ROFI: A Deep Learning-Based Ophthalmic Sign-Preserving and Reversible Patient Face Anonymizer
Year: 2025
Type: preprint
Abstract: Patient face images provide a convenient mean for evaluating eye diseases, while also raising privacy concerns. Here, we introduce ROFI, a deep learning-based privacy protection framework for ophthalmology. Using weakly supervised learning and neural identity translation, ROFI anonymizes facial features while retaining disease features (over 98\% accuracy, $κ> 0.90$). It achieves 100\% diagnostic sensitivity and high agreement ($κ> 0.90$) across eleven eye diseases in three cohorts, anonymizing over 95\% of images. ROFI works with AI systems, maintaining original diagnoses ($κ> 0.80$), and supports secure image reversal (over 98\% similarity), enabling audits and long-term care. These results show ROFI's effectiveness of protecting patient privacy in the digital medicine era.
Source: arXiv (Cornell University)
Authors: Tian, Yuan, Zhou, Min, Chen, Yitong, Li, Fang, Qi, Lingzi, Wang, Shuo, Xu, Xieyang, Yu, Yu, Xu, Shiqiong, Lei, Chaoyu, Jiang, Yankai, Zhang, Rongzhao, Tan, Jia, Wu, Li, Chen, Hong, Liu, Xiaowei, Lu, Wei, Li, Lin, Zhou, Huifang, Song, Xuefei, Zhai, Guangtao, Fan, Xianqun
Institutions: Shanghai Jiao Tong University, Shanghai Ninth People's Hospital, Shanghai Artificial Intelligence Laboratory, Central South University, Xiangya Hospital Central South University, Renmin Hospital of Wuhan University, Huazhong University of Science and Technology, Chinese Academy of Medical Sciences & Peking Union Medical College, Peking Union Medical College Hospital, Dalian Medical University, Second Affiliated Hospital of Dalian Medical University
Language: English
Subfield: Artificial Intelligence
Field: Computer Science
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W7091186309
Cites:
Cited by:
Related to: