Key result
An ECG fusion model integrating features and waveforms detects OMI with a ~0.96 AUROC.
Why the study?
AI in ECG analysis has diverged into feature-based machine learning and raw waveform deep learning, but no framework integrates both to improve detection of acute coronary events.
Does a machine learning fusion model integrating ECG features and raw waveforms improve the diagnostic accuracy of acute coronary events in prehospital patients?
Cohort (n=7,397)
Yes
Does a machine learning fusion model integrating ECG features and raw waveforms improve the diagnostic accuracy of acute coronary events in prehospital patients?
A novel machine learning fusion architecture integrating traditional ECG features and raw waveforms significantly improves the diagnostic accuracy for acute coronary syndrome and occlusion myocardial infarction in prehospital settings.
Answers are grounded in this paper and cite their sources.
Promising for ACS detection via decision fusion but should not change practice; leaves open prospective validation before clinical adoption.
Background and Aims AI in electrocardiography (ECG) has diverged into two paths: traditional signal processing with machine learning, and deep learning of raw waveforms. The former preserves physiological interpretability but may miss novel patterns, while the latter overlooks a century of ECG science. No framework yet integrates both. This study introduces a fusion model combining handcrafted feature-based random forest and waveform-based convolutional neural network to improve diagnosis of acute coronary events. Methods This prospective, multicentre cohort study included prehospital ECGs recorded on emergency medical services from patients with a chief complaint of chest. Using this dataset, we developed four fusion frameworks that integrate ECG features with waveform information: (1) decision-level fusion combining predictions from random forest and CNN; (2) decision-level fusion with retraining; (3) feature-level fusion by concatenating CNN embeddings with ECG features; and (4) feature-level fusion combining CNN embeddings with tree-level outputs from the random forest. Results This study comprised of 10393 ECGs from 7397 patients. On both outcomes, ACS and OMI, the proposed decision fusion with retraining (Approach 2) achieved an AUROC of 0.878 (95% CI, 0.837–0.915) and 0.961 (95% CI, 0.934–0.983) on the test set, and an average precision score of 0.749 (95% CI, 0.681–0.810) and 0.850 (95% CI, 0.763–0.918), respectively, outperforming all other proposed fusion and individual models. Conclusions This study demonstrates the significance of fusion models that integrate complementary ECG representations to improve diagnostic accuracy in acute coronary events, highlighting a promising approach for real-time AI-augmented cardiovascular disease detection across diverse use cases.
No takes yet. Share an insight, caveat, or question.
Ji et al. (2026) conducted a cohort in Acute coronary events (ACS and OMI) (n=7,397). Decision-level fusion model with retraining (Approach 2) vs. Individual models and other fusion frameworks was evaluated on Detection of Acute Coronary Syndrome (ACS) and Occlusion Myocardial Infarction (OMI) (AUROC 0.878 for ACS and 0.961 for OMI, 95% CI 0.837-0.915 for ACS; 0.934-0.983 for OMI). A decision-level fusion model integrating ECG features and waveforms achieved an AUROC of 0.878 for ACS and 0.961 for OMI detection.
Synapse has enriched 10 related papers. Consider them for comparative context: