Computed tomography vertebral segmentation from multi-vendor scanner data

Abstract

Automatic medical image segmentation is a crucial procedure in computer-aided surgery. However, the performance of automatic segmentation algorithms highly depends on the consistent properties of medical images. To address this issue, we propose a network for standardizing computed tomography (CT) images, aiming for the optimal performance of spine segmentation. We compare the synthesized and re-windowed images for performance assessment in terms of structural similarity and segmentation performance.