JP2006271949A - Method for detecting respiration signal from pulse wave - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To establish a method for continuously separating/detecting a respiration signal in real-time for a long period with pulse waves as a signal source since a fluctuation component based on an inspiration/expiration movement is superimposed on the signals of the various pulse waves on an artery with pulsation of the heart. <P>SOLUTION: The pulse waves measured with various methods are adopted as an input signal source 1. The respiration component of the signal is separated/detected from the pulse waves through a BPF (band-pass filter) 2. The components are amplified by an amplifier 3 so as to be taken out as a normal respiration signal. <P>COPYRIGHT: (C)2007,JPO&INPIT

Description

Detailed Description of the Invention

Industrial application fields

The detection of the respiration signal has significance in the evaluation of vital signs in the medical field. Vital sign measurement items are generally four items, blood pressure, pulse, respiration, and body temperature, which are essential basic information for checking that a person is alive. Recently, sleep apnea syndrome (hereinafter referred to as SAS) is becoming a major social problem. SAS is a disease in which apneas of 10 seconds or more occur during sleep at night with the occurrence of 5 or more times per hour, and polysomnography (polysomnography) is used for the definitive diagnosis. The all night inspection is emphasized. Examination items of polysomnography include EEG, respiratory monitor, electrocardiogram, electromyogram, oxygen saturation, eye movement, etc. Among them, the presence or absence of breathing by the respiratory monitor and the level of consciousness by EEG Is extremely important, and in particular, detection of respiratory signals and long-term recording are positioned as the most important measurement items meaningful for SAS.

Conventional methods of monitoring breathing include the method of detecting the airflow in the mouth and nostrils with breathing using a thermistor and pressure sensor, and the method of monitoring the movement of the chest and abdomen due to breathing, using the impedance method and inductance plethysmography. -The electric resistance method by the method and the strain gauge is used for practical use. In addition, many attempts have been made to detect the presence or absence of breathing from snoring sounds (stuttering), breathing sounds, body pressure changes in supine position, changes in body position, etc., but these are still in the stage of practical application. Not reached. Among methods currently in practical use, a method of detecting airflow in the nostril with a pressure sensor is considered highly reliable for the diagnosis of apnea.

Problems to be solved by the invention

It is a fact well known in medicine that the various pulse waves defined in claim 2 are modulated by respiratory fluctuations accompanying respiratory movement. These are generally called respiratory arrhythmia, blood pressure second-class sway, etc., and all of them are explained as a result of changes in intrathoracic pressure accompanying respiratory exercise affecting the cardiac output ability and heart rhythm. Therefore, in the arterial system in which they propagate from the heart to the periphery, this is naturally reflected in the fluctuation of the pulse wave observed in the peripheral vascular bed. In this patent, focusing on this point, a method of separating the two components of respiratory vibration and pulse wave vibration from one pulse wave signal is considered as the subject of the invention. Is. When both are compared by the number of vibrations per minute, the respiration rate at rest is 15 to 20 times, the pulse wave fluctuation number is 60 to 80 times, and this is expressed in terms of frequency, respiration is 0.25 to 0.33 Hz. The pulse wave is 1.0 to 1.33 Hz. Therefore, the solution of the present invention is how to accurately separate both signals and extract only the respiratory signal, that is, to accurately separate and detect the respiratory signal superimposed on the pulse wave based on the frequency difference between the two signals. It is a challenge to try.

Means for solving the problem

In order to separate and detect two signal components inherent in one signal, the frequency spectrum of the original signal is obtained by the fast Fourier transform (FFT) method, etc., and the frequency component of the signal and its bandwidth are determined and separated. It is common to process. If the frequency range is known and the boundary frequency for separating the two signals can be determined, how to accurately separate the frequency components of the two signals in real time becomes the next issue. One of them, a signal separation method using a filter, is simple and reliable as real-time processing. The signal separation method using the BPF of the present invention is claimed based on this idea. . Although the respiration signal existing in the lower frequency region than the pulse wave signal can theoretically be detected only by the low-pass filter, in this patent, the reason for making it a BPF type having two cut-off frequencies is that It is clear from our basic experiments that a filter with an accurate bandwidth in terms of sensitivity and S / N is extremely advantageous in order to accurately separate a signal containing a very low frequency component such as a pulse wave and a pulse wave. Because it became. Further, the BPF cutoff frequency of claim 4 is determined through these basic experiments.

Action

Long-term respiratory monitoring is indispensable for polysomnography used for definitive diagnosis of SAS. The purpose is to accurately test and diagnose apnea conditions of 10 seconds or more during nighttime sleep. At the same time, a heart rate monitor during sleep is one of the important measurement items in this polysomnography. In the present invention, as long as the heart does not stop, the pulse wave signal can be measured without disappearing. Therefore, by processing the pulse wave signal measured by various methods using the respiratory signal detection method of the present invention, the presence or absence of breathing can be easily performed. Time recording is possible. In addition, by using the remaining pulse wave from which the respiratory signal is removed, the signal is waveform-shaped, so that the RR interval and the heart rate can be easily calculated and measured. Conventional heart rate monitoring using an electrocardiogram can be detected also from a pulse wave signal, and if the method of the present invention is expanded, two information of a respiratory monitor and a heart rate monitor can be acquired simultaneously from one pulse wave measurement. Simultaneous use of different devices is avoided. Furthermore, when oxygen saturation, which is a basic item of SAS, is measured with a fingertip-type photoelectric pulse wave oximeter, the method of the present invention is used to simultaneously monitor three items of respiration, heart rate, and oxygen saturation. Is not impossible. As described above, the present invention can greatly contribute to the examination of SAS polysomnography.

FIG. 1 shows a block configuration of the method of the present invention for separating and detecting a respiratory signal from a pulse wave. First, the pulse wave measured by various methods is guided from the input signal source 1 to the BPF 2 that performs signal separation processing, and the respiratory signal is extracted from the pulse wave. After being amplified by the AMP 3, only the respiratory signal is extracted. .

A specific embodiment is shown in FIG. The recording shown in the figure is a part of an experimental result of detecting a respiratory signal from a pulse wave using a method of the present invention in a human. Three sets of measurement records were taken when the posture of the same subject was changed, from the left side to the sitting, standing and supine positions. In each posture, the upper part of the recording shows a recording example of inspiration / expiration change of respiration by the change of the nostril airflow 4 by the thermistor method. The lower row is a recording example of photoelectric pulse wave (finger volume pulse wave) 6 with the second finger. The middle record is a respiration signal 5 obtained by processing the pulse wave 6 with the BPF defined in claims 3 and 4 of the present invention using the input signal source 1 of FIG. The respiratory signal 5 detected from the nostril airflow 4 and the respiratory signal 5 from the pulse wave 6 are different in phase, but the waveform periods, that is, rhythms are in good agreement, and the presence / absence information of respiration is accurately detected. The practical usefulness of the method was fully confirmed.

The invention's effect

The invention of the “method of detecting a respiratory signal from a pulse wave” of this patent contributes to greatly improving the reliability of apnea detection during nighttime sleep by simplifying long-term respiratory monitoring in polysomnography of SAS. it can. Further, not only respiratory monitoring from pulse wave measurement according to the present invention but also simultaneous measurement of heart rate monitoring is possible by processing the waveform. Furthermore, by introducing a pulse oximeter for the measurement of oxygen saturation, it is possible to perform long-term monitoring of three items of oxygen saturation, respiration, and heart rate.

It is a block diagram of the method of detecting a respiration signal from a pulse wave. This is an example in the case where the posture is changed by the same person in the recording of the respiratory curve of the nostril airflow, the respiratory signal from the pulse wave, and the pulse wave (the fingertip volume pulse) in the human.

Explanation of symbols

1 Input signal source 2 Band pass filter (BPF)
3 Amplifier 4 Respiratory curve by thermistor nostril 5 Respiratory signal detected from pulse wave 6 Fingertip volume pulse wave as input signal

Claims (5)

Pressure pulse wave, which is a pressure change in the blood vessel that occurs based on the heart beat, volume pulse wave, which is a pulsation phenomenon of the blood vessel volume that is transmitted to the peripheral blood vessel, and blood pulsation in the narrow blood vessels that accompanies it A method of detecting a respiratory fluctuation component (hereinafter referred to as a respiratory signal) from a pulsation phenomenon such as a pulse wave and a blood flow in which a respiratory fluctuation component is superimposed on any phenomenon such as a flow. As a signal source for detecting a respiratory signal, a pressure pulse wave by a pressure sensor or the like, a transmission type or reflection type photoelectric pulse wave by an optical method, a pulse signal by a pulse oximeter, and a laser that is tissue blood flow. 2. The respiratory signal detection method according to claim 1, wherein a pulse pulsation phenomenon such as a Doppler blood flow signal is defined, and the various pulsation phenomena (hereinafter referred to as pulse waves) are defined as an input signal. The detection of the respiration signal inherent in the pulse wave is performed by passing the various pulse wave signals corresponding to claim 2 through a narrow bandpass filter (hereinafter referred to as BPF). Method 1 (FIG. 1). 4. The BPF according to claim 3, wherein the pass band width of the BPF is a narrow band characteristic of 0.05 Hz to 0.50 Hz, and the attenuation gradient of the filter is 12 dB / oct or more for both the low pass and the high pass. The BPF filter structure is defined as an analog filter method when the input pulse wave signal is an analog signal, and a digital filter method and a Fourier transform filter when the pulse wave signal is a digital signal. Item 4. BPF.

Images