Streamline-Based Analysis: A novel framework for tractogram-driven streamline-wise statistical inference

Most diffusion MRI studies of white matter are interpreted in terms of anatomically defined fibre bundles, yet current statistical frameworks fail to simultaneously provide high-resolution pathway-level localisation, whole-brain coverage, and good statistical power with family-wise error control. We introduce Streamline-Based Analysis (SBA), a novel framework that performs statistical inference directly on individual tractography streamlines. SBA flexibly accommodates any imaging-derived, streamline-wise quantitative metric leveraging streamline similarity to perform streamline-wise data smoothing and statistical enhancement within a non-parametric permutation testing framework that controls family-wise error. Applied to healthy ageing, SBA recapitulated established patterns observed with Fixel-Based Analysis and uncovered previously unreported effects, while yielding more spatially coherent, pathway-level effects with improved anatomical interpretability. By achieving pathway-level specificity, whole-brain coverage, and family-wise error control, SBA fills a key analytical gap between voxel-, fixel-, and tract-level methods, providing a flexible framework for detecting white matter effects across diverse diffusion MRI studies.