A novel statistical morphometry imaging method for differentiating long bone geometry: methodological development and application with adolescent idiopathic scoliosis (AIS) patients.

Non-invasive quantification of bone shape is crucial in orthopaedic research. The primary objective of this study was to develop an automated statistical morphometry method for comparing the cross-sectional images of normal and diseased bones. The secondary objective involved demonstrating the effectiveness of the proposed method in distinguishing AIS patients from normal controls. This framework is composed of bone segmentation followed by measurements of maximum and minimum bone diameters, inter-group and intra-group statistical morphometry, and statistical analysis of bone thickness. The proposed framework was applied to detect bone morphological abnormality in adolescent idiopathic scoliosis (AIS) patients. The forearm bones in cross-sectional peripheral quantitative computed tomography (pQCT) images from 23 AIS patients and 16 normal controls were analyzed. The radius outer contour was found to be rounder and the radius cortical bone was thinner in AIS patients compared to normal controls.