Fixel-Based Analysis Reveals Detailed White Matter Changes in Semantic Dementia

Background and Purpose Accurately characterizing white matter (WM) microstructure is critical for understanding neurodegenerative diseases such as semantic dementia (SD). Regionally constrained techniques like tract-based spatial statistics (TBSS) rely on diffusion-tensor imaging (DTI) and assume a single fiber population per voxel, limiting their sensitivity to complex architecture. Fixel-based morphometry (FBM) overcomes this by assessing multiple fiber populations (fixels) within a single voxel. In this study, we compared TBSS and Fixel-based analysis (FBA) for detecting WM alterations in SD variants associated with anterior temporal lobe (ATL) atrophy. Methods Multi-shell diffusion MRI from 16 left-lateralized semantic-variant PPA (svPPA) and 15 right-lateralized semantic-behavioral fronto-temporal dementia (sbvFTD) cases, plus 44 neurologically healthy controls, underwent both TBSS-DTI and whole-brain FBA. Fiber-specific metrics of fiber density and cross-section were contrasted with conventional DTI measures. Results Both methods confirmed damage to ATL-connected tracts—the uncinate fasciculus, inferior longitudinal fasciculus, inferior fronto-occipital fasciculus, and temporal projections of the arcuate fasciculus. FBA, however, revealed additional involvement of juxtacortical and other previously overlooked pathways, including the tapetum and anterior commissure, projections to the parahippocampal gyrus and amygdala, and longer-range parietal connections. Conclusions By capturing fiber-specific micro- and macrostructural changes, FBA yields a more comprehensive map of WM degeneration in SD than TBSS. The ability to detect early alterations in commissural and mesial-temporal pathways refines our understanding of disease spread and highlights candidate targets for monitoring and intervention aimed at preserving cognitive function.