Cardiopulmonary sound separation method and system based on multi-signal-to-noise-ratio model

Abstract

The invention discloses a cardiopulmonary sound separation method and system based on a multi-signal-to-noise-ratio model. Through establishing a primary function learning network, a primary functioncan be adaptively adjusted according to training data, so that the representation of the transform domain is improved. According to the established single signal-to-noise ratio separation model, maximization of the sum of the signal-to-noise ratios of heart sounds and lung sounds is used as a target function, a primary function of time-frequency transformation can be learned adaptively, and end-to-end learning from a time domain mixed signal to a time domain heart sound signal and a lung sound signal is realized under the condition that the energy ratio of heart and lung sounds is known, so that the purpose of enhancing the heart and lung sound separation effect is achieved. A multi-signal-to-noise-ratio integrated network is established based on a single-signal-to-noise-ratio separation model. An LSTM network is used to learn the mapping weight of the cardiopulmonary sound mixed signal with the unknown energy ratio to each single signal-to-noise ratio separation model. The mapping weight can be adaptively adjusted according to the mixed signals with different energy ratios, and end-to-end learning from the time domain mixed signals to the time domain heart sound and lung sound signals is realized under the condition that the cardiopulmonary sound energy ratio is unknown.

Description

technical field [0001] The invention belongs to the field of cardiopulmonary sound signal separation, and in particular relates to a cardiopulmonary sound separation method and system based on a multi-signal-to-noise ratio model. Background technique [0002] Auscultation medical devices such as stethoscopes are commonly used to auscultate the heart and lung sounds of patients in the preliminary screening of clinical diagnosis of cardiopulmonary diseases. However, the crosstalk of heart and lung sounds in the time and frequency domains will reduce the effectiveness of clinical auscultation. For this reason, researchers Various methods have been proposed on how to isolate cardiopulmonary sounds. Among them, the cardiopulmonary sound separation methods based on Non-Negative Matrix Factorization (NMF) model and Long Short Time Memory (LSTM) network have achieved good results. Both of them first obtain the time spectrum of the cardiopulmonary sound mixed signal through Short-Ti...

AI Technical Summary

Problems solved by technology

[0004] Based on this, the present invention aims to propose a heart-lung sound separation method and system based on a multi-signal-to-noise ratio model, which uses a basis function learning network and a time-domain reconstruction network, and can adaptively adjust the basis function according to the training data to improve the transform domain The representation of the basis function learning network and the LSTM separation network are jointly optimized to realize the end-to-end network learning from the time-domain mixed signal to the time-domain separated signal, obtain a single SNR separation model, and build a multi-SNR based on this model Ratio integration model, to achieve the purpose of adaptively selecting the separation model in the case of unknown energy ratio of heart-lung sound, and solve the technical problem that the existing method cannot separate heart-lung sound with unknown energy ratio

Images