CN-122004812-A - MIMO UWB-based non-contact blood pressure detection method and system

CN122004812ACN 122004812 ACN122004812 ACN 122004812ACN-122004812-A

Abstract

The invention provides a non-contact blood pressure detection method and system based on UWB, which is characterized in that ultra-wideband impulse signals are transmitted through time division duplex of a double-transmitting antenna, echo signals reflected by a part to be detected of a human body are synchronously collected through four paths of receiving antennas, and four paths of synchronous baseband I/Q signals are output; the method comprises the steps of carrying out channel consistency calibration, DC removal, static clutter suppression and differential motion compensation on four paths of synchronous baseband I/Q signals to obtain preprocessed multichannel echo signals, carrying out spatial spectrum estimation and positioning on the preprocessed multichannel echo signals to-be-detected blood vessel areas, extracting continuous phase change sequences through multichannel joint phase unwrapping, converting the continuous phase change sequences into sub-millimeter micro-displacement pulse wave time sequence signals corresponding to human body blood vessel pulsation, extracting multichannel pulse wave space characteristic parameters, calculating to obtain systolic pressure and diastolic pressure values of a human body according to the multichannel pulse wave space characteristic parameters based on a pre-calibrated multichannel characteristic-blood pressure mapping model, and outputting and storing blood pressure detection results.

Inventors

JING ZHENHAI
YANG WEIJUN
WEN MIN

Assignees

长沙驰芯半导体科技有限公司

Dates

Publication Date: 20260512
Application Date: 20260320

Claims (10)

1. A MIMO UWB-based non-contact blood pressure detection method, the method comprising: transmitting ultra-wideband impulse signals through time division duplex of the double transmitting antennas, synchronously collecting echo signals reflected by a part to be detected of a human body through four paths of receiving antennas, and outputting four paths of synchronous baseband I/Q signals; Carrying out channel consistency calibration, DC removal, static clutter suppression and differential motion compensation on the four paths of synchronous baseband I/Q signals to obtain preprocessed multichannel echo signals; Carrying out space spectrum estimation on the preprocessed multichannel echo signals to locate a blood vessel region to be detected, extracting a continuous phase change sequence through multichannel combined phase unwrapping, converting the continuous phase change sequence into sub-millimeter micro-displacement pulse wave time sequence signals corresponding to human body blood vessel pulsation, and extracting multichannel pulse wave space characteristic parameters; Calculating to obtain the systolic pressure and diastolic pressure values of the human body according to the multichannel pulse wave space characteristic parameters based on a pre-calibrated multichannel characteristic-blood pressure mapping model; and outputting and storing a blood pressure detection result to realize communication interaction with the upper terminal.
2. The MIMO UWB-based non-contact blood pressure monitoring method of claim 1 wherein the performing channel consistency calibration, dc removal, static clutter suppression, differential motion compensation processing on the four-channel synchronous baseband I/Q signals comprises: performing amplitude and phase calibration on the four paths of synchronous baseband I/Q signals, and eliminating hardware deviation among channels; the differential processing of the echo signals of the main emission and the auxiliary emission is used for removing millimeter-level macro-displacement interference caused by human respiration and slight body shaking, and retaining sub-millimeter-level micro-displacement signals of vascular pulsation.
3. The MIMO UWB-based non-contact blood pressure monitoring method of claim 1, wherein the performing spatial spectrum estimation on the preprocessed multichannel echo signals to locate a blood vessel region to be measured, extracting a continuous phase change sequence through multichannel joint phase unwrapping, converting the continuous phase change sequence into a sub-millimeter micro-displacement pulse wave time sequence signal corresponding to human body blood vessel pulsation, and extracting multichannel pulse wave spatial characteristic parameters comprises: Performing DOA direction of arrival estimation on the preprocessed multichannel echo signals through a MIMO spatial spectrum estimation algorithm, positioning a blood vessel region to be detected of a radial artery or a precordial region of a human body, and extracting effective echo signals corresponding to the region; Adopting an improved branch cutting method based on the pretreated multi-channel echo signal phase consistency constraint to perform joint phase unwrapping on the multi-channel echo signal, eliminating phase jump under low signal-to-noise ratio, and obtaining a continuous jump-free phase change sequence; According to the linear relation between the phase change and the displacement, converting the phase change sequence into a sub-millimeter micro-displacement time sequence signal corresponding to the pulse of the human blood vessel, namely a multichannel pulse wave signal; And extracting the time domain characteristics of the single-channel pulse wave, and spatial characteristic parameters including pulse wave phase difference, amplitude ratio, pulse wave conduction time difference PATD and wave peak arrival time difference among four paths of receiving channels.
4. The MIMO UWB-based non-contact blood pressure monitoring method of claim 1, wherein the pre-calibrated multi-channel feature-blood pressure mapping model is a LightGBM algorithm-based multi-feature fusion regression model; The input features of the model comprise single-channel time domain features and multi-channel space features, the model is obtained by training crowd samples covering different ages, sexes, body mass indexes and motion states, and the labels of the training samples are blood pressure true values measured by a standard mercury sphygmomanometer; The model supports personalized calibration of a user, and the individual detection precision is improved through at least three groups of standard blood pressure measured values input by the user and corresponding multi-channel pulse wave characteristic parameters based on transfer learning fine adjustment of optimized model parameters.
5. The MIMO UWB-based non-contact blood pressure monitoring method of claim 1, wherein the outputting and storing the blood pressure detection result comprises: Carrying out wireless data transmission with a mobile phone APP and an upper computer through a Bluetooth communication unit; Triggering an alarm when the sound-light alarm unit detects that the blood pressure value exceeds a preset normal range.
6. The method for MIMO UWB-based non-contact blood pressure monitoring of claim 3 further comprising quadrature demodulating the effective echo signal to obtain a multichannel phase sequence wrapped in a [ -pi, pi ].
7. The MIMO UWB-based non-contact blood pressure monitoring method of claim 6, wherein the phase change sequence is converted into a sub-millimeter micro-displacement time sequence signal corresponding to the human vascular pulsation, and is specifically calculated by the following formula: wherein Deltad is a sub-millimeter micro-displacement time sequence signal, For the wavelength corresponding to the center frequency of the UWB signal, Is a continuous, non-hopped sequence of phase changes.
8. The MIMO UWB-based non-contact blood pressure monitoring method of claim 4, wherein the sample data is subjected to min-max normalization preprocessing to eliminate feature dimension differences, and the training set and the test set are divided according to a ratio of 8:2.
9. The MIMO UWB-based non-contact blood pressure monitoring method of claim 8, wherein the model after pre-calibration and personalized fine tuning is iteratively optimized, and the feature weights and iteration parameters of the model are dynamically adjusted based on multiple batches of measured data.
10. A MIMO UWB-based non-contact blood pressure detection system, the system comprising: The MIMO UWB radar front-end module is used for transmitting ultra-wideband impulse signals through time division duplex of the double transmitting antennas, synchronously collecting echo signals reflected by a part to be detected of a human body through four paths of receiving antennas and outputting four paths of synchronous baseband I/Q signals; The multichannel synchronous preprocessing module is used for carrying out channel consistency calibration, DC removal, static clutter suppression and differential motion compensation processing on the four-channel synchronous baseband I/Q signals to obtain preprocessed multichannel echo signals; the MIMO space domain pulse wave extraction module is used for carrying out space spectrum estimation on the preprocessed multichannel echo signals to locate a blood vessel region to be detected, extracting a continuous phase change sequence through multichannel combined phase unwrapping, converting the continuous phase change sequence into sub-millimeter micro-displacement pulse wave time sequence signals corresponding to human body blood vessel pulsation, and extracting multichannel pulse wave space characteristic parameters; the multichannel characteristic fusion blood pressure inversion module is used for calculating and obtaining the systolic pressure and diastolic pressure values of the human body according to the multichannel pulse wave space characteristic parameters based on a pre-calibrated multichannel characteristic-blood pressure mapping model; and the output interaction module is used for outputting and storing the blood pressure detection result and realizing communication interaction with the upper terminal.

Description

MIMO UWB-based non-contact blood pressure detection method and system Technical Field The invention relates to the technical field of ultra wideband, in particular to a non-contact blood pressure detection method and system based on MIMO UWB. Background Blood pressure is one of the core physiological parameters for evaluating cardiovascular health of human bodies, and real-time and accurate blood pressure monitoring has important significance for early screening of hypertension, chronic disease management and early warning of cardiovascular diseases. In the prior art, the main blood pressure detection methods are divided into two types, namely invasive detection and noninvasive detection. The invasive detection method has the highest detection precision, belongs to invasive operation, is only suitable for intensive care scenes of hospitals, and cannot be used for daily monitoring. In noninvasive detection, the most widely applied is a cuff type electronic sphygmomanometer, measurement is realized based on an oscillography method or a Korotkoff sound method, but the cuff is required to be inflated and pressurized in the measurement process, so that discomfort is brought to a user, continuous and dynamic blood pressure monitoring cannot be realized, and meanwhile, the tightness and the measurement posture of the cuff can influence the measurement accuracy. In recent years, non-contact physiological parameter detection technology has rapidly developed, wherein Ultra Wide Band (UWB) radar technology is widely applied to non-contact heart rate and respiration detection. The UWB radar has the advantages of strong penetrability, low power consumption, high ranging precision and strong anti-interference capability by transmitting nanosecond ultra-wideband impulse signals, and is very suitable for non-contact detection of micro-displacement of the body surface of a human body. Therefore, it is necessary to provide a method and a system for non-contact blood pressure detection based on MIMO UWB to effectively solve the above-mentioned problems. Disclosure of Invention Aiming at the problems and the defects existing in the prior art, the invention provides a non-contact blood pressure detection method and a non-contact blood pressure detection system based on MIMO UWB, and high-precision micro-displacement monitoring is realized through the processing of multi-channel combined signals. The embodiment of the invention provides a non-contact blood pressure detection method based on MIMO UWB, which comprises the following steps: transmitting ultra-wideband impulse signals through time division duplex of the double transmitting antennas, synchronously collecting echo signals reflected by a part to be detected of a human body through four paths of receiving antennas, and outputting four paths of synchronous baseband I/Q signals; Carrying out channel consistency calibration, DC removal, static clutter suppression and differential motion compensation on the four paths of synchronous baseband I/Q signals to obtain preprocessed multichannel echo signals; Carrying out space spectrum estimation on the preprocessed multichannel echo signals to locate a blood vessel region to be detected, extracting a continuous phase change sequence through multichannel combined phase unwrapping, converting the continuous phase change sequence into sub-millimeter micro-displacement pulse wave time sequence signals corresponding to human body blood vessel pulsation, and extracting multichannel pulse wave space characteristic parameters; Calculating to obtain the systolic pressure and diastolic pressure values of the human body according to the multichannel pulse wave space characteristic parameters based on a pre-calibrated multichannel characteristic-blood pressure mapping model; and outputting and storing a blood pressure detection result to realize communication interaction with the upper terminal. Preferably, the performing channel consistency calibration, dc removal, static clutter suppression and differential motion compensation on the four-channel synchronous baseband I/Q signal includes: performing amplitude and phase calibration on the four paths of synchronous baseband I/Q signals, and eliminating hardware deviation among channels; the differential processing of the echo signals of the main emission and the auxiliary emission is used for removing millimeter-level macro-displacement interference caused by human respiration and slight body shaking, and retaining sub-millimeter-level micro-displacement signals of vascular pulsation. Preferably, the performing spatial spectrum estimation on the preprocessed multichannel echo signals to locate a blood vessel region to be detected, extracting a continuous phase change sequence through multichannel combined phase unwrapping, converting the continuous phase change sequence into sub-millimeter-level micro-displacement pulse wave time sequence signals corresponding to human body blood vessel pulsa