AI-Driven Optimization of Cloud Resource Allocation for Personalized Medical Imaging in Hospitals: A Case Study from a Major Medical Center

This study is to introduce a cloud-based, AI-driven architecture that can address the above-mentioned challenges properly and demonstrate its feasibility through real-scale implementations and practical testing in a major traditional hospital in Asia that typically operates a variety of diagnostic imaging services. A typical scenario is that vital organ medical images need immediate attention and response from any physician and hospital. CT brain imaging and PET-CT, Gallium imaging of the whole body, glucose, and myocardial ischemia will be implemented and tested in this project. These diagnostic imaging devices usually generate cross-sectional images of patients and typically produce an average of 500 - 600 high-resolution images each. In the local hospital, the human eye ballot reading and deep learning mean image classifications take a considerable amount of time to present the results. This project will take advantage of the cloud, AI, and web-based technology to rapidly assist physicians in providing better medical services. The research objectives of this project are summarized as follows: (1) Gaining experience and lessons by deploying AI technology on a real hospital scale. AI/HA medical imaging services operated in the cloud can be made available for other hospitals to use.