Fusion-Based Deep Learning Ensemble on MIT-BIH and PTB-XL ECG Databases for Enhanced Cardiac Diagnosis

Abstract Electrocardiogram (ECG) analysis plays a critical role in the early detection and diagnosis of cardiac abnormalities. In this study, we propose a fusion-based deep learning ensemble framework that integrates two well-established public ECG databases, MIT-BIH Arrhythmia Database and PTB-XL, to develop a robust and automated cardiac diagnostic system. Our framework employs two base deep learning models — a CNN+LSTM hybrid and a DenseNet1D-inspired network — and combines their predictive features through a meta-learner based on Gradient Boosting. This multi-model integration, designed as a “mini doctor for the heart,” leverages the complementary strengths of both datasets and models. Experimental results demonstrate that the ensemble achieves near-perfect performance with Accuracy up to 100% and ROC-AUC of 1.000, surpassing the performance of individual models. These findings highlight the potential of database fusion and model ensembling for building reliable and scalable solutions in computer-aided cardiac diagnosis. Competing Interest Statement The authors have declared no competing interest. Funding Statement This study did not receive any funding. Author Declarations I confirm all relevant ethical guidelines have been followed, and any necessary IRB and/or ethics committee approvals have been obtained. Yes The details of the IRB/oversight body that provided approval or exemption for the research described are given below: The study used ONLY openly available human data that were originally located at: MIT-BIH Arrhythmia Database, PhysioNet: https://physionet.org/content/mitdb/1.0.0/ PTB-XL Electrocardiography Database, PhysioNet: https://physionet.org/content/ptb-xl/1.0.3/ Both datasets are de-identified, publicly accessible, and widely used for research purposes. I confirm that all necessary patient/participant consent has been obtained and the appropriate institutional forms have been archived, and that any patient/participant/sample identifiers included were not known to anyone (e.g., hospital staff, patients or participants themselves) outside the research group so cannot be used to identify individuals. Yes I understand that all clinical trials and any other prospective interventional studies must be registered with an ICMJE-approved registry, such as ClinicalTrials.gov. I confirm that any such study reported in the manuscript has been registered and the trial registration ID is provided (note: if posting a prospective study registered retrospectively, please provide a statement in the trial ID field explaining why the study was not registered in advance). Yes I have followed all appropriate research reporting guidelines, such as any relevant EQUATOR Network research reporting checklist(s) and other pertinent material, if applicable. Yes Data Availability All data used in the present study are openly available from public repositories. Specifically: MIT-BIH Arrhythmia Database, PhysioNet: https://physionet.org/content/mitdb/1.0.0/ PTB-XL Electrocardiography Database, PhysioNet: https://physionet.org/content/ptb-xl/1.0.3/ No new data were generated in this study.