Anatomy-guided Fiber Trajectory Distribution Estimation for Cranial Nerves Tractography

Diffusion MRI tractography is an important tool for identifying and analyzingthe intracranial course of cranial nerves (CNs). However, the complexenvironment of the skull base leads to ambiguous spatial correspondence betweendiffusion directions and fiber geometry, and existing diffusion tractographymethods of CNs identification are prone to producing erroneous trajectories andmissing true positive connections. To overcome the above challenge, we proposea novel CNs identification framework with anatomy-guided fiber trajectorydistribution, which incorporates anatomical shape prior knowledge during theprocess of CNs tracing to build diffusion tensor vector fields. We introducehigher-order streamline differential equations for continuous flow fieldrepresentations to directly characterize the fiber trajectory distribution ofCNs from the tract-based level. The experimental results on the vivo HCPdataset and the clinical MDM dataset demonstrate that the proposed methodreduces false-positive fiber production compared to competing methods andproduces reconstructed CNs (i.e. CN II, CN III, CN V, and CN VII/VIII) that arejudged to better correspond to the known anatomy.