Key result
Multi-feature Bi-LSTM model achieves ~96% accuracy for ECG heartbeat classification.
Why the study?
Does a multi-feature fusion and compressed Bi-LSTM network improve the accuracy and memory efficiency of ECG heartbeat classification compared to conventional LSTM networks?
Does a multi-feature fusion and compressed Bi-LSTM network improve the accuracy and memory efficiency of ECG heartbeat classification compared to conventional LSTM networks?
A novel compressed Bi-LSTM model with multi-feature fusion enables highly accurate and memory-efficient ECG heartbeat classification, facilitating real-time cardiovascular monitoring on resource-constrained wearable devices.
May enable real-time ECG monitoring on wearables; extends Bi-LSTM models but leaves open clinical validation.
In this paper, we present a memory-efficient ECG based heartbeat classification for wearable devices enabled by multi-feature fusion and compressed bidirectional long short term memory (Bi-LSTM). A multi-feature fusion technique based on time intervals and under-the-curve areas is proposed to extract the main characteristic points of the ECG waveform with high accuracy and robustness against noise and artifacts. A Bi-LSTM network is developed to process the input sequence in both forward and backward directions, resulting in higher accuracy and a 28% smaller network size compared with a conventional LSTM. Multiple neural networks with varying sizes, including tiny (84k), small (150k), medium (478k), and large (1.25M) models, are developed to achieve high accuracy across all classes. The overall accuracy is 96.4% for the large model and 94.6% for the tiny model, while the F1 score across all classes exceeds 89.1% and 85.1%, respectively. The proposed models, compressed using post-training quantization techniques, achieve state-of-the art performance. The compressed large model with 8-bit integer quantization (INT8) achieves an accuracy of 88.4% with 1.3MB of memory, while the compressed tiny model with dynamic range quantization achieves 94.6% accuracy with only 139kB of memory.
No takes yet. Share an insight, caveat, or question.
Nikandish et al. (2026) studied Arrhythmia (n=48). Multi-feature fusion and compressed Bi-LSTM neural networks vs. Conventional LSTM and prior state-of-the-art models was evaluated on Heartbeat classification accuracy. The proposed multi-feature fusion and large Bi-LSTM model achieved an overall classification accuracy of 96.1%, with a compressed version achieving 88.4% accuracy using only 1.3 MB of memory for wearable devices.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: