Key result
Joint DTCWT-ASASS heart sound denoising algorithm outperforms existing techniques, achieving ~9 dB SNR.
Why the study?
Heart sound signals are corrupted by unwanted respiratory sounds and ambient noise, which significantly reduce signal clarity and interpretation.
Does the joint DTCWT and ASASS algorithm improve the Signal-to-Noise Ratio and reduce Root Mean Square Error in heart sound signals compared to existing models?
Does the joint DTCWT and ASASS algorithm improve the Signal-to-Noise Ratio and reduce Root Mean Square Error in heart sound signals compared to existing models?
The proposed joint DTCWT and ASASS algorithm significantly improves the denoising of heart sound signals, enhancing clarity for potential clinical interpretation.
May improve heart sound clarity for diagnosis in noisy settings; extends wavelet-sparsity methods and remains hypothesis-generating for adoption.
Various unwanted and unavoidable noises corrupt Heart Sound Signals (HSSs). It is strongly required to suppress respiratory sound and ambient noise due to significant reduction of clarity and interpretation of HSSs. In this paper, we propose a joint heart sound denoising using Dual-Tree Complex Wavelet Transform (DTCWT) and Adaptive Sparsity-assisted Signal Smoothing (ASASS) algorithm. In this research, the signal is first decomposed by DTCWT to obtain the multi-scale feature representation of the signal. Subsequently, ASASS suppresses pseudo-Gibbs artifacts around signal boundaries of DTCWT while implementing adaptive thresholding strategies to maximize the Signal-to-Noise Ratio (SNR). Experimental validation on the PhysioNet/CinC 2016 database and Open Access Heart Sound Dataset (OAHS Dataset) demonstrates that the proposed method significantly outperforms existing techniques. Under conditions involving Gaussian white noise (GWN) SNR of 0 dB, the proposed method achieves an SNR of 9.01 dB and a Root Mean Square Error (RMSE) of 0.032, outperforming standalone DTCWT and multiple existing models.
No takes yet. Share an insight, caveat, or question.
Hu et al. (2026) studied this question. A joint heart sound denoising algorithm combining DTCWT and ASASS achieved an SNR of 9.01 dB and an RMSE of 0.032 under 0 dB noise conditions, significantly outperforming existing techniques.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: