Triple Model Framework for Medical Diagnosis: Assist in the Decentralization of Medical Resources
C.G. Wang, Hongjun Wang, Wei Li, Shanshan Wang, Jingyan Wang, Kegang Ji · IEEE Access · 2025
The extreme imbalance of medical resources in the world today makes it crucial to achieve the the decentralization of medical resources. However, whether it is the quality of doctors or medical equipment, small hospitals in towns and villages are unable to meet the medical needs of the vast rural population. Therefore, under the premise of existing township medical resources, it is crucial to effectively improve the diagnostic level of doctors. For this purpose, we use inexpensive X-rays as the diagnostic basis and use osteoarthritis as the pathological sample to train an AI diagnostic model to assist township doctors in diagnosis. However, there are three issues with existing AI diagnostics: 1). The disconnect between AI diagnostics and real-world use. 2). The data between various hospitals has not been effectively utilized. 3). The data of the entire society has not been effectively utilized. For this purpose, we have designed Triple Model Framework for Medical Diagnosis with Federated Learning Based on Blockchain (TMMD-FLchain) to address the above three issues. In addition, we have designed Reward mechanism based on expected value (RM-EV) for anonymous federated learning.