A team led by Dr. Yen-Tze Liu, Director of the Big Data Center at Changhua Christian Hospital, has developed an intelligent decision-support system for screening coronary artery stenosis. This system can predict high-risk groups for coronary artery stenosis using only basic health checkup data, enabling physicians to provide personalized preventive medical advice, achieving the goal of early detection and early intervention.
Coronary artery stenosis is a major cause of myocardial infarction and other serious cardiovascular diseases. Traditional risk assessment methods, such as the Framingham Risk Score, are often used to predict future cardiovascular disease risks, but their accuracy varies significantly for Asian populations. While coronary computed tomography angiography (CCTA) can detect coronary artery stenosis, it is costly and carries certain risks, making it unsuitable for widespread screening among the general population.
Health Check Data for Cardiovascular Risk Prediction: Evolutionary Learning for Personalized Risk Assessment
To overcome these challenges, the intelligent decision-support system for coronary artery stenosis screening utilizes innovative evolutionary learning techniques. Using basic health checkup data, such as personal information, medical history, lifestyle questionnaires, and blood test results, the system creates a personalized predictive model to estimate the risk of coronary artery calcification or stenosis. This allows physicians to provide more precise health education and recommendations based on individualized risk information.
The core technology of the system, the Bio-Evolutionary Learning Platform, automatically selects key indicator combinations (e.g., blood pressure, cholesterol, age, gender, family medical history) from vast amounts of health checkup data, building the most accurate model for predicting coronary artery stenosis risk. Empirical data shows that for high-risk individuals with a score greater than 90, the system achieves an 80% positive predictive value, outperforming traditional methods. The system is now integrated into a cloud-based prediction platform, allowing for both single-case and batch input, meeting clinical demands in hospitals.
Clinical Validation and Product Launch: Advancing Precision Preventive Medicine
As a family medicine physician with a focus on preventive medicine, Dr. Liu highlighted that this system offers a powerful tool for the early detection and prevention of coronary artery disease. It not only assists physicians in making more precise screening recommendations but also enhances the convenience and accessibility of screenings while offering educational value. Moving forward, the system will undergo clinical validation at Changhua Christian Hospital and its six branch hospitals, with initial efforts focused on model validation and optimization. In later stages, the product will be launched to provide innovative healthcare services. Dr. Liu expressed hope that, with the aid of AI technology, this system will usher in a new era of preventive medicine, enabling early detection and intervention, and effectively reducing the health risks posed by cardiovascular diseases.
Resource (mandarin): AI幫建個人化模型 基礎健檢即能預測冠狀動脈狹窄！
