MRI-SegFlow: a deep learning-based unsupervised pipeline for vertebral segmentation of spinal MRI image

Abstract

Most deep learning–based vertebral segmentation methods require laborious manual labelling tasks. We aim to establish an unsupervised deep learning pipeline for vertebral segmentation of MR images. We integrate the suboptimal segmentation results produced by a rule-based method with a unique voting mechanism to provide supervision in the training process for the deep learning model. Preliminary validation shows a high segmentation accuracy achieved by our method without relying on any manual labelling. The clinical relevance of this study is that it provides an efficient vertebral segmentation method with high accuracy. Potential applications are in automated pathology detecion and vertebral 3D reconstructions for biomechanical simulations and 3D printing, facilitating clinical decision-making, surgical planning and tissue engineering.