A non-invasive continuous blood pressure measurement method based on deep neural network

Abstract

The present invention proposes a non-invasive continuous blood pressure measurement method based on deep learning, which mainly includes a pulse wave signal preprocessing method, a deep neural network model establishment method, and a blood pressure measurement method. The pulse wave signal input to the neural network is preprocessed, filtered, denoised, and normalized, so that the established model can be more accurate and adaptable. This method uses a deep neural network to avoid subjective feature extraction and complex mathematical modeling problems. At the same time, a large amount of data is used to train the network, so that the network can measure the blood pressure of different forms of pulse waves of different individuals. It can be applied to a wide range of blood pressure prediction scenarios. Compared with the existing technology, it has high objectivity and strong model robustness, and is suitable for integration in home medical monitoring equipment, and also suitable for blood pressure monitoring of wearable equipment.

Description

technical field [0001] The invention relates to the field of digital signal processing, in particular to a method for non-invasive continuous blood pressure calculation based on digital signal processing. Background technique [0002] Blood pressure (BP) is an important parameter that reflects the performance of the human circulatory system. Checking blood pressure is a major way to clinically judge diseases and observe treatment effects. Humans suffer from a number of diseases related to arterial blood pressure (ABP), including high blood pressure and heart disease. Therefore, blood pressure monitoring is of great significance. [0003] At present, blood pressure measurement methods can be roughly divided into two categories: direct measurement method and indirect measurement method: direct measurement method requires the catheter connected to the pressure sensor to be directly inserted into the aorta or heart. This method has the highest measurement accuracy, but because ...

AI Technical Summary

Problems solved by technology

However, this method is very dependent on the selection of features manually, and the quality of feature selection is directly related to the prediction accuracy.

In addition, for different forms of pulse waves collected by different measuring devices, the selected features may vary greatly

This shows that this method is too subjective and not robust enough

Images