Technical Name AI-Enhanced Smart Detection System for Coronary Artery Disease
Project Operator National Health Research Institutes
Project Host 司徒惠康
Currently, detecting CAD in asymptomatic patients before myocardial infarction is challenging. This technology holds three patents, including signal processing, feature engineering, and the use of deep learning and machine learning to provide interpretability of features, improve model accuracy, and expand the input file format, enhancing its applicability. In terms of performance, the algorithm has achieved an 84%-90% accuracy of AUC in detecting asymptomatic CAD patients.
Scientific Breakthrough
This AI-based technology enhances the diagnosis of CAD patients using ECG. About 13% of patients with coronary artery stenosis, as confirmed by coronary angiography, have normal ECGs classified traditional methods. However, this technology can identify 80% of these high-risk asymptomatic patients. Our AI algorithm achieves a performance of 84%-90% accuracy of AUC. Therefore, this technology can detect coronary artery stenosis at early stage and prevent the occurrence of myocardial infarction.
Industrial Applicability
The estimated market size for ECG is projected to reach $10.3 billion USD by 2024. AI has been improving the accuracy and efficiency of ECG interpretation. As a result, the market potential for early detection of cardiovascular diseases using ECG continues to grow in various areas, including telemedicine care, standard equipment in ambulances, value-added services in health checkups, and the development of cardiovascular insurance. There are at least billions of dollars in Taiwan markets.
Keyword Electrocardiography Artificial intelligence Coronary artery disease Myocardial infarction Cardiovascular disease
  • Contact
  • Yi-Ju Chou
other people also saw