The radiation dose and the associated lifetime risk of developing cancer is a problem justifying the use of surface topography (ST) as an alternative to radiographs for monitoring scoliosis. The current ST methods mainly rely on marker placement and fail to describe the 3D aspect of the scoliosis deformity. This thesis describes research and development of a novel method for assessing, quantifying, and monitoring the torso deformities based on their symmetry and curvature.

We present a new 3D asymmetry analysis that discloses the area of abnormality in a geometry using contours of deviation. Its application in biomedical image analysis of torso to quantifying and monitoring torso deformity is described. The deformity pattern of the analyzed torsos are categorized using the classification system. The reliability of the classification system along with its correlation to the radiograph measurement are investigated.

Internal characteristics observed in radiographs such as magnitude, number, and location of the scoliosis curve are estimated using the ST measurements. The difference of ST measurements between two successive acquisitions, with one year gap between, is used to determine the change of the thoracic/thoracolumbar and lumbar scoliosis curves. The capability of the ST measurements in classifying the scoliosis curves into progression and non-progression group is assessed with the aim of replacing some of the X-ray acquisition used to monitor patients with adolescent idiopathic scoliosis.