Shape characterization of the gonarthrosis in the X-ray images

Abstract

Gonarthrosis is a degenerative disease of the knee joint that involves damage of the articular cartilage, formation of the osteophytes and reactive changes in the synovial membrane and in the synovial liquid. Location of the initial change in the knee joint is unknown, i. e. initial change can occur anywhere. Diagnosis of the gonarthrosis is based on the use of the clinical and radiological methods. In this study, X – ray images were used. Resolution and dimensions of the X – ray images can vary and influence precision in their reading and analysis. Taking into account a possible complexity of images, it is necessary to perform a few processing steps in order to obtain measurable information. In order to eliminate imperfections of images, a non-linear median filter is used for image filtering. The segmentation of the characteristic regions in the image is done by using active contour segmentation which is based on the curve flow, curvature and contour of the desired region. Therefore, the quantification of the obtained information is necessary in order to perform precise classification of the gonarthrosis grade. Quantification of the segmented regions was carried out by measuring space between femur and tibia and by comparing it with measured space between femur and tibia in healthy persons. The precise diagnosis of this disease is of great importance to preserve objectivity during decision making, for further treatment and to facilitate the performance of everyday activities of the patients.