Key result
Mobile ECG random forest model achieves ~97% accuracy for MI detection.
Why the study?
Cardiovascular diseases, particularly myocardial infarction and arrhythmias, require timely and accurate computational support for screening and follow-up within a single mobile-health workflow.
Does a unified IoT-enabled mobile ECG framework using Random Forest classifiers accurately detect myocardial infarction and classify multiclass arrhythmias from 12-lead ECG row-level signal samples?
Does a unified IoT-enabled mobile ECG framework using Random Forest classifiers accurately detect myocardial infarction and classify multiclass arrhythmias from 12-lead ECG row-level signal samples?
This study demonstrates an integrated mobile-health prototype combining ECG filtering and Random Forest classifiers that achieves high within-dataset row-level accuracy for both myocardial infarction detection (96.89%) and multiclass arrhythmia classification (89.59%), though patient-disjoint clinical validation is still required.
Answers are grounded in this paper and cite their sources.
Supports mobile IoT ECG ML tools; leaves open prospective validation before clinical adoption.
Cardiovascular diseases (CVDs), particularly myocardial infarction (MI) and arrhythmias, require timely and accurate computational support for screening and follow-up. This study presents a unified internet of things (IoT)-enabled electrocardiogram (ECG) framework for mobile MI detection and multiclass arrhythmia classification. The framework integrates ECG preprocessing using biorthogonal and median filtering, machine learning (ML)-based interpretation, and mobile application delivery for prompt diagnostic feedback. The main contribution of this work is the integration of MI detection and multiclass arrhythmia classification within a single mobile-health workflow. The best random forest (RF) model achieved 96.89% accuracy for MI detection. For arrhythmia classification, the selected RF model achieved 89.59% accuracy on the held-out test set. These findings support the potential of the proposed framework as a mobile decision-support system for ECG interpretation and remote follow-up. Because evaluation was performed on row-level samples rather than patient-disjoint data, the reported results should be interpreted as within-dataset decision-support performance. Overall, the framework provides a practical link between ECG interpretation and mobile-health delivery while supporting further validation in real-world deployment settings.
No takes yet. Share an insight, caveat, or question.
Hasan et al. (2026) studied Myocardial infarction and cardiac arrhythmias. Unified IoT-enabled mobile ECG framework using Random Forest (Gini criterion) with biorthogonal and median filtering vs. HistGradientBoosting, GradientBoosting, LightGBM, and Entropy-based Random Forest classifiers was evaluated on Held-out test classification accuracy for myocardial infarction detection (95% CI 96.83-96.94). Random forest within a mobile ECG framework achieved 96.89% accuracy for MI detection and 89.59% accuracy for multiclass arrhythmia classification.
Synapse has enriched 10 related papers. Consider them for comparative context: