Survey of Advances in Cobb Angle Measurement for Automatic Spine Detection in X-Ray

Abstract

Chronic back pain is a bending-induced malformation of the human spinal column that can cause severe pain as well as cosmetic and pulmonary issues. The external appearance of a human back in scoliosis is generally the reflection of internal deformation. Spinal curvature is usually measured in degrees using the Cobb angle, the standard method for evaluating scoliosis patients. This article highlights the review of earlier research articles on scoliosis to provide insight into the existing knowledge, which aids in the robust identification and monitoring of scoliosis. However, many researchers have worked in this field for many decades yet there is no reliable, easily available, and universal tool for Cobb angle estimation. Hence, the present article enlightens the existing information and the lacunae in the field to aid further scope for research opportunities available for future consideration. Using RGB and complexity photos collected by an RGB-complexity device Microsoft, a modified convolutional network (MCN) named fuse-Unet is the proposal to provide automatic recognition of the human spine area and which was before the imaging route. A normal-vector-based approach and two force sensors are used to ensure that the probe fits the spine area well a 6-degree-of-freedom robotic arm in the role of a doctor who completes the automatic scanning along the pre-planned path. Furthermore, Cobb angles for morphological structural analysis of the spine are determined using 3-D ultrasound modeling and scanning of the spine. The suggested system's performance is evaluated using phantom and in vivo tests.