Key result
CNN-Transformer ensemble identifies myocardial infarction and ischemia on ECG with ~94% AUC.
Why the study?
Silent myocardial ischemia in diabetic patients often goes undetected, and wearable ECG devices record only two or three leads, reducing sensitivity for ischemic events. An optimal 12-lead ECG teacher model was needed to provide a high-capacity source for knowledge distillation to train lightweight wearable models.
Does an ensemble of CNN-Transformer hybrid deep learning models improve 12-lead ECG classification performance for detecting myocardial infarction and ischemia in the PTB-XL dataset?
Observational (n=18,885)
Does an ensemble of CNN-Transformer hybrid deep learning models improve 12-lead ECG classification performance for detecting myocardial infarction and ischemia in the PTB-XL dataset?
Answers are grounded in this paper and cite their sources.
This study develops a reproducible CNN-Transformer ensemble teacher model that achieves 93.97% AUC for detecting myocardial ischemia and infarction on 12-lead ECGs, providing a calibrated foundation for knowledge distillation into lightweight wearable device models.
Background: Diabetes affects more than 537 million adults worldwide. Cardiovascular complications cause about half of all deaths in this population. Silent myocardial ischemia is common in diabetic patients. Autonomic neuropathy often hides the typical chest pain. As a result, ischemia often goes undetected until it becomes severe. Wearable ECG devices usually record only two or three leads. This reduces sensitivity for detecting ischemic events. Objective: We aimed to find an optimal teacher model for 12‑lead ECG classification of myocardial infarction (MI) and ischemia. Our goal was to provide a high‑capacity source for knowledge distillation. This source can train lightweight models for wearable devices. Methods : We performed a systematic hyperparameter search on the PTB‑XL dataset (21,837 records). We tested five architectures with different Transformer depths, hidden dimensions, dropout rates, and learning rates. We also trained two extended variants (V2 and V3) with higher positive‑class weights and longer schedules. We built ensembles by averaging the logits of the best models. Results : The wider transformer model (hidden dimension 1024, 3 layers, dropout 0.20) achieved 86.58% validation accuracy and 80.74% sensitivity. Extended training (V2: weight 3.0, 12 epochs) raised validation sensitivity to 88.82% with only a small accuracy drop (85.39%). The ensemble of V1 and V2 gave the best test performance: accuracy 87.40%, sensitivity 76.18%, specificity 92.65%, F1 79.41%, and AUC 93.97%. Adding V3 did not improve performance. Conclusions : The V1+V2 ensemble is a well‑calibrated teacher model for knowledge distillation. Its strong discriminative ability (AUC 93.97%) makes it suitable for training lightweight student networks. These networks can be used in reduced‑lead, resource‑limited settings. All code and data are publicly available.
No takes yet. Share an insight, caveat, or question.
Soltanzadeh et al. (2026) conducted an observational in Myocardial infarction and myocardial ischemia (n=18,885). CNN-Transformer hybrid ensemble model (V1+V2) vs. Single baseline and extended CNN-Transformer configurations was evaluated on Test set classification performance (AUC and accuracy) for myocardial infarction and ischemia vs. other findings. The V1+V2 CNN-Transformer ensemble achieved 87.40% accuracy, 76.18% sensitivity, 92.65% specificity, and 93.97% AUC for classifying myocardial infarction and ischemia on 12-lead ECGs.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: