Anatomy-to-Tract Mapping: Inferring White Matter Pathways Without Diffusion Streamline Propagation
Diffusion tractography, a cornerstone of white matter mapping, relies on point-to-point streamline propagation---a process often compromised by errors stemming from inadequate signal-to-noise ratio and limited spatial-angular resolution in diffusion MRI (dMRI) data. Here, we present Anatomy-to-Tract Mapping (ATM), a novel deep learning framework that, for the first time, generates complete, bundle-specific streamlines directly from anatomical MRI, entirely bypassing the limitations of streamline propagation. ATM leverages the superior quality and minimal distortion of anatomical MRI, learning from multi-subject datasets to deliver robust, subject-specific streamline bundles while accurately preserving structural connectivity. This paradigm-shifting approach overcomes challenges associated with complex configurations, such as crossing, kissing, bending, and bottlenecks, providing anatomically precise reconstructions. Beyond individual-level mapping, ATM facilitates the creation of population-level streamline bundles aligned to average anatomical templates derived from diverse datasets. By offering global connectivity insights less affected by local uncertainties, ATM complements diffusion tractography, advancing white matter pathway reconstruction with a more reliable, anatomy-driven perspective.