A T2-MRI Dataset of Cervical Spondylotic Myelopathy for Automated Segmentation, Radiomics, and Neurological Assessment

Cervical Spondylotic Myelopathy (CSM) is a compression of the spinal cord and commonly linked with spinal dysfunction in older adults, primarily due to spinal cord compression. Its diagnosis and prognosis are restricted by the complex nature of its presentation and the limitations of conventional assessment methods. This work presents the first publicly available T2-MRI dataset for CSM, offering a unique compilation of manually segmented images and neurological assessment ratings. The dataset is poised to advance research in automated CSM segmentation, biomarker discovery, and prediction of neurological scores. Additionally, the inclusion of patient neurological indices (JOA and NDI) and their correlation with biomarkers enriches the dataset, promising to propel MRI-based evaluation of neurological impairment in CSM. The deployment of this dataset for radiomics studies and the development of machine learning algorithms holds the promise to transform the landscape of patient self-assessment, paving the way for automated, MRI-based diagnostic tools that could offer rapid, reliable, and non-invasive alternatives to traditional method.