Our research paper, entitled “Towards AI-driven radiology education: A self-supervised segmentation-based framework for high-precision medical image editing” has been accepted at the top medical AI conference, the 26th International Conference on Medical Image Computing and Computer Assisted Intervention (MICCAI 2023). The paper is among the top 14% that receives early acceptance.
Medical education is essential for providing the best patient care in medicine, but creating educational materials using real-world data poses many challenges. For example, the diagnosis and treatment of a disease can be affected by small but significant differences in medical images; however, collecting images to highlight such differences is often costly.
Our accepted paper proposes a novel educational medical image editing system, which allows users to create their intended disease characteristics. We present a novel algorithm to edit anatomical elements using segmentation labels acquired through self-supervised learning. The self-supervised segmentation achieves pixel-wise clustering under the constraint of invariance to photometric and geometric transformations, which are assumed not to change the clinical interpretation of anatomical elements. The user then edits the segmentation map to produce a medical image with the intended detailed findings. Evaluation by five expert physicians demonstrated that the edited images appeared natural as medical images and that the disease characteristics were accurately reproduced.
We believe the potential impact of our research will change the landscape of medical education, making it more accurate, less expensive, and more accessible to improve patient care worldwide.
- Authors:
Kazuma Kobayashi (National Cancer Center Research Institute, RIKEN Center for Advanced Intelligence Project)
Lin Gu (RIKEN Center for Advanced Intelligence Project)
Ryuichiro Hataya (RIKEN Information R&D and Strategy Headquarters, National Cancer Center Research Institute)
Mototaka Miyake (National Cancer Center Hospital)
Yasuyuki Takamizawa (National Cancer Center Hospital)
Sono Ito (National Cancer Center Hospital)
Hirokazu Watanabe (National Cancer Center Hospital)
Yukihiro Yoshida (National Cancer Center Hospital)
Hiroki Yoshimura (Hiroshima University School of Medicine)
Tatsuya Harada (The University of Tokyo, RIKEN AIP)
Ryuji Hamamoto (National Cancer Center Research Institute, RIKEN Center for Advanced Intelligence Project)
