Millimeter wave sensing-based breathing machine adaptive regulation and control method and system

Abstract

The invention provides a breathing machine self-adaptive regulation and control method and system based on millimeter wave perception, and when a breathing machine works, the method comprises the following steps: transmitting an original signal to the chest position of a user by using a millimeter wave radar, and collecting an echo signal of the original signal reflected by a human body; the method comprises the following steps: decomposing echo signals collected at a current moment and in a previous preset time period into a plurality of distance units, screening effective distance units from the plurality of distance units, screening the effective distance units with the maximum phase fluctuation as a target distance unit, and extracting original phase information from the target distance unit; purifying the extracted original phase information by adopting a sliding window; and dynamically generating detection thresholds of a peak value and a valley value at the current moment according to historical original phase information by utilizing a peak value extraction algorithm, comparing the original phase information at the current moment with the detection thresholds of the peak value and the valley value at the current moment so as to identify the breathing state of the current user, and regulating and controlling the working state of the breathing machine according to the breathing state of the user.

Description

Technical Field

[0001] This invention relates to the field of millimeter-wave sensing technology, and in particular to a method and system for adaptive control of ventilators based on millimeter-wave sensing. Background Technology

[0002] In recent years, respiratory infectious diseases have shown a global pandemic trend, leading to a significant increase in the number of severe cases. Against this backdrop, a large number of critically ill patients are experiencing severe complications such as acute respiratory failure, making their clinical need for respiratory support equipment, especially ventilators, particularly urgent. This surge in demand for medical equipment not only reflects the severity of the current public health crisis but also poses a significant challenge to the emergency response capabilities of healthcare systems in various countries. During the initial setup and use of ventilators in clinical and home settings, users often experience "air choking," a sensation of airflow impact during inhalation or exhalation, leading to aversion to ventilator use. This phenomenon mainly stems from the asynchrony between the patient's voluntary rhythm and the airflow output by the ventilator (such as pressure, flow rate, and ventilation rate), especially common in patients experiencing ventilator use for the first time, as well as those who are unconscious or unable to regulate their breathing independently. Furthermore, traditional ventilators rely on preset or passive feedback methods to adjust ventilation parameters, making it difficult to promptly identify the patient's real-time respiratory status, thus exacerbating the discomfort caused by patient-ventilator asynchrony.

[0003] Currently, there are four main methods for controlling airflow switching in ventilators. The first is time-based switching, where the ventilator automatically switches to the expiratory phase based on a preset inspiratory time. However, this method cannot adjust to the patient's actual breathing needs, leading to asynchrony between the patient and the ventilator, causing significant discomfort. The second is pressure-based switching, where the ventilator switches to the expiratory phase when the airway pressure reaches a predetermined value. However, this method has the drawback of unstable tidal volume, heavily influenced by lung compliance and airway resistance; improper settings can cause lung damage. The third is flow-based switching, where the ventilator switches to the expiratory phase when the inspiratory flow rate drops to a set threshold. This method requires highly sensitive sensors and may fail to accurately detect the correct inspiratory-expiratory switching time, leading to user discomfort. The fourth method is volume-based switching, where the ventilator switches to the expiratory phase when the delivered tidal volume reaches a predetermined value. However, the tidal volume actively received by a person with each breath is not perfectly consistent; fluctuations can increase the risk of lung damage. In addition, there is a composite switching strategy that combines two or more of the above switching methods, such as time and pressure, or time and flow. However, this method requires setting more parameters and is relatively more complex to operate. Overall, current ventilator airflow control strategies are unable to accurately reflect the rhythm and needs of the patient's spontaneous breathing in real time, and problems such as patient-ventilator asynchrony still exist. This may still lead to a mismatch between respiratory control and the patient's actual rhythm, thereby affecting user comfort and treatment effectiveness.

[0004] Another feasible ventilator airflow control solution involves monitoring the user's breathing status in real time using a breathing belt and dynamically adjusting ventilator parameters based on this data. However, a significant drawback of this method is that the breathing belt itself physically restricts the user's chest movement, limiting the smoothness of natural breathing and significantly increasing user discomfort. Especially in long-term use scenarios, this discomfort may lead to decreased user compliance.

[0005] Therefore, developing an adaptive control method for ventilators that can both ensure the accuracy of monitoring the user's respiratory status and maximize the user's comfort is a technical problem that urgently needs to be solved and has become a pressing need for current clinical applications. Summary of the Invention

[0006] In view of this, embodiments of the present invention provide a method and system for adaptive control of ventilators based on millimeter wave sensing, in order to eliminate or improve one or more defects existing in the prior art.

[0007] One aspect of the present invention provides a millimeter-wave sensing-based adaptive control method for a ventilator, the method comprising the following steps: transmitting a raw signal toward the user's chest cavity using a millimeter-wave radar and acquiring the echo signal after the raw signal is reflected by the human body; decomposing the echo signal acquired at the current moment and within a previously preset time period into multiple distance units, selecting those distance units whose average signal intensity exceeds a first preset threshold as effective distance units, selecting those effective distance units with the largest phase fluctuation as target distance units, and extracting raw phase information from the target distance units; purifying the extracted raw phase information using a sliding window; dynamically generating detection thresholds for peak and trough values ​​at the current moment using a peak extraction algorithm based on historical raw phase information, comparing the raw phase information at the current moment with the detection thresholds for peak and trough values ​​at the current moment to identify the user's current respiratory state, and adjusting the ventilator's operating state according to the user's respiratory state.

[0008] In some embodiments of the present invention, the original signal is a frequency-modulated continuous wave signal, and the echo signal is composed of multiple signal segments; the step of decomposing the echo signal collected at the current time and within a previously preset time period into multiple distance units includes: performing a fast Fourier transform in the range direction on each signal segment in the echo signal, thereby decomposing the echo signal into multiple distance units, each distance unit corresponding to a specific spatial distance position, representing the reflected echo information at that position, the original phase information can be extracted based on the reflected echo information, and the phase fluctuation can be analyzed based on the reflected echo information at the current time and within a previously preset time period.

[0009] In some embodiments of the present invention, the plurality of range units and the spatial distance position corresponding to each range unit are stored in the form of a two-dimensional radar spatiotemporal information range information complex matrix in the distance-time dimension; the step of extracting the original phase information from the target range unit includes: filtering out a one-dimensional complex sequence corresponding to the target range unit from the two-dimensional radar spatiotemporal information range information complex matrix; extracting phase information from the one-dimensional complex sequence, and performing phase unwrapping processing on the extracted phase information to finally obtain the original phase information.

[0010] In some embodiments of the present invention, the step of purifying the extracted raw phase information using a sliding window includes: for the raw phase information, removing the baseline using a first type of sliding window and removing high-frequency noise using a second type of sliding window; the length of the first type of sliding window is greater than the length of the second type of sliding window.

[0011] In some embodiments of the present invention, the step of removing the baseline using a first type of sliding window includes: using a first type of sliding window to identify the baseline offset included in the original phase information, and removing the baseline by subtracting the original phase information from the baseline offset; the step of removing high-frequency noise using a second type of sliding window includes: using a second type of sliding window to perform signal smoothing operation, thereby removing high-frequency noise.

[0012] In some embodiments of the present invention, the step of dynamically generating the detection thresholds for peak and valley values ​​at the current moment using a peak extraction algorithm based on historical raw phase information includes: obtaining raw phase information extracted from the target distance unit within the most recent preset time period; wherein the preset time period includes multiple peak and valley cycles; analyzing the raw phase information to obtain the average amplitude of peak and valley values, and calculating the average time interval between peak and valley values; dynamically generating the detection thresholds for peak and valley values ​​at the current moment based on the average amplitude of peak and valley values ​​and the average time interval between peak and valley values.

[0013] In some embodiments of the present invention, the step of comparing the original phase information at the current moment with the detection thresholds for peak and trough values ​​at the current moment to identify the current user's breathing state includes: when the amplitude increase of the original phase information at the current moment is detected to exceed the peak detection threshold, the original phase information at the current moment is determined to be a peak value, and the current user's breathing state is switching from exhalation to inhalation; when the amplitude decrease of the original phase information at the current moment is detected to exceed the trough detection threshold, the original phase information at the current moment is determined to be a trough value, and the current user's breathing state is switching from inhalation to exhalation; otherwise, it is regarded as an ordinary point that is neither a peak nor a trough, and the current user's breathing state is inhalation or exhalation.

[0014] In some embodiments of the present invention, the step of adjusting the working state of the ventilator according to the user's breathing state includes: when the user's breathing state is switching from exhalation to inhalation, adjusting the expiratory valve of the ventilator from open to closed and the inhalation valve from closed to open; when the user's breathing state is switching from inhalation to exhalation, adjusting the expiratory valve of the ventilator from closed to open and the inhalation valve from open to closed; when the user's breathing state is inhalation or exhalation, maintaining the current working state of the ventilator.

[0015] Corresponding to the above method, the present invention also provides a millimeter-wave sensing-based adaptive control system for ventilators. This system includes: a millimeter-wave radar for transmitting raw signals toward the user's chest cavity; a signal receiving module for acquiring echo signals after the raw signals are reflected by the human body; a signal processing module for decomposing the echo signals acquired at the current moment and within a previously preset time period into multiple distance units, selecting those with an average signal intensity exceeding a first preset threshold as effective distance units, selecting the distance units with the largest phase fluctuation as target distance units, and extracting raw phase information from the target distance units; further, it is used to purify the extracted raw phase information using a sliding window; an event recognition module for dynamically generating detection thresholds for peak and trough values ​​at the current moment based on historical raw phase information using a peak extraction algorithm, comparing the raw phase information at the current moment with the detection thresholds for peak and trough values ​​at the current moment to identify the current user's respiratory state; and an event control module for controlling the ventilator's operating state according to the user's respiratory state. This adaptive control system for ventilators is used to implement the steps of any of the methods described in the above embodiments.

[0016] In some embodiments of the present invention, the original signal is a frequency-modulated continuous wave signal, and the echo signal is composed of multiple signal segments; the multiple range units and the spatial distance position corresponding to each range unit are stored in the form of a two-dimensional radar spatiotemporal information range information complex matrix in the distance-time dimension; the user's breathing state includes switching from exhalation to inhalation, switching from inhalation to exhalation, and inhalation and exhalation.

[0017] The adaptive control method and system for ventilators based on millimeter-wave sensing proposed in this invention can monitor the user's respiratory status (respiratory rhythm) using millimeter-wave radar, accurately determine the timing of the switch between inhalation and exhalation, and thus perform real-time adaptive control based on the user's respiratory status, dynamically controlling the opening and closing states of the inspiratory and expiratory valves, making the ventilator more in line with the individual's natural breathing habits, and significantly improving the comfort and synchronization during ventilator use.

[0018] Additional advantages, objects, and features of the invention will be set forth in part in the description which follows, and will also become apparent in part to those skilled in the art upon studying the description, or may be learned by practice of the invention. The objects and other advantages of the invention can be realized and obtained by means of the structures specifically pointed out in the description and drawings.

[0019] Those skilled in the art will understand that the objectives and advantages achievable with the present invention are not limited to those specifically described above, and that the above and other objectives achievable with the present invention will become clearer from the following detailed description. Attached Figure Description

[0020] The accompanying drawings, which are included to provide a further understanding of the invention and form part of this application, are not intended to limit the scope of the invention. In the drawings: Figure 1 This is a flowchart of a ventilator adaptive control method based on millimeter wave sensing in one embodiment of the present invention.

[0021] Figure 2 This is a diagram of the ventilator event response model architecture in one embodiment of the present invention. Detailed Implementation

[0022] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the embodiments and accompanying drawings. Here, the illustrative embodiments and descriptions of this invention are used to explain the invention, but are not intended to limit the invention.

[0023] It should also be noted that, in order to avoid obscuring the invention with unnecessary details, only the structures and / or processing steps closely related to the solution according to the invention are shown in the accompanying drawings, while other details that are not closely related to the invention are omitted.

[0024] It should be emphasized that the term "including / comprises" as used herein refers to the presence of a feature, element, step, or component, but does not exclude the presence or addition of one or more other features, elements, steps, or components.

[0025] It should also be noted that, unless otherwise specified, the term "connection" in this article can refer not only to a direct connection, but also to an indirect connection involving an intermediary.

[0026] In the following description, embodiments of the invention will be illustrated with reference to the accompanying drawings. In the drawings, the same reference numerals represent the same or similar parts, or the same or similar steps.

[0027] To overcome the problems of existing technologies, this invention proposes an adaptive control method and system for ventilators based on millimeter-wave sensing. This method combines traditional ventilators with millimeter-wave sensing technology and dynamically adjusts key parameters in the ventilator ventilation process based on real-time respiratory information acquired by millimeter-wave radar.

[0028] Millimeter-wave radar is a non-contact vital sign sensing technology that can use high-precision, real-time monitoring of the subtle chest and abdominal movements caused by respiration, thereby calculating precise respiratory information such as respiratory rate and the timing of inhalation and exhalation. By integrating millimeter-wave sensing into the ventilator control system, control signals can be promptly sent to the ventilator based on the respiratory information sensed by the millimeter waves. This monitoring and control method facilitates more intelligent and precise human-machine synchronization, significantly improving the comfort and compliance of ventilator users.

[0029] Figure 1 This is a flowchart of a ventilator adaptive control method based on millimeter-wave sensing in one embodiment of the present invention. When the ventilator is operating, the method includes the following steps: Step S110: Use millimeter-wave radar to transmit the original signal toward the user's chest cavity and collect the echo signal after the original signal is reflected by the human body.

[0030] Step S120: Decompose the echo signal collected at the current time and within the previous preset time period into multiple distance units, select the effective distance units from the multiple distance units whose average signal strength exceeds a first preset threshold, select the target distance unit from the effective distance units whose phase fluctuation is the largest, and extract the original phase information from the target distance unit.

[0031] It should be noted that, considering the fact that the human body generally does not undergo significant movements during ventilator use, only the (chest) movement caused by breathing needs to be considered. Therefore, the distance unit with the largest phase fluctuation represents the area on the user's chest that fluctuates most dramatically with breathing. The raw phase information (i.e., chest micro-movement information) extracted from the target distance unit best reflects whether the user is currently inhaling, exhaling, switching from exhalation to inhalation, or switching from inhalation to exhalation.

[0032] Step S130: Use a sliding window to purify the extracted raw phase information.

[0033] Step S140: Using a peak extraction algorithm, dynamically generate detection thresholds for peak and trough values ​​at the current moment based on historical raw phase information. Compare the raw phase information at the current moment with the detection thresholds for peak and trough values ​​at the current moment to identify the current user's breathing state and adjust the working state of the ventilator according to the user's breathing state.

[0034] The aforementioned respiratory states include switching from exhalation to inhalation, switching from inhalation to exhalation, and inhalation and exhalation. Adjusting the ventilator's operating state refers to adjusting the opening and closing of the ventilator's inspiratory and expiratory valves. When switching from exhalation to inhalation, the inspiratory valve opens and the expiratory valve closes; when switching from inhalation to exhalation, the expiratory valve opens and the inspiratory valve closes. Switching from exhalation to inhalation and from inhalation to exhalation can be summarized as respiratory switching events.

[0035] The millimeter-wave sensing-based adaptive control method for ventilators proposed in this invention can monitor the user's respiratory status (breathing rhythm) using millimeter-wave radar, accurately determine the timing of the switch between inhalation and exhalation, and thus perform real-time adaptive control based on the user's respiratory status, dynamically controlling the opening and closing states of the inspiratory and expiratory valves, making the ventilator more in line with the individual's natural breathing habits, and significantly improving the comfort and synchronization during ventilator use.

[0036] In some embodiments of the present invention, the original signal is a frequency-modulated continuous wave signal, and the echo signal is composed of multiple signal segments.

[0037] Accordingly, the step of decomposing the echo signal acquired at the current time and within the previously preset time period into multiple distance units includes: performing a fast Fourier transform in the range direction on each signal segment in the echo signal, thereby decomposing the echo signal into multiple distance units, each distance unit corresponding to a specific spatial distance position, representing the reflected echo information at that position, extracting the original phase information based on the reflected echo information, and analyzing phase fluctuations based on the reflected echo information at the current time and within the previously preset time period.

[0038] Using this embodiment of the invention, it is possible to acquire raw phase information containing information on the fluctuations of the user's chest cavity position using millimeter-wave radar, and then analyze the user's breathing (switching) state based on the raw phase information.

[0039] In some embodiments of the present invention, the plurality of distance units and the spatial distance position corresponding to each distance unit are stored in the form of a two-dimensional radar spatiotemporal information distance information complex matrix in the distance-time dimension.

[0040] Accordingly, the step of extracting the original phase information from the target range cell includes: filtering out a one-dimensional complex sequence corresponding to the target range cell from the complex matrix of the two-dimensional radar spatiotemporal range information; extracting phase information from the one-dimensional complex sequence; and performing phase unwrapping processing on the extracted phase information to finally obtain the original phase information.

[0041] Using this embodiment of the invention, a complex matrix can be used to store distance units and the spatial distance position corresponding to each distance unit, and subsequent calculations can be supported.

[0042] In the field of radar technology, range cell is an existing concept. A range cell is essentially a discrete "basket" or "grid" dividing the continuous distance between the radar and the target. Each grid represents a specific range, and the echo signal received by the radar is placed into the corresponding "grid" according to its arrival time. In this scheme, the target refers to the user's chest cavity. Decomposing the echo signal collected at the current moment and within a previously preset time period into multiple range cells means dividing the user's chest cavity position into discrete regions. The selected target range cell is the most representative and best reflects the "rise and fall" of the chest cavity. Therefore, by extracting the original phase information of the target range cell, the "rise and fall" state of the user's chest cavity at this moment can be analyzed, thereby determining whether the user is inhaling, exhaling, switching from exhalation to inhalation, or switching from inhalation to exhalation.

[0043] In some embodiments of the present invention, the step of purifying the extracted raw phase information using a sliding window includes: for the raw phase information, removing the baseline using a first type of sliding window and removing high-frequency noise using a second type of sliding window; the length of the first type of sliding window is greater than the length of the second type of sliding window.

[0044] Using this embodiment of the invention, baseline removal and high-frequency noise removal can be achieved respectively by using sliding windows of different lengths.

[0045] In some embodiments of the present invention, the step of removing the baseline using a first type of sliding window includes: using the first type of sliding window to identify the baseline offset included in the original phase information, and performing baseline removal by subtracting the original phase information from the baseline offset. In other embodiments of the present invention, the step of removing high-frequency noise using a second type of sliding window includes: performing signal smoothing operation using the second type of sliding window to remove high-frequency noise. The length of the first type of sliding window can be 2 seconds, and the length of the second type of sliding window can be 0.1 seconds.

[0046] The above embodiments of the invention provide detailed examples of achieving baseline removal and high-frequency noise removal.

[0047] In some embodiments of the present invention, the step of dynamically generating the detection thresholds for peak and valley values ​​at the current moment using a peak extraction algorithm based on historical raw phase information includes: obtaining raw phase information extracted from the target distance unit within the most recent preset time period; wherein the preset time period includes multiple peak and valley cycles; analyzing the raw phase information to obtain the average amplitude of peak and valley values, and calculating the average time interval between peak and valley values; dynamically generating the detection thresholds for peak and valley values ​​at the current moment based on the average amplitude of peak and valley values ​​and the average time interval between peak and valley values.

[0048] In some specific embodiments of the present invention, the detection thresholds for the peak and valley values ​​at the current moment are dynamically generated based on the average amplitude of the peak and valley values ​​and the average time interval between the peak and valley values. This can be achieved by multiplying the average amplitude and average time interval by a fixed coefficient, and the result is used as the detection threshold for the peak and valley values ​​at the current moment. The fixed coefficient can be set to 0.90, 0.95, 0.99, etc.

[0049] By employing this embodiment of the invention, detection thresholds for peak and trough values ​​at the current moment can be dynamically generated based on historical data, and peak and trough data can be updated in real time, thereby ensuring the validity of the data.

[0050] In some specific embodiments of the present invention, the step of comparing the original phase information at the current moment with the detection thresholds for peak and trough values ​​at the current moment to identify the current user's breathing state includes: when the amplitude increase of the original phase information at the current moment exceeds the peak detection threshold, the original phase information at the current moment is determined to be a peak value, and the current user's breathing state is switching from exhalation to inhalation; when the amplitude decrease of the original phase information at the current moment exceeds the trough detection threshold, the original phase information at the current moment is determined to be a trough value, and the current user's breathing state is switching from inhalation to exhalation; otherwise, it is regarded as an ordinary point that is neither a peak nor a trough, and the current user's breathing state is either inhalation or exhalation.

[0051] Using this embodiment of the invention, the current user's breathing status can be determined in real time based on the original phase information and the pre-calculated detection thresholds for the current peak and trough values.

[0052] In some specific embodiments of the present invention, the step of adjusting the working state of the ventilator according to the user's breathing state includes: when the user's breathing state is switching from exhalation to inhalation, adjusting the expiratory valve of the ventilator from open to closed and the inhalation valve from closed to open; when the user's breathing state is switching from inhalation to exhalation, adjusting the expiratory valve of the ventilator from closed to open and the inhalation valve from open to closed; when the user's breathing state is inhalation or exhalation, maintaining the current working state of the ventilator.

[0053] By employing this embodiment of the invention, the operating state of the ventilator can be adjusted in a targeted manner based on the user's breathing state determined in a prior step, especially the expiratory valve and inspiratory valve of the ventilator, thereby achieving targeted adaptive control of the ventilator and providing the user with a more comfortable experience.

[0054] In some embodiments of the present invention, the adaptive control of the ventilator may be achieved by sending the real-time detected breathing switching event to the ventilator for adaptive adjustment, or by the event control module generating a ventilator control command based on the breathing switching event and sending the ventilator control command to the ventilator for adaptive adjustment.

[0055] From another perspective, this scheme can be divided into two stages: acquiring thoracic micromotion information (using raw phase information to characterize raw phase information) and determining respiratory status (especially respiratory switching events).

[0056] The steps for acquiring information on thoracic cavity micromovements are as follows: First, the echo signal sensed by the millimeter wave at that moment and within the previous 5 seconds was acquired. This echo signal consists of multiple signal segments.

[0057] Then, a Fast Fourier Transform (FFT) is performed on each signal segment in the range direction. This is a processing method for FMCW signals. After FFT processing, the frequency domain information of the signal segment can be converted into range cells, that is, the range dimension corresponding to each slow time segment. This decomposes the continuous echo signal into multiple range cells, each range cell corresponding to a specific spatial range location, representing the reflected echo information at that location. The final result is a two-dimensional radar spatiotemporal information range information complex matrix in the range-time dimension.

[0058] Furthermore, considering that the human body generally does not undergo significant movements during ventilator use, with only breathing movements, this invention filters out effective distance units with average signal strength exceeding the noise threshold based on the average signal strength of each distance unit, thus eliminating background noise interference. Among these effective distance units, the fluctuation amplitude of their phase over time is further analyzed, and the distance unit with the largest phase fluctuation is selected as the final target distance unit for respiratory signal extraction.

[0059] Finally, based on the selected target range cells, a one-dimensional complex sequence corresponding to the range cell is selected from the complex matrix of range information in the two-dimensional radar spatiotemporal information. Then, phase information is extracted from this one-dimensional complex sequence and phase unwrapping processing is performed to obtain the original phase information. The output original phase information is used as the basis for subsequent signal processing and analysis.

[0060] During the respiratory status determination phase, the respiratory switching events reflected in the chest cavity micro-movement information at the current moment are monitored in real time. If no events are detected, it means that the expiratory and inspiratory valves of the ventilator do not need to be adjusted. The steps for determining the respiratory status phase are as follows: First, after acquiring the original phase information, the first type of sliding window is used to identify the baseline offset included in the original phase information, and the baseline offset is removed by subtracting the original phase information from the baseline offset (the original phase information is a one-dimensional real number sequence with a length equal to the number of slow time sampling points).

[0061] Then, a second type of sliding window is used to smooth the signal, thereby removing high-frequency noise. The output is the final smoothed aroma information, which is used as the respiratory micro-movement information to be processed.

[0062] Then, the peak and valley information corresponding to a specific time window are extracted using a peak extraction algorithm.

[0063] Finally, based on the historical peak and trough values, as well as the intervals between peaks and troughs, the user's breathing state at that moment is obtained. If the moment is identified as a peak, it indicates a switch from exhalation to inhalation; conversely, if it is a trough, it indicates a switch from inhalation to exhalation.

[0064] In some embodiments of the present invention, the detailed process of the peak extraction algorithm is as follows: First, the average amplitude of several peaks and troughs closest to the current moment, as well as the average time interval between them, are calculated. Based on the average amplitude of several peaks and troughs and the average time interval between them, the detection thresholds for peaks and troughs are dynamically generated to adapt to real-time changes in respiratory signals.

[0065] Then, several sampling points are taken forward from the end point, where the end point represents the current time. The amplitude difference between the end point and its adjacent points is calculated, and these differences are compared with the above thresholds: if the amplitude increase of the end point (relative to the previous two points) exceeds the peak threshold, it is determined to be a peak; if the relative amplitude decrease exceeds the valley threshold, it is determined to be a valley; otherwise, it is regarded as an ordinary point that is neither a peak nor a valley.

[0066] Finally, the discrimination results are output, and the peak / valley history list is updated in a timely manner when a peak or valley is identified, providing support for the next round of threshold calculation, thereby ensuring that even the respiratory event of the last sampling point can be accurately detected and used for ventilator adaptive control.

[0067] This invention proposes a robust signal peak extraction algorithm. The algorithm first dynamically generates detection thresholds for peaks and troughs, adapting to signal variations based on the average amplitude and time interval of recent peaks and troughs. Then, it determines whether a point is a peak or trough by comparing the amplitude difference between the last point and the two preceding points with the threshold: if the amplitude increase exceeds the peak threshold, it is a peak; if the amplitude decrease exceeds the trough threshold, it is a trough; otherwise, it is a normal point. Finally, the results are output and the historical peak-trough list is updated to ensure the accuracy of real-time detection and adaptive ventilation control, including the last sampling point.

[0068] Figure 2 This is a diagram of the ventilator event response model architecture in one embodiment of the present invention.

[0069] like Figure 2 In the illustrated embodiment, when the ventilator is started, the millimeter-wave radar is also simultaneously activated, continuously monitoring the user's breathing events. When the radar detects a switch from exhalation to inhalation, the millimeter-wave radar sends two commands to the ventilator: closing the expiratory valve and opening the inspiratory valve (i.e., the expiratory valve goes from open to closed + the inspiratory valve goes from closed to open); conversely, when a switch from inhalation to exhalation is detected, the millimeter-wave radar sends two commands to the ventilator: closing the inspiratory valve and opening the expiratory valve (i.e., the expiratory valve goes from closed to open + the inspiratory valve goes from open to closed), thereby achieving valve switching that matches the user's natural breathing rhythm.

[0070] Corresponding to the above method, the present invention also provides a millimeter-wave sensing-based adaptive control system for ventilators. This system includes: a millimeter-wave radar for transmitting raw signals toward the user's chest cavity; a signal receiving module for acquiring echo signals after the raw signals are reflected by the human body; a signal processing module for decomposing the echo signals acquired at the current moment and within a previously preset time period into multiple distance units, selecting those with an average signal intensity exceeding a first preset threshold as effective distance units, selecting the distance units with the largest phase fluctuation as target distance units, and extracting raw phase information from the target distance units; further, it is used to purify the extracted raw phase information using a sliding window; an event recognition module for dynamically generating detection thresholds for peak and trough values ​​at the current moment based on historical raw phase information using a peak extraction algorithm, comparing the raw phase information at the current moment with the detection thresholds for peak and trough values ​​at the current moment to identify the current user's respiratory state; and an event control module for controlling the ventilator's operating state according to the user's respiratory state. This adaptive control system for ventilators is used to implement the steps of any of the methods described in the above embodiments.

[0071] In another embodiment of the present invention, the specific functions of each module of the millimeter-wave sensing-based adaptive control system for ventilators are illustrated, as follows: Signal receiving module: While the ventilator is operating, the millimeter-wave radar is aimed at the user's chest cavity, transmitting the raw signal towards the user and receiving the echo signal after the raw signal is reflected by the body. The raw signal can be an FMCW (Frequency Modulated Continuous Wave) signal, and the echo signal consists of multiple signal segments. The sampling rate of these signal segments is called the fast time, and multiple signal segments are spliced ​​together to form a large signal set; the overall sampling rate of this signal set is called the slow time.

[0072] Signal processing module: The received echo signal undergoes differential processing, phase extraction, outlier removal, and bandpass filtering sequentially to extract the purified respiratory signal / chest cavity micro-movement information caused by respiration. The differential processing refers to performing differential processing on the echo signal based on the original signal.

[0073] Event recognition module: It uses peak extraction algorithm to extract and detect the user's current breathing status reflected in the chest cavity micro-movement information in real time, especially breathing switching events (including the transition from exhalation to inhalation and the transition from inhalation to exhalation).

[0074] Event control module: Based on the real-time detected breathing switching event, the working state of the ventilator is adaptively controlled. The adaptive control includes adjusting the opening and closing state of the inspiratory valve and expiratory valve of the ventilator.

[0075] The aforementioned millimeter-wave sensing-based adaptive control system for ventilators combines a millimeter-wave sensing module with a traditional ventilator module. By using millimeter-wave sensing to detect the user's real-time breathing information, it can dynamically control the opening and closing of the expiratory and inspiratory channels, thereby achieving adjustments that better match the user's natural breathing rhythm and significantly improving the comfort of using the ventilator.

[0076] In some embodiments of the present invention, the original signal is a frequency-modulated continuous wave signal, and the echo signal is composed of multiple signal segments.

[0077] In some embodiments of the present invention, the plurality of distance units and the spatial distance position corresponding to each distance unit are stored in the form of a two-dimensional radar spatiotemporal information distance information complex matrix in the distance-time dimension.

[0078] In some embodiments of the present invention, the user's breathing state includes switching from exhalation to inhalation, switching from inhalation to exhalation, and inhalation and exhalation.

[0079] The adaptive control method and system for ventilators based on millimeter-wave sensing proposed in this invention can monitor the user's respiratory status (respiratory rhythm) using millimeter-wave radar, accurately determine the timing of the switch between inhalation and exhalation, and thus perform real-time adaptive control based on the user's respiratory status, dynamically controlling the opening and closing states of the inspiratory and expiratory valves, making the ventilator more in line with the individual's natural breathing habits, and significantly improving the comfort and synchronization during ventilator use.

[0080] Those skilled in the art will understand that the exemplary components, systems, and methods described in conjunction with the embodiments disclosed herein can be implemented in hardware, software, or a combination of both. Whether implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this invention. When implemented in hardware, it can be, for example, electronic circuits, application-specific integrated circuits (ASICs), appropriate firmware, plug-ins, function cards, etc. When implemented in software, the elements of this invention are programs or code segments used to perform the desired tasks. The programs or code segments can be stored in a machine-readable medium or transmitted over a transmission medium or communication link via data signals carried in a carrier wave.

[0081] It should be clarified that the present invention is not limited to the specific configurations and processes described above and shown in the figures. For the sake of brevity, detailed descriptions of known methods are omitted here. In the above embodiments, several specific steps are described and shown as examples. However, the method process of the present invention is not limited to the specific steps described and shown. Those skilled in the art can make various changes, modifications, and additions, or change the order of steps, after understanding the spirit of the present invention.

[0082] In this invention, features described and / or illustrated for one embodiment may be used in the same or similar manner in one or more other embodiments, and / or combined with or in place of features of other embodiments.

[0083] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. For those skilled in the art, various modifications and variations of the embodiments of the present invention are possible. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for millimeter wave perception based adaptive regulation of a ventilator, the method comprising: In the method, when the ventilator is working, the method comprises: ​ The millimeter wave radar emits original signals towards the chest position of the user and collects echo signals of the original signals reflected by the human body; The echo signals collected at the current time and in the previous preset period are decomposed into multiple distance units, effective distance units with average signal strength exceeding a first preset threshold are screened from the multiple distance units, a target distance unit with the largest phase fluctuation is screened from the effective distance units, and original phase information is extracted from the target distance unit; The extracted original phase information is purified by using a sliding window; A peak extraction algorithm is used to dynamically generate detection thresholds of the current time peak value and valley value according to historical original phase information, the original phase information at the current time is compared with the detection thresholds of the current time peak value and valley value to identify the breathing state of the user, and the working state of the ventilator is regulated according to the breathing state of the user.

2. The method of claim 1, wherein, The original signals are frequency-modulated continuous wave signals, and the echo signals are composed of multiple signal segments; The step of decomposing the echo signals collected at the current time and in the previous preset period into multiple distance units comprises: The echo signals are decomposed into multiple distance units by performing fast Fourier transform on each signal segment in the echo signals in the distance direction, each distance unit corresponds to a specific spatial distance position, represents reflection echo information of the position, original phase information can be extracted based on the reflection echo information, and phase fluctuation can be analyzed based on the reflection echo information in the previous preset period.

3. The method of claim 1, wherein, The multiple distance units and the spatial distance positions corresponding to each distance unit are stored in the form of a two-dimensional radar space-time information distance information complex matrix in the distance-time dimension; The step of extracting original phase information from the target distance unit comprises: A one-dimensional complex sequence corresponding to the target distance unit is screened from the two-dimensional radar space-time information distance information complex matrix; Phase information is extracted from the one-dimensional complex sequence, and the extracted phase information is subjected to phase unwrapping processing to finally obtain the original phase information.

4. The method of claim 1, wherein, The step of purifying the extracted original phase information by using a sliding window comprises: for the original phase information, a first type of sliding window is used to remove a baseline, and a second type of sliding window is used to remove high-frequency noise; the length of the first type of sliding window is greater than the length of the second type of sliding window.

5. The method of claim 4, wherein, The step of removing the baseline by using the first type of sliding window comprises: identifying a baseline offset included in the original phase information by using the first type of sliding window, and removing the baseline by using the difference between the original phase information and the baseline offset; The step of removing the high-frequency noise by using the second type of sliding window comprises: performing signal smoothing operation by using the second type of sliding window to remove the high-frequency noise.

6. The method of claim 1, wherein, The step of dynamically generating detection thresholds of the current time peak value and valley value according to historical original phase information by using a peak extraction algorithm comprises: The original phase information extracted from the target distance unit in the recent preset period is obtained; wherein, the preset period contains multiple peak-valley cycles; According to the original phase information analysis, the average amplitude of the peak and the valley is obtained, and the average time interval between the peak and the valley is calculated; According to the average amplitude of the peak and the valley, and the average time interval between the peak and the valley, the detection threshold of the current time peak and valley is dynamically generated.

7. The method of claim 1, wherein, The step of comparing the original phase information at the current time with the detection threshold of the current time peak and valley to identify the current user's breathing state, including: When the amplitude of the original phase information at the current time is detected to increase by more than the detection threshold of the peak, it is determined that the original phase information at the current time is the peak, and the current user's breathing state is switched from exhale to inhale; When the amplitude of the original phase information at the current time is detected to decrease by more than the detection threshold of the valley, it is determined that the original phase information at the current time is the valley, and the current user's breathing state is switched from inhale to exhale; Otherwise, it is considered as neither peak nor valley, and the current user's breathing state is in inhale or exhale.

8. The method of claim 1, wherein, The step of adjusting the working state of the breathing machine according to the user's breathing state, including: When the user's breathing state is switched from exhale to inhale, the exhale valve of the breathing machine is adjusted from open to closed, and the inhale valve is adjusted from closed to open; When the user's breathing state is switched from inhale to exhale, the exhale valve of the breathing machine is adjusted from closed to open, and the inhale valve is adjusted from open to closed; When the user's breathing state is in inhale or exhale, the current working state of the breathing machine is maintained. 9.A ventilator adaptive regulation system based on millimeter wave perception, characterized in that, The system includes: Millimeter wave radar for emitting original signals towards the chest position of the user; Signal receiving module for collecting echo signals of the original signals reflected by the human body; Signal processing module for decomposing the echo signals collected at the current time and in the previous preset period into multiple distance units, selecting the effective distance unit with an average signal strength exceeding a first preset threshold from the multiple distance units, selecting the target distance unit with the largest phase fluctuation from the effective distance units, and extracting the original phase information from the target distance unit; also for purifying the extracted original phase information by using a sliding window; Event recognition module for dynamically generating the detection threshold of the current time peak and valley according to the historical original phase information by using a peak extraction algorithm, and comparing the original phase information at the current time with the detection threshold of the current time peak and valley to identify the current user's breathing state; Event control module for adjusting the working state of the breathing machine according to the user's breathing state; The breathing machine adaptive control system is used to realize the steps of the method of any one of claims 1-8.

10. The system of claim 9, wherein, The original signal is a frequency-modulated continuous wave signal, and the echo signal is composed of multiple signal segments; The multiple distance units and the spatial distance position corresponding to each distance unit are stored in the form of a two-dimensional radar space-time information distance information complex matrix in distance-time dimension; The user's breathing state includes switching from exhale to inhale, switching from inhale to exhale, and inhaling or exhaling.

Images