Lower back MRI region-based segmentation for computer-aided systems

Abstract

Vertebral column (spine) and intervertebral disc (IVD) diseases are major sources of lower back pain. Exquisite segmentation of the spinal anatomy is salutary and substantial so as to be utilized for computer-aided diagnosis and computer-assisted surgery systems. A modified region-based segmentation approach was applied to magnetic resonance (MR) sagittal and axial plane images of different lumbosacral spine patients. It confers distinguished mean dice similarity coefficient of 0.99755 with low standard deviation of about 0.001 for axial plane IVD segmentation.