WO2024174124A1 - Non-contact blood pressure monitoring method and apparatus based on multi-spectral pulse waves - Google Patents

Abstract

Disclosed in the present invention are a non-contact blood pressure monitoring method and apparatus based on multi-spectral pulse waves. The method comprises: S1, acquiring a continuous video image containing human skin, selecting an effective skin region in the video image, and extracting pulse wave signals from pixels of the skin region; S2, performing denoising and component decomposition on pulse wave signals of a green-light waveband and a near-infrared light waveband by using a pulse wave signal of a blue-light waveband among multi-spectral pulse wave signals, and extracting a pulse wave transit time; and S3, establishing a regression model regarding the pulse wave transit time and a blood pressure, and calibrating and continuously and dynamically estimating the blood pressure. In the present invention, a single camera is used to acquire a continuous video image of human skin, such that the extraction of pulse wave signals in visible light and near infrared from the image can be simultaneously performed, thereby significantly enhancing the independence of different wavebands in pulse wave signal monitoring, and after a regression model is established using a calculated pulse wave transit time, blood pressure parameters can be calibrated, thereby ensuring the accuracy of blood pressure values.

Description

A non-contact blood pressure monitoring method and device based on multi-spectral pulse wave Technical Field

The present invention relates to the technical field of non-contact heart rate measurement, and in particular to a non-contact blood pressure monitoring method and device based on multi-spectral pulse wave.

Background Art

As my country gradually enters a stage of low fertility and aging society, the health of newborns and the elderly will be one of the social problems that need to be solved urgently. In order to achieve health monitoring and disease prevention for these two groups, it is urgent to continuously monitor physiological information such as pulse, blood pressure and electrocardiogram. Currently, the commonly used and relatively mature wearable physiological continuous monitoring products are mostly of two categories: (1) sports bracelets, which monitor heart rate signals through a single green band (the green light band PPG signal is strong and can achieve a certain degree of motion robustness, but the monitored physiological indicators are too single); (2) finger oximeters, which monitor heart rate and blood oxygen saturation through red and infrared dual bands (the infrared band PPG signal is weak, so the heart rate monitoring stability is low). Wearable devices can easily cause damage and infection to the delicate and fragile skin of newborns, so non-contact monitoring methods have been further developed.

Camera-PPG-based video sensing technology has been used for non-contact vital sign monitoring and health monitoring, using image/signal processing algorithms to extract human physiological signals from a video image containing continuous frames. Among them, the methods of video monitoring of heart rate and respiratory rate are relatively mature, while the research and application of blood pressure monitoring are relatively few, mainly because of the high complexity of its technical implementation. At present, the research on non-contact camera-PPG signal monitoring of blood pressure can be divided into two categories: (i) based on the characteristics of pulse transit time (PTT). In contact monitoring, the pulse transit time can be calculated by the time difference between the R peak of the ECG electrocardiogram and the systolic peak of the contact PPG of the fingertips. In non-contact blood pressure monitoring, the time difference and phase difference of the PPG signals of two different body parts can be used to be equivalent to the pulse transit time. Therefore, it is necessary to monitor multiple body parts simultaneously through video, which is easily limited by the scene in practical applications; (ii) Blood pressure estimation based on the waveform characteristics of PPG. This method has extremely high requirements for the quality of pulse wave (PPG) and extremely high requirements for personalized parameters (such as peripheral vascular resistance) in blood pressure calibration, and it only stays in the feasibility analysis stage. Therefore, the existing monitoring methods cannot achieve stable and non-contact blood pressure monitoring from a single body part (such as facial skin), and there are certain defects.

Technical Solutions

The purpose of the present invention is to provide a non-contact blood pressure monitoring method and device based on multi-spectral pulse wave to solve the problems mentioned in the above background technology.

In order to achieve the above object, the solution of the present invention is: a non-contact blood pressure monitoring method based on multi-spectral pulse wave, comprising:

S1. Acquire continuous video images of the monitored person including human skin through a multi-spectral optical imaging device, select a valid skin area in the video image, and extract pulse wave signals of multiple bands in the time domain from the pixels of the skin area;

S2, using the pulse wave signal in the blue light band of the multi-spectral pulse wave signal to reduce noise and decompose the pulse wave signals in the green light band and the near-infrared light band, and then extracting the pulse wave transmission time from the decomposed green light and near-infrared light pulse wave signals;

S3. Use the polynomial regression method to establish a regression model between pulse wave transmission time and blood pressure, calibrate and continuously dynamically estimate blood pressure, and complete the monitoring of blood pressure values.

Furthermore, the multi-spectral optical imaging device includes a light source transmitter and a camera; a polarizer 1 is arranged in front of the light source transmitter; a camera sensor is arranged on the camera, and a polarizer 2 perpendicular to the polarization direction of the light source is arranged in front of the camera sensor.

Furthermore, the light source emitter is a Phosphor LED light source with a continuous broadband spectrum.

Furthermore, the camera sensor is an RGB camera sensor and a NIR sensor.

Furthermore, the multiple band signals extracted from the pixels of the skin area are B-G-IR three-channel time domain signals, and the three bands in the B-G-IR are B (450 nm±10 nm) - G (550 nm±10 nm) - IR (805 nm±10 nm); after simple detrending and bandpass filtering, the B-G-IR three-channel time domain signal removes non-pulse wave components in the signal to generate a three-channel pulse wave signal.

Furthermore, the pulse wave signal in the green light band represents the blood pulsation volume in the arterioles.

Furthermore, the pulse wave signal in the near-infrared light band represents the blood pulsation volume in the artery.

Furthermore, when extracting the pulse wave conduction time from the decomposed green light and near-infrared light pulse wave signals, the systolic peaks of the pulse wave signals in the green light band and the near-infrared light band are detected to calculate the systolic peak spacing of the pulse wave signals in different bands in the same cardiac cycle, and the calculated peak spacing is the pulse wave conduction time.

A non-contact blood pressure monitoring device based on multi-spectral pulse wave, comprising:

Multispectral optical imaging equipment for collecting and acquiring continuous video images including human skin;

A multi-spectral pulse wave signal extraction module is used to extract pulse wave signals of multiple bands in the skin pixel in the time domain;

The pulse wave signal noise reduction and separation module uses the pulse wave signal in the blue light band of the multi-spectral pulse wave signal to reduce noise and decompose the pulse wave signals in the green light band and the near-infrared light band;

A pulse wave transmission time estimation module is used to extract the pulse wave transmission time from the decomposed green light and near-infrared light pulse wave signals;

The blood pressure calibration module is used to calibrate the extracted pulse wave transmission time, obtain the calibrated regression model parameters, and generate continuous dynamic estimation values of blood pressure.

Furthermore, the device also includes a UI display interface for displaying the monitoring video and the blood pressure values obtained through monitoring.

Beneficial Effects

The beneficial effects of the present invention compared with the prior art are:

The present invention uses a single camera to obtain continuous video images of human skin, so that the obtained images can be used to extract pulse wave signal data in visible light and near-infrared at the same time, which significantly enhances the independence of different bands in pulse wave signal monitoring. Then, the systolic peak spacing of pulse wave signals of different bands in the same cardiac cycle is calculated through the extracted pulse wave signal data, and the pulse wave conduction time is obtained, so that a regression model can be established through the pulse wave conduction time and blood pressure, and the continuous dynamic estimation of blood pressure parameters can be calibrated to obtain more accurate blood pressure data; and the use of multi-spectral pulse wave signals can suppress motion interference and environmental noise that occur during monitoring, further improving the accuracy of monitoring.

BRIEF DESCRIPTION OF THE DRAWINGS

Fig. 1 is a flow chart of the method of the present invention;

FIG. 2 is a structural block diagram of the blood pressure monitoring device of the present invention.

Embodiments of the present invention

Embodiments of the present invention are described in detail below, examples of which are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are intended to be used to explain the present invention, and should not be construed as limiting the present invention.

Example

As shown in FIG1-2, a non-contact blood pressure monitoring method based on multi-spectral pulse wave includes:

S1. Obtain continuous video images of the monitored person including human skin through a multi-spectral optical imaging device, select an effective skin area in the video image, and extract pulse wave signals of multiple bands in the time domain from the pixels of the skin area; in this embodiment, the multi-spectral optical imaging device includes a light source transmitter and a camera; a polarizer 1 is provided in front of the light source transmitter; a camera sensor is provided on the camera, and a polarizer 2 perpendicular to the polarization direction of the light source is provided in front of the camera sensor; in this embodiment, the light source transmitter is a Phosphor LED light source with a continuous broadband spectrum; in this embodiment, the camera sensor is an RGB camera sensor and a NIR sensor; in this embodiment, the multiple band signals extracted from the pixels of the skin area are B-G-IR three-channel time domain signals, and the three bands in the B-G-IR are B (450 nm±10 nm) - G (550 nm±10 nm) - IR (805 nm±10 nm); after simple detrending and bandpass filtering, the non-pulse wave components in the B-G-IR three-channel time domain signal are removed to generate a three-channel pulse wave signal; wherein the skin area can be the facial skin area of the human body; the camera sensor is equipped with a three-band narrow-band filter for extracting the band signal in the skin area pixel; the light source emitter adopts a Phosphor LED light source with a continuous broadband spectrum, and the generated continuous spectrum includes visible light and near-infrared bands (450-950 nm), the emitted photons pass through polarizer 1 and become a single vibration direction to illuminate the human skin. The polarized photons emitted by the light source are absorbed and scattered by the skin tissue and then reflected and can pass through the polarizer 2 on one side of the camera. This process is a depolarization process, which will prevent the specular reflection light that has not entered the skin tissue from passing through, so that the polarized light can ensure that the video mainly monitors the diffuse reflection signal containing physiological information from the skin tissue layer; in this embodiment, the polarizer 1 of the light source transmitter is 0°, and the polarizer 2 of the camera sensor is 90°. Such a setting can make the polarization direction of the polarizer vertical, thereby achieving the purpose of suppressing the specular reflection of the skin and ensuring the clarity of the shooting; and the polarizer 2 at the camera sensor can change the vibration direction of the light, and the three-band narrowband filter of the camera sensor will force the camera's shooting channel from the R-G-B channel to the narrowband B-G-IR channel, specifically using B (450 nm±10 nm) - G (550 nm±10 nm) - IR (805 nm±10 The use of narrow bands can enhance the independence between bands and strengthen the perception of physiological information of different tissue layers, so that a single camera can simultaneously obtain images in visible light and near-infrared. Moreover, after simple detrending and bandpass filtering, the B-G-IR three-channel time domain signal can effectively reduce noise infection, ensuring accurate and stable transmission of signals.

S2. Use the pulse wave signal of the blue light band in the multi-spectral pulse wave signal to reduce noise and decompose the pulse wave signals of the green light band and the near-infrared light band, and then extract the pulse wave transmission time from the decomposed green light and near-infrared light pulse wave signals; in this embodiment, the pulse wave signal of the green light band represents the blood pulsation volume in the arterioles; in this embodiment, the pulse wave signal of the near-infrared light band represents the blood pulsation volume in the arteries; in this embodiment, the decomposed green light and near-infrared light When extracting the pulse wave transmission time from the optical pulse wave signal, the systolic peaks of the pulse wave signals in the green light band and the near-infrared light band are detected to calculate the systolic peak spacing of the pulse wave signals in different bands in the same cardiac cycle. The calculated peak spacing is the pulse wave transmission time. The pulse wave signal in the blue light band is used as the reference signal. When the pulse wave signals in the green light band and the near-infrared light band are decomposed into components, the photons of the pulse wave signal in the blue light band can only reach the subcutaneous <0.5 mm, which mainly senses the physiological information of the epidermis (i.e., the mechanical compression of capillaries by arterial pulsation), so it can remove the interference of non-pulsating blood in the superficial skin tissue in the pulse wave signals of the green light band and the near-infrared light band, making the pulse wave signals of the green light band and the near-infrared light band purer; and the extracted pulse wave transmission time is calculated by the peak detection algorithm, by detecting the systolic peaks of the pulse wave signals of the green light band and the near-infrared light band, the systolic peak spacing (i.e., PTT characteristics) of the pulse wave signals of different bands in the same cardiac cycle can be calculated;

S3. Use the polynomial regression method to establish a regression model between pulse wave transmission time and blood pressure, calibrate and continuously dynamically estimate the blood pressure, and complete the monitoring of blood pressure values; wherein, when using the polynomial regression method to establish a regression model between pulse wave transmission time and blood pressure, blood pressure includes systolic pressure, diastolic pressure, and average blood pressure. These data, systolic pressure, diastolic pressure, and average blood pressure, are obtained from the user's standard monitor, and the blood pressure calibration model extracts features from multi-spectral polarized light signals and maps them to blood pressure values through the model, so that the calibrated continuous dynamic estimation data also includes systolic pressure, diastolic pressure, and average blood pressure, and finally accurate systolic pressure, diastolic pressure, and average blood pressure values are obtained.

As a preferred embodiment of the present invention, a non-contact blood pressure monitoring device based on multi-spectral pulse wave is also included, comprising:

Multispectral optical imaging equipment for collecting and acquiring continuous video images including human skin;

A multi-spectral pulse wave signal extraction module is used to extract pulse wave signals of multiple bands in the skin pixel in the time domain;

The pulse wave signal noise reduction and separation module uses the pulse wave signal in the blue light band of the multi-spectral pulse wave signal to reduce noise and decompose the pulse wave signals in the green light band and the near-infrared light band;

A pulse wave transmission time estimation module is used to extract the pulse wave transmission time from the decomposed green light and near-infrared light pulse wave signals;

The blood pressure calibration module is used to calibrate the extracted pulse wave transmission time, obtain the calibrated regression model parameters, and generate continuous dynamic estimation values of blood pressure.

In this embodiment, the device further includes a UI display interface for displaying the monitoring video and the blood pressure value obtained by monitoring;

In summary, the present invention provides a non-contact blood pressure monitoring method and device based on multi-spectral pulse wave, which uses a single camera to obtain continuous video images of human skin, so that the obtained image can simultaneously extract pulse wave signal data in visible light and near-infrared, significantly enhancing the independence of different bands in pulse wave signal monitoring, and then calculating the systolic peak spacing of pulse wave signals of different bands in the same cardiac cycle through the extracted pulse wave signal data, and obtaining the pulse wave conduction time, so as to establish a regression model through the pulse wave conduction time and blood pressure, calibrate the continuous dynamic estimation of blood pressure parameters, and obtain more accurate blood pressure data; and the use of multi-spectral pulse wave signals can suppress motion interference and environmental noise occurring in monitoring, further improving the accuracy of monitoring.

In the description of this specification, the description with reference to the terms "one embodiment", "some embodiments", "example", "specific example", or "some examples" etc. means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representation of the above terms should not be understood as necessarily being directed to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described may be combined in any one or more embodiments or examples in a suitable manner. In addition, those skilled in the art may combine and combine different embodiments or examples described in this specification.

Although the embodiments of the present invention have been shown and described above, it is to be understood that the above embodiments are exemplary and are not to be construed as limitations of the present invention. A person skilled in the art may change, modify, replace and vary the above embodiments within the scope of the present invention.

Claims (10)

A non-contact blood pressure monitoring method based on multi-spectral pulse wave, characterized by comprising: S1. Acquire continuous video images of the monitored person including human skin through a multi-spectral optical imaging device, select a valid skin area in the video image, and extract pulse wave signals of multiple bands in the time domain from the pixels of the skin area; S2, using the pulse wave signal in the blue light band of the multi-spectral pulse wave signal to reduce noise and decompose the pulse wave signals in the green light band and the near-infrared light band, and then extracting the pulse wave transmission time from the decomposed green light and near-infrared light pulse wave signals; S3. Use the polynomial regression method to establish a regression model between pulse wave transmission time and blood pressure, calibrate and continuously dynamically estimate blood pressure, and complete the monitoring of blood pressure values. A non-contact blood pressure monitoring method based on multispectral pulse wave as described in claim 1, characterized in that: the multispectral optical imaging device includes a light source transmitter and a camera; a polarizer 1 is provided in front of the light source transmitter; a camera sensor is provided on the camera, and a polarizer 2 perpendicular to the polarization direction of the light source is provided in front of the camera sensor. A non-contact blood pressure monitoring method based on multi-spectral pulse wave as described in claim 2, characterized in that: the light source emitter is a Phosphor LED light source with a continuous broadband spectrum. The non-contact blood pressure monitoring method based on multi-spectral pulse wave transit time as described in claim 2 is characterized in that: the camera sensor is an RGB camera sensor and a NIR sensor. A non-contact blood pressure monitoring method based on multispectral pulse wave transmission time as described in claim 1, characterized in that: the multiple band signals extracted from the skin area pixels are B-G-IR three-channel time domain signals, and the three bands in the B-G-IR are B (450 nm±10 nm) - G (550 nm±10 nm) - IR (805 nm±10 nm); after simple detrending and bandpass filtering, the B-G-IR three-channel time domain signal removes non-pulse wave components in the signal to generate a three-channel pulse wave signal. A non-contact blood pressure monitoring method based on multi-spectral pulse wave as described in claim 1, characterized in that: the pulse wave signal in the green light band represents the blood pulsation volume in the arterioles. A non-contact blood pressure monitoring method based on multi-spectral pulse wave as described in claim 6, characterized in that: the pulse wave signal in the near-infrared light band represents the blood pulsation volume in the artery. A non-contact blood pressure monitoring method based on multi-spectral pulse wave as described in claim 7, characterized in that: when extracting the pulse wave transmission time from the decomposed green light and near-infrared light pulse wave signals, the systolic peaks of the pulse wave signals in the green light band and the near-infrared light band are detected to calculate the systolic peak spacing of the pulse wave signals in different bands in the same cardiac cycle, and the calculated peak spacing is the pulse wave transmission time. A non-contact blood pressure monitoring device based on multi-spectral pulse wave, characterized by comprising: Multispectral optical imaging equipment for collecting and acquiring continuous video images including human skin; A multi-spectral pulse wave signal extraction module is used to extract pulse wave signals of multiple bands in the skin pixel in the time domain; The pulse wave signal noise reduction and separation module uses the pulse wave signal in the blue light band of the multi-spectral pulse wave signal to reduce noise and decompose the pulse wave signals in the green light band and the near-infrared light band; A pulse wave transmission time estimation module is used to extract the pulse wave transmission time from the decomposed green light and near-infrared light pulse wave signals; The blood pressure calibration module is used to calibrate the extracted pulse wave transmission time, obtain the calibrated regression model parameters, and generate continuous dynamic estimation values of blood pressure. The non-contact blood pressure monitoring device based on multi-spectral pulse wave as described in claim 9 is characterized by: it also includes a UI display interface for displaying the monitoring video and the blood pressure value obtained by monitoring.