DiffECG, a versatile probabilistic diffusion model, effectively synthesized, imputed, and forecasted ECG signals, outperforming standard GANs across multiple metrics and improving downstream arrhythmia classification accuracy to 99%.
Does DiffECG improve ECG signal synthesis and classifier performance compared to other generative models?
A novel diffusion probabilistic model, DiffECG, effectively synthesizes ECG signals for data augmentation, outperforming existing generative models.
Within cardiovascular disease detection using deep learning applied to ECG signals, the complexities of handling physiological signals have sparked growing interest in leveraging deep generative models for effective data augmentation. In this paper, we introduce a novel versatile approach based on denoising diffusion probabilistic models for ECG synthesis, addressing three scenarios: (i) heartbeat generation, (ii) partial signal imputation, and (iii) full heartbeat forecasting. Our approach presents the first generalized conditional approach for ECG synthesis, and our experimental results demonstrate its effectiveness for various ECG-related tasks. Moreover, we show that our approach outperforms other state-of-the-art ECG generative models and can enhance the performance of state-of-the-art classifiers.
Neifar et al. (Fri,) conducted a other in Arrhythmia (n=48). DiffECG (Diffusion Denoising Probabilistic Model) vs. GANs, LSTM, VAE, and other state-of-the-art generative models was evaluated on ECG signal generation quality (RMSE, MAE, FID, DTW, EMD, MMD) and downstream classification accuracy. DiffECG, a versatile probabilistic diffusion model, effectively synthesized, imputed, and forecasted ECG signals, outperforming standard GANs across multiple metrics and improving downstream arrhythmia classification accuracy to 99%.