Fetal heart sound signal noise reduction method based on non-negative matrix factorization

Abstract

The invention discloses a fetal heart sound signal noise reduction method based on non-negative matrix factorization. An existing filtering method has limitation in improving the signal-to-noise ratio of fetal heart sound signals. The method comprises the steps of performing short-time Fourier transform on a signal to obtain an amplitude information matrix and a phase angle information matrix; performing singular value decomposition and non-negative matrix decomposition on the amplitude information matrix, selecting useful features according to the feature matrix and the coefficient matrix, and combining to obtain a denoised amplitude information matrix; performing phase compensation on the phase angle information matrix to obtain a compensated phase matrix; and performing short-time inverse Fourier transform on the amplitude information matrix and the compensated phase matrix to obtain a denoised time domain signal. The singular value decomposition method adopted by the invention avoids the influence of the randomness of the initial value on the matrix iteration result, and the adopted phase compensation method uses the compensated phase to replace the original noisy signal phase, so that the credibility of the phase signal after noise reduction is improved, and the noise reduction effect of the fetal heart sound signal is obviously improved.

Description

technical field [0001] The invention relates to a signal noise reduction method, in particular to a fetal heart sound signal noise reduction method based on non-negative matrix decomposition. Background technique [0002] In the process of actually collecting the fetal heart sound signal, there is a lot of noise, not only the noise generated by the movement of the fetus itself in the womb, the beating sound of the mother's heart, and the sound of other organs will be mixed in the useful signal. The sound of friction with the mother's abdomen and environmental noise are also the main components of noise. Therefore, it is necessary to reduce the noise of the collected signal and improve the signal-to-noise ratio in order to ensure the accuracy of the subsequent calculation of the fetal heart rate. [0003] At present, many noise reduction methods have been applied to fetal heart sound signal processing, among which wavelet transform and empirical mode decomposition are common...

AI Technical Summary

Problems solved by technology

The filter based on wavelet transform, the wavelet mother function and the selection of wavelet threshold have a great influence on the noise reduction effect. Secondly, wavelet transform is mainly used to suppress the noise outside the frequency band of the useful signal, but it cannot suppress the noise in the same frequency band as the useful signal; Empirical mode decomposition is an empirical algorithm. At present, there is no precise mathematical reasoning, and there are key defects such as envelope fitting, endpoint effect, and mode mixing. These defects will lead to unsatisfactory results of empirical mode decomposition; The heart sound signal is a non-stationary and nonlinear signal, and the noise frequency band overlaps with the useful signal frequency band, so it is difficult to use wavelet transform and empirical mode decomposition to achieve optimal filtering

In summary, although the above-mentioned existing filtering methods have practical significance for improving the signal-to-noise ratio of the fetal heart sound signal, they have certain limitations, which affect the reliability of the noise reduction results.

Images