Identification of Osteoporotic Changes of Vertebral Bodies on Computed Tomography Images Based on the Analysis of Groups of Textural Features

Abstract

The study presented in this paper is devoted to the development of technology for detecting osteoporotic changes in vertebral bodies according to computed tomography data based on the analysis of groups of textural features. This technology will be of interest to radiologists when assessing the structure of vertebral bodies in patients with osteoporosis, and especially in patients with malignant neoplasms, in which the disease proceeds with a decrease in bone mineral density. Automation of the analysis process will significantly reduce the time spent by a radiologist on the evaluation of vertebral bodies, including when evaluating the effectiveness of osteoporosis treatment. During the development process, the most informative group of signs was identified, which allows achieving high accuracy in detecting signs of osteoporotic changes in the vertebral bodies in 94.5%.