Corticospinal Tractometry and Whole-Brain Connectometry of Hand Dexterity in Chronic Stroke and Traumatic Brain Injury

Hand dexterity impairment is a major contributor to long-term disability after acquired brain injury, yet the white matter substrates supporting residual dexterity remain incompletely understood. We investigated diffusion MRI markers of hand dexterity in individuals with chronic stroke (n = 9) and traumatic brain injury (TBI; n = 8) using complementary tract-specific and whole- brain approaches. Partial least squares regression (PLSR) was used to evaluate the cross-validated predictive relevance of bilateral corticospinal tract (CST) diffusion and tractometry features, while quantitative anisotropy (QA)-based correlational tractography was used to identify distributed white matter pathways associated with dexterity performance measured using Box and Block Test (BBT) and MusicGlove Dexterity Test(MGDT). In stroke, CST features predicted BBT performance (Q2 = 0.69, r = 0.85, permutation p = .010) and, more modestly, MGDT performance (Q2= 0.22, r = 0.72, permutation p = .008). In contrast, CST-based models showed no predictive relevance for dexterity outcomes in TBI. Whole-brain connectometry revealed that better dexterity after stroke was associated with greater QA across distributed pathways extending beyond the CST, including commissural, association, and projection fibers. Box and Block Test performance was prominently associated with callosal and cingulum-related pathways, whereas MusicGlove performance showed greater representation of CST and projection pathways. In TBI, significant connectometry findings for the BBT similarly implicated distributed commissural and association pathways, whereas no significant pathways were identified for the MusicGlove test. Together, these findings suggest that the structural correlates of hand dexterity extend beyond the CST and vary across dexterity measures and injury populations. Although preliminary given the small cohorts, the complementary tractometry and connectometry findings support a network-level characterization of residual hand function after acquired brain injury and motivate validation in larger cohorts.