Longitudinal automated fiber quantification reveals corticospinal tract dynamics and prognostic significance after primary basal ganglia hemorrhage.

BACKGROUND: The integrity of the corticospinal tract (CST) is a critical determinant of functional recovery after intracerebral hemorrhage. However, the longitudinal evolution of CST microstructure after primary basal ganglia hemorrhage (PBGH) and its relevance for clinical outcome prediction remain incompletely understood. METHODS: Thirty-nine patients with PBGH underwent diffusion tensor imaging within 3 days of onset (T1), 14 ± 2 days (T2), 30 ± 2 days (T3), and 90 ± 2 days (T4). The CST was segmented with the automated fiber quantification technique, and diffusion tensor metrics-fractional anisotropy (FA), axial diffusivity (AD), mean diffusivity (MD), and radial diffusivity (RD)-were derived. Functional outcome at 3 months was evaluated with the modified Rankin Scale (mRS). RESULTS: Thirty-nine patients and 32 matched healthy controls were recruited. The ipsilesional CST (IL-CST) demonstrated persistently reduced FA and elevated MD and RD across multiple tract segments compared with controls. In contrast, contralesional CST alterations were restricted to scattered elevations in MD, RD, and AD, with no significant changes in FA. Longitudinal analysis indicated that FA decreased from T1 to T2 and partially recovered by T3-T4, whereas AD, MD, and RD displayed biphasic or region-specific temporal patterns. An elastic-net logistic regression model combining baseline DTI features with clinical variables achieved robust prognostic performance for 3-month outcome (AUC = 0.902, permutation P < 0.001). CONCLUSIONS: Longitudinal AFQ analysis revealed dynamic, region-specific CST microstructural alterations after PBGH. Baseline tract-specific DTI metrics, when integrated with clinical variables, provided complementary prognostic information for 3-month functional outcomes, highlighting their potential clinical value.