A Bayesian Residual Transform for Signal Processing

TL;DR

This paper introduces a Bayesian Residual Transform (BRT), a novel multi-scale signal decomposition method for physiological signals, demonstrating its effectiveness in noise suppression through ECG analysis.

Contribution

The study presents a new Bayesian-based multi-scale decomposition method, BRT, with a deep cascading framework, offering an alternative to traditional scale-space and wavelet approaches.

Findings

BRT effectively suppresses noise in ECG signals.

BRT achieves higher SNR compared to traditional methods.

Feasibility demonstrated for physiological signal processing.

Abstract

Multi-scale decomposition has been an invaluable tool for the processing of physiological signals. Much focus in multi-scale decomposition for processing such signals have been based on scale-space theory and wavelet transforms. In this study, we take a different perspective on multi-scale decomposition by investigating the feasibility of utilizing a Bayesian-based method for multi-scale signal decomposition called Bayesian Residual Transform (BRT) for the purpose of physiological signal processing. In BRT, a signal is modeled as the summation of residual signals, each characterizing information from the signal at different scales. A deep cascading framework is introduced as a realization of the BRT. Signal-to-noise ratio (SNR) analysis using electrocardiography (ECG) signals was used to illustrate the feasibility of using the BRT for suppressing noise in physiological signals. Results in this study show that it is feasible to utilize the BRT for processing physiological signals for tasks such as noise suppression.