MR-Neurography with a Fully Automated Nerve Segmentation Workflow

Motivation: There is no established workflow for an automated segmentation of peripheral nerves for MR-Neurography (MRN). Existing methods provide poor nerve visibility. Goal(s): Provide an improved workflow for cross-patient MRN for healthy and pathologic peripheral nerves, enabling automated nerve-segmentation and improving precision of clinical applications. Approach: 3D T1w-FLASH and DWI SE-EPI sequences were optimized for MRN. A fully automated nerve-segmentation algorithm for healthy and pathologic nerves was developed. Results: Our workflow delivers ultra-high resolution MRN-images with improved nerve visibility and a fully automated identification and segmentation of peripheral nerves. Impact: The developed workflow enhances clinical MR-Neurography by improving image quality, automating the identification and segmentation of nerves, rendering diagnoses more precise and efficient, and allowing cross-patient application with both healthy and pathological peripheral nerves.