Pulse wave multi-feature fusion-based respiratory rate extraction method

Abstract

The invention discloses a pulse wave multi-feature fusion-based respiratory rate extraction method, which comprises the following steps of: acquiring signals at the wrist and fingers of a user by using an array pulse wave acquirer to obtain a multi-channel pulse wave file, and performing data preprocessing on the pulse wave file to obtain a multi-channel fused pulse wave; multiple time-frequency characteristics of the fused pulse waves are extracted; constructing a credible measure of each time-frequency feature through time domain analysis; taking the credible measure of each feature as the learning weight of each feature in a neural network by using an attention mechanism, and extracting a respiration feature map by using a convolutional neural network; and fusing the extracted breathing feature maps, and inputting the fused breathing feature maps into a VGG regression model to obtain a final breathing rate. According to the method, on the basis of credible measurement of multiple features, the respiration feature map is better extracted in combination with an attention mechanism of the neural network, and the accuracy of pulse wave respiration rate extraction under the condition of big data is improved.

Description

technical field [0001] The invention relates to the technical field of respiration rate monitoring, in particular to a respiration rate extraction method based on pulse wave multi-feature fusion. Background technique [0002] Respiration rate is an important parameter that characterizes respiratory function and an important physiological indicator that reflects the health status of an individual. Chronic respiratory disease is one of the four major chronic diseases with the heaviest burden in the world. Whether it is the disease of the respiratory system itself or the disease of other important organs, the respiratory center will be affected to a certain extent if it develops to a certain extent. Therefore, continuous and accurate respiratory rate monitoring is beneficial to the prevention of lesions in the lungs, respiratory tract and other parts. [0003] Currently, methods for detecting respiration rate are mainly divided into electrical signal detection methods, acousti...

AI Technical Summary

Problems solved by technology

At present, the respiratory rate extraction based on photoplethysmography is mainly based on the modulation of respiration on the pulse wave baseline. The extraction methods include low-pass filtering, EMD decomposition, wavelet decomposition, or pulse wave-based baseline signal processing. The complexity of the system and the pulse wave acquisition process are extremely susceptible to interference. The baseline component of the pulse wave is not single, but also includes low-frequency signals related to temperature regulation and nervous system regulation, resulting in certain uncertainties in the respiratory signal extracted by the above method. The accuracy of respiration rate extraction is not high when the amount of data is large

Images