PulseExploreJournal ClubDebatesTrendingResearchersJournals
Instagram
HomeExploreJournal ClubTrending
Synapse
⌘+K
Synapse
February 8, 2026Scientific Reports0 citationsOpen Access

A hybrid learning framework for automated multiclass electrocardiogram classification with SimCardioNet

MMMuhammad Dawood MajidMAMuhammad AnwarSBSyed Fakhar Bilal

Key Result

SimCardioNet achieved 0.921 accuracy and 0.921 F1-score on the PTB-XL dataset, outperforming current state-of-the-art models for automated multiclass electrocardiogram classification.

Key Points

  • The aim is to improve the accuracy of electrocardiogram classification using a hybrid learning framework that combines self-supervised and supervised techniques.
  • Developed SimCardioNet, a multi-scale convolutional neural network with attention mechanisms.
  • Implemented a modified SimCLR approach for self-supervised pretraining followed by supervised fine-tuning.
  • Evaluated performance on three ECG datasets, including clinical and benchmark datasets.
  • Achieved 0.975 accuracy on the clinical ECG dataset with high precision and recall.
  • Demonstrated perfect classification on an external dataset, indicating excellent generalization.
  • Outperformed existing models with a 0.921 accuracy on the PTB-XL dataset, highlighting robustness.

Structured PICO

Does SimCardioNet improve multiclass ECG image classification accuracy compared to existing state-of-the-art models?

P
Population
Three distinct ECG image datasets: (1) a 4-class Pakistani clinical ECG dataset (Dataset I), (2) an external Kaggle electrocardiography dataset for out-of-distribution validation (Dataset II), and (3) the large-scale PTB-XL benchmark (Dataset III) covering five diagnostic superclasses.
I
Intervention
SimCardioNet, a hybrid self-supervised and supervised deep learning framework leveraging a custom multi-scale convolutional neural network backbone enhanced with residual connections and multi-head self-attention, pretrained via a modified SimCLR contrastive learning strategy.
C
Comparator
Current state-of-the-art models including dual-branch CNNs, entropy-enhanced CNNs, and Bi-GRU architectures.
O
Outcome
Classification accuracy, precision, recall, and F1-score.surrogate

SimCardioNet, a hybrid self-supervised and supervised deep learning framework, achieves high accuracy in multiclass ECG image classification, demonstrating strong generalization and potential to reduce reliance on labeled data in resource-constrained settings.

Limitations

  • Intrinsic complexity and inter-class similarity of ECG morphologies, particularly between pathological categories
  • Computational intensity of deep learning approaches

Abstract

Electrocardiography is a cornerstone in the diagnosis of cardiovascular diseases; however, accurate interpretation demands expert knowledge and is often impeded by data scarcity and annotation costs. To address these challenges, we propose SimCardioNet, a hybrid self-supervised and supervised deep learning framework for multi-class electrocardiography image classification. SimCardioNet leverages a custom multi-scale convolutional neural network backbone enhanced with residual connections and multi-head self-attention, pretrained via a modified SimCLR contrastive learning strategy that integrates a hybrid loss combining InfoNCE and cosine similarity. Following self-supervised pretraining, the model undergoes supervised fine-tuning with progressive layer unfreezing to mitigate overfitting and preserve meaningful representations. We evaluate SimCardioNet across three distinct ECG image datasets: (1) a 4-class Pakistani clinical ECG dataset (Dataset I), (2) an external Kaggle electrocardiography dataset for out-of-distribution validation (Dataset II), and (3) the large-scale PTB-XL benchmark (Dataset III) covering five diagnostic superclasses. On Dataset I, SimCardioNet achieves 0.975 accuracy, 0.973 precision, 0.973 recall, and 0.972 F1-score under 3-fold cross-validation. On Dataset II, the model demonstrates perfect classification performance (1.00 accuracy, precision, recall, and F1-score), highlighting strong generalization. On the PTB-XL dataset (Dataset III), SimCardioNet attains 0.921 accuracy and 0.921 F1-score, outperforming current state-of-the-art models including dual-branch CNNs, entropy-enhanced CNNs, and Bi-GRU architectures. Ablation studies confirm the critical contributions of self-supervised pretraining, attention mechanisms, and domain-specific augmentations. Grad-CAM visualizations further validate the model's focus on clinically relevant Electrocardiography regions. Our results underscore SimCardioNet's potential to reduce reliance on labeled data while delivering robust, interpretable, and clinically viable Electrocardiography classification especially valuable in resource-constrained settings.

Ask AI
Helpful
Bookmark
Share
View Full Paper

Cite This Study

Majid et al. (2026) studied Cardiovascular diseases (Electrocardiogram classification) (n=21,837). SimCardioNet vs. State-of-the-art models (dual-branch CNNs, entropy-enhanced CNNs, Bi-GRU architectures) was evaluated on Classification accuracy on the PTB-XL dataset (Dataset III). SimCardioNet achieved 0.921 accuracy and 0.921 F1-score on the PTB-XL dataset, outperforming current state-of-the-art models for automated multiclass electrocardiogram classification.

synapsesocial.com/papers/698828990fc35cd7a8848353https://doi.org/10.1038/s41598-026-36932-1
Ask AI
Helpful
Bookmark
Share
View Full Paper