Algorithm for Evaluation of Vertebral Body Deformity for Detection of Osteoporosis Using Computed Tomography Data

The paper presents a method of determining the degree of deformation of vertebral bodies to detect osteoporosis from computed tomography data based on an algorithm for evaluating their geometric features. The results of the algorithm’s work on real images are presented. The study revealed high separability of the features of compactness, shape, Haralick and Blair–Biss factors, as well as the contour shape feature for the classification of normal and pathologic vertebrae. The developed algorithm improves the accuracy of osteoparosis diagnosis and reduces the time required for image analysis.