Spinal cord microstructure-based tissue classification in cervical myelopathy

Motivation: Degenerative cervical myelopathy (DCM) due to chronic compression is generally examined using qualitative imaging, which does not specify how tissue microstructure changes within the cord. Goal(s): To implement a clustering-based tissue classification framework to identify tissue classes in DCM spinal cord. Approach: A variant of a clustering framework was implemented to classify cervical cord tissue based on 3 quantitative MRI measures without spatial input in 25 participants with DCM and one healthy control. "Tissue microstructure maps" were created for each participant. Results: Cluster sizes of white matter-like tissue were significantly lower in immediate above-compression regions compared to the whole cord, indicating upstream damage. Impact: Cervical cord tissue was classified in people with degenerative cervical myelopathy based only on clustering quantitative MRI measures. Different proportions of tissue clusters were seen in regions immediately above compression sites compared to the entire imaged cord (C2-C5).