3D Cascaded Convolutional Networks for Multi-vertebrae Segmentation.
Journal:
Current medical imaging
PMID:
32133953
Abstract
BACKGROUND: Automatic approach to vertebrae segmentation from computed tomography (CT) images is very important in clinical applications. As the intricate appearance and variable architecture of vertebrae across the population, cognate constructions in close vicinity, pathology, and the interconnection between vertebrae and ribs, it is a challenge to propose a 3D automatic vertebrae CT image segmentation method.