Key result
A novel framework combining a residual network with a squeeze and excitation block and bidirectional LSTM achieved superior arrhythmia classification performance compared with conventional methods.
Why the study?
Classification performance for arrhythmia types other than atrial fibrillation has not been satisfactory for clinical use due to minority classes.
Does a novel deep learning framework combining a residual network and bidirectional LSTM improve automatic cardiac arrhythmia classification compared to conventional methods?
Does a novel deep learning framework combining a residual network and bidirectional LSTM improve automatic cardiac arrhythmia classification compared to conventional methods?
The proposed deep learning framework improves the accuracy of automatic arrhythmia classification, particularly for minority classes, potentially aiding direct clinical diagnosis.
May aid automated ECG arrhythmia detection; leaves open prospective clinical validation before practice change.
Diagnosis and classification of arrhythmia, which is associated with abnormal electrical activities in the heart, are critical for clinical treatments. Previous studies focused on the diagnosis of atrial fibrillation, which is the most common arrhythmia in adults. The classification performance achieved by studies on other arrhythmia types is not satisfactory for clinical use owing to the small number of classes (minority classes). In this study, we propose a novel framework for automatic classification that combines a residual network with a squeeze and excitation block, and a bidirectional long short-term memory. 8-class, 4-class, and 2-class performances were evaluated on the MIT-BIH arrhythmia database (MITDB), MIT-BIH atrial fibrillation database (AFDB), and PhysioNet/Computing in the cardiology challenge 2017 database (CinC DB), respectively, and they were superior to performance achieved by conventional methods. In addition, the class-wise F1-score in the minority classes was higher than those of the methods adopted in existing studies. To measure the generalization ability of the proposed framework, AFDB and CinC DB were tested using a MITDB-trained model, and superior performance was achieved compared with ShallowConvNet and DeepConvNet. We performed a cross-subject experiment using AFDB and obtained a statistically higher performance using the proposed method compared with typical machine learning methods. The proposed framework can enable the direct diagnosis of arrhythmia types in clinical trials based on the accurate detection of the minority class.
No takes yet. Share an insight, caveat, or question.
Kim et al. (2022) studied Arrhythmia. Residual network combined with a squeeze and excitation block, and a bidirectional long short-term memory vs. Conventional methods (ShallowConvNet, DeepConvNet, typical machine learning methods) was evaluated on Classification performance. A novel framework combining a residual network with a squeeze and excitation block and bidirectional LSTM achieved superior arrhythmia classification performance compared with conventional methods.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: