Pulse wave reconstruction method for extracting time domain feature points

Abstract

The invention discloses a pulse wave reconstruction method for extracting time domain feature points. The method comprises the following steps of: a), recording a human body surface pulse wave signal by using a pulse wave signal measuring device, and converting the human body surface pulse wave signal into a digital signal; b), converting the digital signal into a power spectrum chart in a fast Fourier transform mode; c), calculating the volume of base frequency and obtaining a frequency value F0 corresponding to the highest peak value; d), dividing a frequency range into areas within the range of 0.5-1.5F0, 1.5-2.5F0...6.5-7.5F07 by using the frequency value as the interval size; and respectively solving a maximum value ai and corresponding frequency values fi in each area, wherein i can be a natural number from 1 to 7; and e), respectively calculating a phase value theta i corresponding to the frequency fi in a phase frequency chart; and f), substituting ai, fi and theta i into a formula to obtain a reconstruction signal of which the length is equal to that of an original signal, wherein t is equal to 0: 1 / fs: 4,000 / fs. The reconstructed waveform is highly similar to the original waveform on the time domain.

Description

technical field [0001] The invention relates to a processing method for human pulse wave and other periodic biomedical signals. Background technique [0002] Pulse diagnosis is an ancient medical diagnosis method in my country, but it has the disadvantages of being highly subjective and relying on the experience of doctors. Modern medicine requires the expression and analysis of pulse conditions to be objective, digital, and scientific. Pulse wave measurement and analysis technology has also become an important scientific research field. The pulse wave and other periodic biomedical signals measured from the human body surface, due to various reasons, often have certain regularity and inevitably contain a lot of noise, which makes the further analysis and calculation of the pulse wave difficult. difficulty. [0003] The traditional method is to use various methods to smooth the pulse wave in the time domain, the common ones are the median method, the linear moving average m...

AI Technical Summary

Problems solved by technology

However, the effect of these smoothing methods on removing noise is limited

For example, the median method and the straight-line moving average method are better for smoothing the rising edge of the pulse wave, but for the falling edge and trough position, there is still a lot of noise after smoothing

For the five-point triple smoothing method, if the smoothing times are small, the smoothing effect will not be good, and if the smoothing times are too many, the peak value will be reduced, the body shape will be wider, and the basic shape of the waveform will change greatly.

However, common filtering techniques, such as band-pass and low-pass filters, still cannot remove low-power noise near the passband.

Images