BCG signal processing method and device

Abstract

The embodiment of the invention provides a method and device for processing a BCG signal. The method comprises the following steps of obtaining a prescreening signal of target data from the BCG signal, and performing segment treatment on the prescreening signal to obtain time frequency image features of each segment prescreening signal; obtaining all the energy gathering regions in the time frequency image features for each segment prescreening signal, inputting the energy gathering regions to a neural network which is trained in advance, and outputting the quality classifying result of the time frequency image features, and according to the quality classifying result, determineing the energy gathering regions of the signal energy ingredients from all the energy gathering regions to be used as a target region; and according to the target region of all the segment prescreening signals, designing a time frequency fractional wave filter, inputting the prescreening signal to the time frequency fractional wave filter, and outputting a rescreening signal of the target data. The method disclosed by the embodiment of the invention has the advantages of being high in accuracy and stable.

Description

technical field [0001] The present invention relates to the technical field of signal processing, and more specifically, to a method and device for processing BCG signals. Background technique [0002] Sleep occupies about one-third of people's day. Sleep health is very important to human health. More and more people need to monitor and analyze their sleep conditions. Conventional medical sleep monitoring technology relies on PSG (sleep polysomnography, Polysomnography) monitoring technology, but the operation of medical equipment is complicated, and intrusive measurement will have a serious impact on the subject, so the non-contact BCG signal collects the key parameters of the human body. Applications such as speed, respiration rate, etc. are becoming more and more common. [0003] BCG (ballistocardiogram) does not need to touch the subject's body when collecting signals, thereby avoiding the discomfort caused by contact. The BCG signal collected by the micro-motion mattr...

AI Technical Summary

Problems solved by technology

[0005] However, the conditions for the above two types of algorithms to be tested are relatively ideal signals, approximate processing with low precision requirements in a short period of time, poor environmental anti-interference and universal applicability, and the influence of long-term trend changes in heart rate is not considered.

In a stable state, the range of heart rate changes is small. For example, the fluctuation of the heart rate within half an hour is within 20bpms and the heartbeat frequency changes within about 0.3hz. The precision requirement is very high. If the filter parameter is too wide, it will cause serious noise interference, and if it is too narrow, it will lead to the loss of key information.

However, the BCG signal environment is complex, the energy of the heartbeat information is weak and the signal is unstable, and there will be serious time-frequency aliasing noise in the actual time frequency. It is difficult to guarantee a high signal-to-noise ratio for the extracted signal.

When the environment is unstable and the acquisition equipment cannot be high-precision and standardized, different individuals, different time periods, emotional states and other external influences will lead to large differences in heart rate

Therefore, it is difficult to use simple statistical rules to design parameters, and it is difficult to meet the requirements of accuracy, anti-interference and universality with conventional thresholds and empirical models.

Images