Why the study?
Interpreting myocardial infarction via ECG is challenging due to morphological variation, and existing machine learning algorithms rely on heuristic features with shallow architectures.
Does an LSTM-based deep learning architecture improve ECG-rhythm signal classification for myocardial infarction compared to standard RNN and GRU?
Population
15-lead ECG signals of MI and healthy controls from the PhysioNet PTB Diagnostic ECG Database
Comparison
RNN vs LSTM vs GRU recurrent network architectures
Design
Machine learning model development and validation study
Authors
Loading...
Supports LSTM integration for automated ECG-based MI detection; extends recurrent network benchmarks in an RCT.
Does an LSTM-based deep learning architecture improve ECG-rhythm signal classification for myocardial infarction compared to standard RNN and GRU?
Deep learning using an LSTM architecture provides highly accurate automated interpretation of myocardial infarction from ECG signals, outperforming other recurrent network structures.
Darmawahyuni et al. (2019) studied this question.
Synapse has enriched 4 closely related papers on similar clinical questions. Consider them for comparative context: