Fig. 1From: A spine segmentation method based on scene aware fusion networkSpine parsing results. Spine parsing refers to the multi-class segmentation of both the vertebrae and intervertebral discs, whereby each individual vertebra or intervertebral disc is assigned its own unique label. The letters T, L, and S are used to represent thoracic, lumbar, and sacral vertebrae, respectively. (BG represents the background.)Back to article page