The spring scenery of West Lake returns, and the spring water is dyed green. In April, Hangzhou welcomed the annual grand event in the field of health management - the 13th "West Lake Health Forum" - amidst the bright spring scenery. Top health management experts and scholars from home and abroad gathered to deeply analyze industry development trends and the latest concepts from a cutting-edge perspective, fully showcasing and exchanging new technologies, practices, and progress, injecting new impetus into the development of health management.

At the "AI Empowered Health Management" sub forum held on the afternoon of April 26th, Professor Sun Yunfan from the Department of Hepatology and Oncology at Zhongshan Hospital Fudan University presented a wonderful report titled "Intelligent Fusion and Precise Exploration: Research and Practice of AI Driven miRNA7 Multimodal Large Model". Focusing on the core issues of early diagnosis and treatment of liver cancer, the miRNA7 model and the achievements of scientific research transformation and clinical application were systematically demonstrated. Through innovative integration of artificial intelligence and molecular detection technology, a breakthrough solution was provided for early diagnosis and treatment of liver cancer.

Professor Sun clearly pointed out the severe situation of early diagnosis and treatment of liver cancer in China. 70%-80% of liver cancer patients are diagnosed in the middle and late stages, and the 5-year survival rate after surgery is relatively low; If patients can undergo radical surgery in the early stages of liver cancer, the 5-year survival rate after surgery is expected to significantly increase to over 70%. So early screening and diagnosis are key breakthroughs in improving patient prognosis. However, current early diagnostic methods for liver cancer are limited and rely heavily on traditional testing methods such as alpha fetoprotein, ultrasound, and CT. He further pointed out that from a molecular biology perspective, the changes in miRNA molecules during the progression of liver cancer occur much earlier than pathological morphological changes. The diagnosis model of liver cancer based on seven kinds of miRNAs in plasma, established by the team of academician Fan Jia of Zhongshan Hospital Fudan University, has been optimized and verified by thousands of large samples and multi center studies, and can accurately identify liver cancer, healthy people, chronic hepatitis B and cirrhosis patients. MiRNA7, as the first liver cancer molecular detection product approved by NMPA, can significantly improve the detection rate of very early liver cancer and has now become a first-line liver cancer screening method in China.

Professor Sun vividly demonstrated the clinical value brought by technological innovation through typical cases - a patient showed miRNA7 positive and AFP negative in three consecutive physical examinations. After 18 months of follow-up, a 2cm hepatocellular carcinoma (HCC) tumor was discovered and minimally invasive resection was performed in a timely manner. The patient was cured without the need for adjuvant therapy after surgery, greatly advancing the prevention and treatment of liver cancer. MiRNA7 has also demonstrated excellent diagnostic performance in clinical real-world data, with a detection rate for very early liver cancer that is 30% higher than traditional methods such as AFP. Professor Sun also stressed that with the improvement of countrymen's living conditions, the proportion of liver cancer caused by metabolic and fat related liver diseases has increased significantly, which undoubtedly brings higher requirements to the health management of liver diseases and community screening of liver cancer, and also highlights the urgency of developing more efficient and accurate screening tools with the help of artificial intelligence technology.

Subsequently, Professor Sun emphasized the innovative breakthrough of AI in early diagnosis of liver cancer - the miRNA7 multimodal fusion model. This model integrates patient basic information, traditional serological markers, and 7 plasma miRNAs to construct a multimodal fusion algorithm for analysis, thereby achieving accurate diagnosis. From 2020 to 2022, nearly a thousand large sample studies (covering HBV related HCC, chronic liver disease, and healthy populations) showed that the GALADM model was superior to single indicator, single combination of previous indicators, and GALAD model in distinguishing HCC from non-HCC populations in the very early and early stages of liver cancer. This model is based on the optimal early diagnosis plan for liver cancer discovered through the integration of hospital big data, which compensates for the limitations of single detection and fully reflects the medical value of multi-dimensional data complementarity.

Professor Sun finally pointed out that for extremely early liver cancer, it is more necessary to use a liver cancer molecular typing system integrated with multiple indicators and multimodal AI to provide more clues about the risk of liver cancer occurrence for clinical practice. For the field of health management, this technology has broad application prospects in physical examination scenarios, which can help with precise screening of high-risk populations, early warning and hierarchical management of asymptomatic populations. With the continuous iteration of AI technology, we look forward to seeing the integration of more dimensional data in the future, including the deep fusion of genomics and imaging information, as well as the construction of dynamic analysis and evaluation, in order to promote the development of liver cancer prevention and treatment towards more accurate and personalized directions.

With the deepening of the "Healthy China 2030" strategy, health management is transforming from "disease treatment" to "full life cycle protection". Dunwill Medical adheres to its original intention and is committed to making cutting-edge achievements truly benefit billions of people, promoting the deep integration of industry, academia and research, and providing solid technical support for the "Healthy China 2030" strategy. Based on the successful clinical practice of the miRNA7 multimodal fusion model, we look forward to working with more health management experts to provide smarter and more accurate solutions for the comprehensive management of liver disease and early warning of liver cancer, and to protect the light of life with the power of cutting-edge science!