Related Experiment Video
Updated: Jun 28, 2026
07:45
Acute Myocardial Infarction in Rats
Published on: February 16, 2011
An Artificial Intelligence based model for predicting long-term all-cause mortality after acute Myocardial Infarction
Linghan Xue1, Wenmiao Wang2, Qianli Zhao3
1Department of Cardiology, Fuwai Hospital, National Center for Cardiovascular Diseases, Peking Union Medical College and Chinese Academy of Medical Sciences, No. 167 North Lishi Road, Xicheng District, Beijing 100037, China.
European Heart Journal. Digital Health
|June 18, 2026
Summary
Predicting long-term mortality after acute myocardial infarction (AMI) is difficult. An Artificial Intelligence-based model (AIMI) shows superior accuracy in predicting long-term all-cause mortality compared to traditional risk scores.
Area of Science:
- Cardiology
- Artificial Intelligence
- Predictive Analytics
Background:
- Predicting long-term mortality after acute myocardial infarction (AMI) is crucial for patient management but remains challenging.
- Existing clinical risk scores have limitations in accurately stratifying long-term risk post-AMI.
Purpose of the Study:
- To develop and validate an Artificial Intelligence-based model, termed the AIMI model, for predicting long-term all-cause mortality following AMI.
- To compare the performance of the AIMI model against established clinical risk scores.
Main Methods:
- The AIMI model was developed using Random Forest (RF) on a dataset of 4825 AMI patients who underwent emergent coronary angiography or PCI.
- The model incorporated 15 variables and was validated externally on 723 AMI patients.
- Performance was assessed using time-dependent ROC curves, Kaplan-Meier analyses, and Brier scores, comparing AIMI against GRACE and TIMI risk scores.
Main Results:
- The AIMI model demonstrated robust performance with a C-index of 0.81 in the test set.
- AIMI outperformed GRACE and TIMI risk scores in predicting mortality across short-, mid-, and long-term periods, particularly for long-term outcomes (1, 3, and 5 years).
- External validation confirmed the model's good generalization and robustness, with strong discrimination between risk groups (all P < 0.05) and good individual-level performance (Brier scores ~0.021-0.022).
Conclusions:
- The AI-based AIMI model significantly surpasses traditional methods for predicting long-term all-cause mortality after AMI.
- The AIMI model shows potential to enhance risk stratification and guide post-discharge management for AMI patients.
