SCIseg: Automatic Segmentation of T2-weighted Intramedullary Lesions in Spinal Cord Injury.
Enamundram Naga KarthikJan ValošekAndrew C SmithDario PfyfferSimon Schading-SassenhausenLynn FarnerKenneth A WebberPatrick FreundJulien Cohen-AdadPublished in: medRxiv : the preprint server for health sciences (2024)
An open-source, automatic method, SCIseg , was trained and evaluated on a dataset of 191 spinal cord injury patients from three sites for the segmentation of the spinal cord and T2-weighted lesions. SCIseg generalizes across traumatic and non-traumatic lesions, scanner manufacturers, and heterogeneous image resolutions, enabling the automatic extraction of lesion morphometrics in large multi-site cohorts. Quantitative MRI biomarkers, namely, lesion length and maximal axial damage ratio derived from the automatic predictions showed no statistically significant difference when compared with manual ground truth, implying reliability in SCIseg's predictions.
Keyphrases
- spinal cord injury
- deep learning
- spinal cord
- convolutional neural network
- neuropathic pain
- contrast enhanced
- machine learning
- end stage renal disease
- magnetic resonance
- magnetic resonance imaging
- ejection fraction
- newly diagnosed
- chronic kidney disease
- oxidative stress
- resistance training
- peritoneal dialysis
- computed tomography
- high resolution
- patient reported outcomes
- mass spectrometry
- patient reported