI and DKI metrics (FA, MD, KA and MK) were derived. Group comparisons were conducted for each metric across 17 brain regions selected based on their relevance to SCI from previous studies. Multiple comparison corrections were applied per metric to account for the number of examined regions. Effect sizes were calculated using Cohen's d. For regions showing significant group differences, Spearman correlations were performed to assess associations between imaging metrics and clinical outcomes, including neurological status (ISNCSCI) and functional independence (SCIM III), with correction for multiple comparisons. Results: MD was significantly higher in the right genu of the corpus callosum in the SCI group (adjusted p = 0.021). In this region, MD negatively correlated with SCIM scores (r = -0.51, p = 0.022), whereas MK showed a positive correlation (r = 0.482, p = 0.038). Discussion: Structural changes in the corpus callosum may reflect impaired interhemispheric communication, linked to reduced functional independence after SCI. DTI and DKI could serve as complementary tools for identifying brain-based biomarkers, potentially informing recovery trajectories.