Automated diagnosis of disc bulge and disc desiccation from lumbar mri using concatenated shape and texture features

Abstract

The disc degeneration is the most common abnormalities occurringin the lumbar spine which leads to severe low back pain. It is one of the mostsuffered problem by a large population. Among many imaging modalities, the Magnetic Resonance Imaging (MRI) is more sophisticated and preferredmodality for viewing the lumbar region. It shows that the soft tissues arebetter than X-ray and Computer Tomography (CT) modalities. From the MRI,the difference between the normal and the degenerated InterVertebral Discs(IVDs) are visible to the human eyes. To classify the degeneration, aradiologist s help is needed and it can also be a time consuming process. So,an automated diagnosis system is proposed to classify and to diagnose thedisc bulge and the disc desiccation. An automated diagnosis system consistsof four steps, namely localization, segmentation, feature extraction andclassification. newline