HKUST's medical ChatGPT aims to save diagnosis time by 40%
During a recent media briefing, Assistant Professor Chen Hao from the HKUST Department of Computer Science presented MedDr, an innovative AI tool that acts as a versatile medical generalist, similar to a "medical GPT." This advanced tool can generate reports and offer preliminary diagnoses based on medical images submitted by users for analysis.
Chen Hao emphasized that the primary goal of MedDr is to reduce doctors' workload by up to 40%, thus enhancing their overall efficiency and allowing them to focus on more critical tasks. The university plans to collaborate with local hospitals and medical professionals to ensure the tool's accuracy and effectiveness.
According to HKUST, MedDr stands out as one of the most extensive open-source software projects in the field of general medicine. A recent study conducted by the Shanghai AI Laboratory recognized MedDr as one of the top-performing generalist models currently available.
Enhancing Cancer Diagnosis with MOME
Another remarkable AI model, known as MOME, specializes in diagnosing breast cancer, which is among the most prevalent types of cancer affecting women in Hong Kong. By analyzing patients' breast MRIs, MOME aids doctors in accurately differentiating between benign and malignant breast tumors, potentially avoiding unnecessary biopsies.
Moreover, MOME has the capability to predict patient responses to chemotherapy, assisting healthcare providers in devising personalized treatment plans. Chen Hao highlighted that the model's accuracy level is comparable to that of experienced radiologists, exceeding 90%. This feat was accomplished through extensive collaboration with five hospitals in mainland China and the analysis of over 10,000 patient cases.
Advancements in AI Medical Systems
In addition to MedDr and MOME, the suite of AI models includes pathology assistant tools like mSTAR and XAIM, which offer insights into how AI systems make medical decisions. HKUST emphasized that the research and development of these AI models were conducted utilizing the university's robust AI computing resources, enabling these systems to process vast amounts of data efficiently and achieve exceptional performance.
During the media briefing, Provost Guo Yike highlighted this initiative as a pivotal step towards revolutionizing healthcare practices. He mentioned that HKUST is actively working towards establishing Hong Kong's third medical school, with a core focus on integrating AI technologies into modern healthcare as a key direction for the university's future endeavors.
