Title: | A deep registration method for accurate quantification of joint space narrowing progression in rheumatoid arthritis |
Authors: | Wang, Haolin Browse this author |
Ou, Yafei Browse this author |
Fang, Wanxuan Browse this author |
Ambalathankandy, Prasoon Browse this author |
Goto, Naoto Browse this author |
Ota, Gen Browse this author |
Okino, Taichi Browse this author |
Fukae, Jun Browse this author |
Sutherland, Kenneth Browse this author →KAKEN DB |
Ikebe, Masayuki Browse this author →KAKEN DB |
Kamishima, Tamotsu Browse this author →KAKEN DB |
Keywords: | Rheumatoid arthritis |
Joint space narrowing |
Image registration |
Deep learning |
Radiology |
Computer-aided diagnosis |
Issue Date: | Sep-2023 |
Publisher: | Elsevier |
Journal Title: | Computerized Medical Imaging and Graphics |
Volume: | 108 |
Start Page: | 102273 |
Publisher DOI: | 10.1016/j.compmedimag.2023.102273 |
Abstract: | Rheumatoid arthritis (RA) is a chronic autoimmune inflammatory disease that leads to progressive articular destruction and severe disability. Joint space narrowing (JSN) has been regarded as an important indicator for RA progression and has received significant attention. Radiology plays a crucial role in the diagnosis and monitoring of RA through the assessment of joint space. A new framework for monitoring joint space by quantifying joint space narrowing (JSN) progression through image registration in radiographic images has emerged as a promising research direction. This framework offers the advantage of high accuracy; however, challenges still exist in reducing mismatches and improving reliability. In this work, we utilize a deep intra-subject rigid registration network to automatically quantify JSN progression in the early stages of RA. In our experiments, the mean-square error of the Euclidean distance between the moving and fixed images was 0.0031, the standard deviation was 0.0661 mm and the mismatching rate was 0.48%. Our method achieves sub-pixel level accuracy, surpassing manual measurements significantly. The proposed method is robust to noise, rotation and scaling of joints. Moreover, it provides misalignment visualization, which can assist radiologists and rheumatologists in assessing the reliability of quantification, exhibiting potential for future clinical applications. As a result, we are optimistic that our proposed method will make a significant contribution to the automatic quantification of JSN progression in RA. Code is available at https://github.com/pokeblow/Deep-Registration-QJSN-Finger.git. |
Rights: | © <2023>. This manuscript version is made available under the CC-BY-NC-ND 4.0 license http://creativecommons.org/licenses/by-nc-nd/4.0/ |
http://creativecommons.org/licenses/by-nc-nd/4.0/ |
Type: | article |
URI: | http://hdl.handle.net/2115/92807 |
Appears in Collections: | 量子集積エレクトロニクス研究センター (Research Center for Integrated Quantum Electronics) > 雑誌発表論文等 (Peer-reviewed Journal Articles, etc)
|