BackgroundQuantitative MRI biomarkers in spinal cord injury (SCI) can help understand the extent of the focal injury. However, due to the lack of automatic segmentation methods, these biomarkers are derived manually, which is a time-consuming process prone to intra- and inter-rater variability, thus limiting large multi-site studies and translation to clinical workflows.PurposeTo develop a deep learning tool for the automatic segmentation of T2-weighted hyperintense lesions and the spinal cord in SCI patients.Material and MethodsThis retrospective study included a cohort of SCI patients from three sites enrolled between July 2002 and February 2023 who underwent clinical MRI examination. A deep learning model,SCIseg, was trained on T2-weighted images with heterogeneous image resolutions (isotropic, anisotropic), and orientations (axial, sagittal) acquired using scanners from different manufacturers (Siemens, Philips, GE) and different field strengths (1T, 1.5T, 3T) for the automatic segmentation of SCI lesions and the spinal cord. The proposed method was visually and quantitatively compared with other open-source baseline methods. Quantitative biomarkers (lesion volume, lesion length, and maximal axial damage ratio) computed from manual ground-truth lesion masks and automaticSCIsegpredictions were correlated with clinical scores (pinprick, light touch, and lower extremity motor scores). A between-group comparison was performed using the Wilcoxon signed-rank test.ResultsMRI data from 191 SCI patients (mean age, 48.1 years ± 17.9 [SD]; 142 males) were used for training. Compared to existing methods,SCIsegachieved the best segmentation performance for both the cord and lesions and generalized well to both traumatic and non-traumatic SCI patients.SCIsegis open-source and accessible through the Spinal Cord Toolbox.ConclusionAutomatic segmentation of intramedullary lesions commonly seen in traumatic SCI replaces the tedious manual annotation process and enables the extraction of relevant lesion morphometrics in large cohorts. The proposed model generalizes across lesion etiologies (traumatic, ischemic), scanner manufacturers and heterogeneous image resolutions.SummaryAutomatic segmentation of the spinal cord and T2-weighted hyperintense lesions in spinal cord injury on MRI scans across different treatment strategies, lesion etiologies, sites, scanner manufacturers, and heterogeneous image resolutions.Key ResultsAn open-source, automatic method,SCIseg, was trained on a dataset of 191 spinal cord injury patients from three sites for the segmentation of spinal cord and T2-weighted hyperintense lesions.SCIseggeneralizes across traumatic and non-traumatic lesions, scanner manufacturers, and heterogeneous image resolutions, enabling the automatic extraction of lesion morphometrics in large multi-site cohorts.Morphometrics derived from the automatic predictions showed no statistically significant difference when compared with manual ground truth, implying reliability inSCIseg’spredictions.