Abstract—Early detection of rheumatoid arthritis is very important for its treatment. However, it can be difficult to detect changes in medical conditions by visually inspecting medical images. Computer-based applications that can support doctors’ diagnoses can be helpful. In this study, we propose a diagnostic application based on computer-based recognition of features in medical images. Specifically, joint learning is performed by using Haar-like features those capture differences in brightness of joints. Furthermore, we aimed to improve detection accuracy by removing false positives based on pixel values and positional relationships of joint detection results. As a case study, we apply it to the detection of third finger joints in X-ray images of hands. The application is able to correctly identify these regions in most cases, thereby aiding doctors in the early detection of rheumatoid arthritis.

Index Terms—rheumatoid arthritis, third finger joint detection, joint space

Cite: Kosuke Goto, Tomio Goto, and Koji Funahashi, "Automatic Joint Part Detection Method for Joint Space Measurement," Journal of Image and Graphics, Vol. 9, No. 3, pp. 103-108, September 2021. doi: 10.18178/joig.9.3.103-108

Copyright © 2021 by the authors. This is an open access article distributed under the Creative Commons Attribution License (CC BY-NC-ND 4.0), which permits use, distribution and reproduction in any medium, provided that the article is properly cited, the use is non-commercial and no modifications or adaptations are made.