A Novel Data Augmentation Strategy for Robust Deep Learning Classification of Biomedical Time-Series Data: Application to ECG and EEG Analysis
Researchers have developed a new deep learning framework for classifying biomedical time-series data like ECG and EEG signals. The approach integrates a ResNet-based CNN with an attention mechanism and a novel data augmentation technique involving time-domain concatenation of augmented signal variants. This method achieved state-of-the-art accuracies of up to 100% on benchmark datasets while managing class imbalance and requiring minimal computational resources, making it suitable for deployment on low-end devices. AI
IMPACT Enhances accuracy and efficiency in biomedical signal analysis, potentially improving patient diagnostics and enabling deployment on resource-constrained devices.