Self-Supervised Learning for Spinal MRIs
This work addresses the challenge of data scarcity in medical imaging for clinicians and researchers, though it is incremental as it applies existing self-supervised learning techniques to a new domain.
The authors tackled the problem of limited annotated medical imaging data by using longitudinal spinal MRI scans as free self-supervision to pre-train a deep network, resulting in superior performance on a disc degeneration grading task with fewer annotated samples compared to training from scratch.
A significant proportion of patients scanned in a clinical setting have follow-up scans. We show in this work that such longitudinal scans alone can be used as a form of 'free' self-supervision for training a deep network. We demonstrate this self-supervised learning for the case of T2-weighted sagittal lumbar Magnetic Resonance Images (MRIs). A Siamese convolutional neural network (CNN) is trained using two losses: (i) a contrastive loss on whether the scan is of the same person (i.e. longitudinal) or not, together with (ii) a classification loss on predicting the level of vertebral bodies. The performance of this pre-trained network is then assessed on a grading classification task. We experiment on a dataset of 1016 subjects, 423 possessing follow-up scans, with the end goal of learning the disc degeneration radiological gradings attached to the intervertebral discs. We show that the performance of the pre-trained CNN on the supervised classification task is (i) superior to that of a network trained from scratch; and (ii) requires far fewer annotated training samples to reach an equivalent performance to that of the network trained from scratch.