Key result
Soft-voting deep ensemble improves PTB-XL ECG classification by ~0.02 macro F1 over MiniRocket.
Why the study?
The optimal architecture for classifying 12-lead ECGs into diagnostic superclasses remains somewhat unsettled in the literature.
Does a soft-voting ensemble of deep learning models improve 12-lead ECG diagnostic superclass classification compared to single deep architectures and classical machine learning models on the PTB-XL dataset?
Does a soft-voting ensemble of deep learning models improve 12-lead ECG diagnostic superclass classification compared to single deep architectures and classical machine learning models on the PTB-XL dataset?
In classifying 12-lead ECGs on the PTB-XL dataset, a soft-voting deep learning ensemble and a classical MiniRocket-LightGBM model achieved nearly identical macro AUROC (0.916 vs 0.914) and accuracy, though lack of external validation precludes immediate clinical readiness.
Answers are grounded in this paper and cite their sources.
Supports ensemble ECG AI refinement; leaves open translation to clinical diagnostic gains.
Which architecture is best for sorting 12-lead ECGs into diagnostic superclasses still remains somewhat unsettled in the literature. In this study we compared six different models on PTB-XL (21,375 records) under one single patient-level split: a Mamba-style state space model (M1), a 1D residual CNN (M2), a BiLSTM (M3), a CRNN with attention (M4), a soft-voting ensemble of M2 to M4 (M5), and MiniRocket features with LightGBM (M6). Each record was labeled with the superclass of its highest-likelihood diagnostic code. We then scored the saved per-record predictions with paired patient-level bootstrap intervals, McNemar tests, per-class AUROC and AUPRC, sensitivity, specificity, and calibration. The ensemble reached the highest macro F1 (0.665, 95% CI 0.644-0.685), ahead of every single deep model by 0.020-0.041 (all paired CIs excluded zero) and ahead of M6 (0.645) by 0.021 (95% CI 0.002-0.040). We chose its members after seeing test results, so these margins are optimistic: ensembles built from validation-selected members, or from all four deep models, scored 0.656-0.657, about 0.009 lower (paired CIs included zero). M5 and M6 did not differ significantly in accuracy, and their macro AUROC was nearly identical (0.916 and 0.914). M1 trailed M2 and M3 by 0.021. Hypertrophy (HYP) was the hardest class (F1 0.33-0.41). Differences in HYP recall mostly reflected where each model's threshold sat: M6 had the lowest recall but the highest AUROC (0.871). We used only a single dataset and split, with no pretraining and no external validation, and so these results do not show that any model is ready for clinical use.
No takes yet. Share an insight, caveat, or question.
Bhatt et al. (2026) studied 12-lead ECG diagnostic superclass classification (n=21,375). Soft-voting ensemble of 1D residual CNN, BiLSTM, and CRNN with attention (M5) vs. Single deep models (Mamba-style SSM, 1D residual CNN, BiLSTM, CRNN with attention) and MiniRocket features with LightGBM (M6) was evaluated on Macro F1 score for ECG diagnostic superclass classification (MD 0.021, 95% CI 0.002-0.040). A soft-voting deep ensemble achieved the highest macro F1 on PTB-XL ECGs, outperforming MiniRocket with LightGBM.
Synapse has enriched 8 closely related papers on similar clinical questions. Consider them for comparative context: