Ventilation method based on automatic airway pressure feedback for cardiopulmonary resuscitation, and respirator

Abstract

Disclosed in the present invention are a ventilation method based on automatic airway pressure feedback for cardiopulmonary resuscitation, and a respirator. The method comprises: according to a preset sampling frequency, acquiring a current airway pressure signal during cardiac compression for a target person; dynamically correcting the current airway pressure signal on the basis of figure parameters of the target person, so as to acquire a target airway pressure signal; performing continuous analysis on the target airway pressure signal, so as to acquire a compression frequency and a compression phase regarding the target person, and high-dimensional signal information; training a machine learning model by means of preset compression parameters and ventilation rule parameters; and inputting into the machine learning model the compression frequency and the compression phase regarding the target person, the high-dimensional signal information, and mode setting parameters of a ventilation device, so as to determine a ventilation probability and a ventilation time window of the target pressure signal, and on the basis of the ventilation time window, controlling the ventilation device to automatically ventilate the target person. The method improves levels of safety, reliability and stability.

Description

A cardiopulmonary resuscitation airway pressure automatic feedback ventilation method and ventilator Technical Field

[0001] The present invention relates to the technical field of medical first aid, and in particular to a cardiopulmonary resuscitation airway pressure automatic feedback ventilation method and a ventilator. Background Art

[0002] Currently, cardiopulmonary resuscitation (CPR) is the most basic and important medical method for saving the lives of patients experiencing cardiac arrest. Chest compressions combined with forced ventilation are typically the first-aid measures required for these patients to maintain basic blood circulation, heart rate, and breathing. Existing CPR typically includes two key components: cardiac compressions and ventilation (breathing). Cardiac compressions typically take the form of manual compressions and mechanical (equipment) compressions. Manual compressions are used when unavailable, while mechanical (equipment) compressions are used when available in hospitals. Clinical requirements for compressions generally adhere to: an effective chest compression frequency of 100 to 120 times per minute and a depth of 5 to 6 centimeters. Ventilation (breathing) is typically provided through manual breathing, balloon ventilation, and mechanical ventilation during CPR. Ventilation requirements are based on clinical requirements, with 2 ventilations between 30 compressions, 2 ventilations after 15 compressions, and 1 ventilation after every 10 compressions. The existing coordination of cardiopulmonary resuscitation compressions and ventilations is usually coordinated by medical staff, or the cardiopulmonary resuscitation compression equipment and the ventilator are coordinated through a communication protocol. There is a problem that it cannot be coordinated automatically, and the manual coordination method is difficult to accurately and stably control for a long time, which reduces the practicality.

[0003] Summary of the Invention

[0004] In response to the above-mentioned problems, the present invention provides a cardiopulmonary resuscitation airway pressure automatic feedback ventilation method and a ventilator to solve the problem mentioned in the background technology that the existing cardiopulmonary resuscitation compression and ventilation coordination is usually coordinated by medical staff, or the cardiopulmonary resuscitation compression equipment and the ventilator are coordinated through a communication protocol, which cannot be coordinated automatically, and the manual coordination method is difficult to accurately and stably control for a long time, which reduces the practicality.

[0005] A cardiopulmonary resuscitation airway pressure automatic feedback ventilation method comprises the following steps:

[0006] Acquire a current airway pressure signal when compressing the heart of a target person according to a preset sampling frequency;

[0007] Dynamically correct the current airway pressure signal according to the body parameters of the target person to obtain the target airway pressure signal;

[0008] Continuously analyze the target airway pressure signal to obtain the compression frequency and compression phase of the target person and high-dimensional signal information;

[0009] Train the machine learning model by presetting compression parameters and ventilation rule parameters;

[0010] The compression frequency and compression phase of the target person, as well as the high-dimensional signal information and the mode setting parameters of the ventilation equipment are input into the machine learning model to determine the ventilation probability and ventilation time window of the target pressure signal, and the ventilation equipment is controlled to automatically ventilate the target person according to the ventilation time window.

[0011] Preferably, dynamically correcting the current airway pressure signal according to the body shape parameters of the target person to obtain the target airway pressure signal includes:

[0012] Detecting the compression force and compression speed when compressing the target person's heart;

[0013] Set a preset sampling frequency based on the pressing force and speed, obtain the area where the target person's mouth and nose are located and set it as the sampling area;

[0014] The air pressure of the target person is sampled in the sampling area at a preset sampling frequency through the air pressure sensor;

[0015] The current airway pressure signal when compressing the heart of the target person is obtained according to the sampling result.

[0016] Preferably, before continuously analyzing the current airway pressure signal to obtain the compression frequency and compression phase of the target person and the high-dimensional signal information, the method further includes:

[0017] confirming the current airway pressure signal as the initial signal;

[0018] Using a low-frequency filter to remove high-frequency noise signals and other clutter signals in the initial signal to obtain the initial signal after the first processing;

[0019] Eliminating the baseline offset signal in the initial signal after the first processing by eliminating the DC signal to obtain the initial signal after the second processing;

[0020] A clean proximal pressure signal is generated based on the second processed initial signal.

[0021] Preferably, dynamically correcting the current airway pressure signal according to the body shape parameters of the target person to obtain the target airway pressure signal includes:

[0022] Determining the current chest cavity resonance parameters of the target person according to the body shape parameters of the target person;

[0023] Determining a dynamic correction factor based on current chest cavity resonance parameters and preset acquisition parameters of current airway pressure signals;

[0024] Dynamically correct the current airway pressure signal using a dynamic correction factor;

[0025] The target airway pressure signal is obtained according to the correction result.

[0026] Preferably, the current airway pressure signal is continuously analyzed to obtain the compression frequency and compression phase of the target person and high-dimensional signal information, including:

[0027] Analyzing the proximal pressure signal using a time domain analysis algorithm to obtain a first analysis result;

[0028] Analyzing the proximal pressure signal using a frequency domain analysis algorithm to obtain a second analysis result;

[0029] Determining the compression frequency, compression interval, and single compression duration for the target person based on the first analysis result, and determining the compression phase based on the compression interval and single compression duration;

[0030] The signal frequency domain features are obtained according to the second analysis result, and the signal frequency domain features are deeply analyzed to obtain the signal high-dimensional information.

[0031] Preferably, analyzing the proximal pressure signal using a frequency domain analysis algorithm to obtain a second analysis result includes:

[0032] The proximal pressure signal is subjected to time-frequency conversion using a window function and Fourier transform to obtain a spectrum signal;

[0033] Mark the peak of the spectrum signal, and extract the area around the peak, the area of ​​the low-frequency area, and the area of ​​the high-frequency area;

[0034] Obtaining regional signals corresponding to the peak surrounding area, the low-frequency area, and the high-frequency area;

[0035] A second analysis result is obtained by performing geometric operations on the regional signals corresponding to the peak peripheral area, the low-frequency area, and the high-frequency area using a frequency domain analysis algorithm.

[0036] Preferably, the training of the machine learning model by presetting compression parameters and ventilation rule parameters includes:

[0037] Generate compression pattern parameters based on preset compression parameters based on a preset machine learning algorithm;

[0038] generating a first training sample according to the compression rule parameters, and setting model learning parameters and model output parameters based on the ventilation rule parameters;

[0039] Based on the model learning parameters and the model output parameters, the machine learning model is trained using the first training sample.

[0040] Preferably, the step of inputting the compression frequency and compression phase of the target person, the high-dimensional signal information, and the mode setting parameters of the ventilation device into a machine learning model to determine the ventilation probability and ventilation time window of the target pressure signal, and controlling the ventilation device to automatically ventilate the target person according to the ventilation time window includes:

[0041] Obtain ventilation compression ratio, target frequency and ventilation threshold of different ventilation modes according to the mode setting parameters of the ventilation equipment;

[0042] Determine the target ventilation mode based on the compression rate for the target person and the ventilation compression ratio of different modes, target rate and ventilation threshold;

[0043] Obtain the ventilation-compression phase in the target ventilation mode, and match the ventilation-compression phase according to the compression phase of the target person using a dynamic threshold method;

[0044] Determine the ventilation probability of the target pressure signal based on the matching result, and determine the ventilation time window threshold based on the positional relationship between the ventilation-compression phase and the compression phase for the target person;

[0045] According to the ventilation time window, the ventilation equipment is controlled to automatically ventilate the target person according to the ventilation threshold.

[0046] Preferably, determining the dynamic correction factor according to the current chest cavity resonance parameter and the preset acquisition parameters of the current airway pressure signal includes:

[0047] Determining a set chest compression variation amplitude parameter based on preset acquisition parameters of the current airway pressure signal;

[0048] Determine the airway resistance index based on the chest compression variation parameter and the current chest resonance parameter;

[0049] Generate a respiratory signal sample set using a respiratory mechanics model based on the airway resistance index;

[0050] Determine the morphological parameter values ​​of each point on the airway of the target person according to the respiratory signal sample set;

[0051] Determining the force tensor range of the target person's airway through the morphological parameter values ​​of each point on the target person's airway;

[0052] Acquire characteristic parameters for maintaining steady-state airway performance of the target person, and acquire airway performance indicators of the target person based on the characteristic parameters;

[0053] Determine the force differentiation of the target person's airway based on the target person's airway performance index and the airway force tensor range;

[0054] Determine the standard data of chest compression force on the target person and the error range of the actual test data based on the force differentiation;

[0055] The dynamic correction factor of the current airway pressure signal is determined based on the standard data of the chest compression force of the target person, the error range of the actual detection data, and the detection pressure data corresponding to the detected current airway pressure signal.

[0056] A ventilator, the device comprising: a ventilator body, a ventilation line, an air flow and air pressure detection device, a ventilation mask and a pressure plate;

[0057] The pressure plate is used to perform manual or machine chest compressions on the target person;

[0058] The ventilation mask is used to cover the target person's mouth and nose area;

[0059] The airflow and air pressure detection equipment is used to detect the airway pressure signal of the target person's heart during compression;

[0060] The ventilator body is used to analyze the airway pressure signal, obtain the target person's cardiopulmonary resuscitation and compression information, determine the ventilation time window based on the target person's cardiopulmonary resuscitation and compression information, issue ventilation commands and perform ventilation operations;

[0061] The ventilation tube is used to pass the gas to be ventilated in the ventilator body into the mouth and nose of the target person.

[0062] Other features and advantages of the present invention will be described in the following description, and in part will become apparent from the description, or will be understood by practicing the present invention. The purpose and other advantages of the present invention can be realized and obtained by the structures particularly pointed out in the written description and the accompanying drawings.

[0063] The technical solution of the present invention is further described in detail below through the accompanying drawings and embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0064] The accompanying drawings are used to provide further understanding of the present invention and constitute a part of the specification. They are used to explain the present invention together with the embodiments of the present invention and do not constitute a limitation of the present invention.

[0065] FIG1 is a flowchart of a method for automatic airway pressure feedback ventilation for cardiopulmonary resuscitation provided by the present invention;

[0066] FIG2 is another workflow diagram of a cardiopulmonary resuscitation airway pressure automatic feedback ventilation method provided by the present invention;

[0067] FIG3 is another workflow diagram of a cardiopulmonary resuscitation airway pressure automatic feedback ventilation method provided by the present invention;

[0068] FIG4 is a schematic structural diagram of a ventilator provided by the present invention;

[0069] FIG5 is a screenshot of an embodiment generated by a method for automatic feedback ventilation of airway pressure for pulmonary resuscitation provided by the present invention;

[0070] Figure 6 is an airway pressure curve during cardiac compression;

[0071] FIG7 is a schematic diagram showing the principle of preprocessing a signal;

[0072] FIG8 is a schematic diagram showing the principle of performing frequency domain analysis on a signal;

[0073] FIG9 is a schematic diagram showing the principle of performing time domain analysis on a signal;

[0074] FIG10 is a schematic diagram showing the principle of using a dynamic threshold method to determine ventilation timing based on machine compression signals;

[0075] FIG11 is a schematic diagram showing the principle of using the dynamic threshold method to determine the ventilation timing based on manual compression signals. DETAILED DESCRIPTION

[0076] Exemplary embodiments will be described in detail herein, with examples illustrated in the accompanying drawings. In the following description, when referring to the drawings, identical numerals in different figures represent identical or similar elements, unless otherwise indicated. The embodiments described in the following exemplary embodiments are not intended to represent all possible embodiments consistent with the present disclosure. Rather, they are merely examples of apparatus and methods consistent with certain aspects of the present disclosure, as detailed in the appended claims.

[0077] Currently, cardiopulmonary resuscitation (CPR) is the most basic and important medical method for saving the lives of patients experiencing cardiac arrest. Chest compressions combined with forced ventilation are typically the first-aid measures required for these patients to maintain basic circulation, heart rate, and breathing. Current CPR typically includes two key components: chest compressions and ventilation (breathing). Heart compressions are typically performed manually or mechanically (with equipment). Manual compressions are used when unavailable, while mechanical compressions are used when available in hospitals. Clinical requirements for chest compressions generally adhere to: an effective chest compression frequency of 100 to 120 times per minute and a depth of 5 to 6 centimeters. Ventilation (breathing) is typically provided through manual breathing, balloon ventilation, and mechanical ventilation. Ventilation requirements are based on clinical requirements, including two ventilations between 30 compressions, two ventilations after 15 compressions, and one ventilation after every 10 compressions. Existing CPR compression and ventilation coordination is usually coordinated by medical staff, or the CPR compression equipment and ventilator are coordinated through a communication protocol. However, there are problems with the inability to automatically coordinate, and manual coordination methods are difficult to accurately and stably control for a long time, reducing practicality. To address the above issues, this embodiment discloses a CPR airway pressure automatic feedback ventilation method.

[0078] A cardiopulmonary resuscitation airway pressure automatic feedback ventilation method, as shown in FIG1 , comprises the following steps:

[0079] Step S101: acquiring a current airway pressure signal when compressing the heart of a target person according to a preset sampling frequency;

[0080] Step S102: dynamically correcting the current airway pressure signal according to the body shape parameters of the target person to obtain a target airway pressure signal;

[0081] Step S103: Continuously analyze the target airway pressure signal to obtain compression frequency and compression phase for the target person and high-dimensional signal information;

[0082] Step S104: training a machine learning model using preset compression parameters and ventilation rule parameters;

[0083] Step S105: Input the compression frequency and compression phase of the target person, the high-dimensional signal information, and the mode setting parameters of the ventilation equipment into the machine learning model to determine the ventilation probability and ventilation time window of the target pressure signal, and control the ventilation equipment to automatically ventilate the target person according to the ventilation time window.

[0084] In this embodiment, the preset sampling frequency is represented by the acquisition cycle frequency of the sampling signal;

[0085] In this embodiment, the body shape parameters are parameters describing the body shape of the target person, such as height, weight, and specific indicators;

[0086] In this embodiment, dynamic correction is represented by correcting the current airway pressure signal according to the influence factor of the target person's body shape on the force during compression;

[0087] In this embodiment, the continuous analysis is represented by performing a continuous analysis of the target airway pressure signal in the time domain and frequency;

[0088] In this embodiment, the compression frequency and compression phase of the target person and the high-dimensional signal information are represented as the frequency and amplitude of chest compression of the target person;

[0089] In this embodiment, the preset compression parameters and ventilation rule parameters are represented by the corresponding rule parameters between the preset number of compressions and ventilation timing, for example: 30 compressions for one ventilation;

[0090] In this embodiment, the mode setting parameters of the ventilation device are represented as ventilation setting parameters of the ventilation device in each working mode;

[0091] In this embodiment, the ventilation probability and ventilation time window are expressed as a description time window of the probability of ventilation during each chest compression and the number of chest compressions required to confirm ventilation.

[0092] The working principle of the above technical solution is: obtaining the current airway pressure signal when compressing the heart of the target person according to the preset sampling frequency; dynamically correcting the current airway pressure signal according to the body parameters of the target person to obtain the target airway pressure signal; continuously analyzing the target airway pressure signal to obtain the compression frequency and compression phase of the target person and high-dimensional signal information; training the machine learning model through preset compression parameters and ventilation rule parameters; inputting the compression frequency and compression phase of the target person and the high-dimensional signal information and the mode setting parameters of the ventilation equipment into the machine learning model to determine the ventilation probability and ventilation time window of the target pressure signal, and controlling the ventilation equipment to automatically ventilate the target person according to the ventilation time window.

[0093] The beneficial effects of the above technical solution are: by automatically predicting the ventilation time window based on the compression frequency and compression phase of the target person and performing automatic ventilation, the device can automatically perform cardiac compression and timely ventilation on the target person without the need for a communication protocol, reducing labor costs while accurately controlling the timing of ventilation, improving safety, reliability and stability, avoiding the influence of subjective factors of medical staff, improving objectivity, and solving the problem that the existing cardiopulmonary resuscitation compression and ventilation coordination is usually coordinated by medical staff, or the cardiopulmonary resuscitation compression equipment and the ventilator are coordinated through a communication protocol, which cannot be automatically coordinated, and the manual coordination method is difficult to accurately and stably control for a long time, reducing practicality.

[0094] In one embodiment, as shown in FIG2 , dynamically correcting the current airway pressure signal according to the body shape parameters of the target person to obtain the target airway pressure signal includes:

[0095] Step S201: detecting the compression force and compression speed when compressing the heart of a target person;

[0096] Step S202: Set a preset sampling frequency according to the pressing force and pressing speed, obtain the area where the target person's mouth and nose are located and set it as the sampling area;

[0097] Step S203: sampling the air pressure of the target person in the sampling area at a preset sampling frequency using an air pressure sensor;

[0098] Step S204: Acquire a current airway pressure signal when compressing the heart of the target person according to the sampling result.

[0099] In this embodiment, the sampling area refers to the range of the current airway pressure signal for signal sampling.

[0100] The beneficial effects of the above technical solution are: by intelligently setting the sampling frequency, the sampling frequency can be reasonably set according to the real-time pressing situation of the target person, thereby ensuring the high quality of the sampling data. Furthermore, by locating the sampling area, the airflow model and air pressure signal detection of the mouth and nose of the target person can be performed more accurately, thereby ensuring the accuracy of the detection signal.

[0101] In one embodiment, as shown in FIG3 , before continuously analyzing the current airway pressure signal to obtain the compression frequency and compression phase of the target person and high-dimensional signal information, the process further includes:

[0102] Step S301: confirming the current airway pressure signal as an initial signal;

[0103] Step S302: Use a low-frequency filter to remove high-frequency noise signals and other clutter signals in the initial signal to obtain the initial signal after the first processing;

[0104] Step S303: Eliminating the baseline offset signal in the initial signal after the first processing by eliminating the DC signal to obtain the initial signal after the second processing;

[0105] Step S304: Generate a clean proximal pressure signal according to the initial signal after the second processing.

[0106] In this embodiment, the baseline shift signal is represented by a target signal in which the baseline of the signal shifts in the initial signal after the first processing.

[0107] The beneficial effects of the above technical solution are: it can ensure the purity of the signal and remove the influence of interference signals, lay the foundation for subsequent signal analysis, and improve practicality.

[0108] In one embodiment, dynamically correcting the current airway pressure signal according to the body shape parameters of the target person to obtain the target airway pressure signal includes:

[0109] Determining the current chest cavity resonance parameters of the target person according to the body shape parameters of the target person;

[0110] Determining a dynamic correction factor based on current chest cavity resonance parameters and preset acquisition parameters of current airway pressure signals;

[0111] Dynamically correct the current airway pressure signal using a dynamic correction factor;

[0112] The target airway pressure signal is obtained according to the correction result.

[0113] In this embodiment, the current chest cavity resonance parameter is represented as a parameter describing the resonance between the chest cavity and the compression operation caused by the target person's body shape;

[0114] In this embodiment, the preset acquisition parameters are set parameters such as the acquisition frequency and acquisition intensity of the current airway pressure signal;

[0115] In this embodiment, the dynamic correction factor is expressed as a correction factor for the signal phase and signal frequency of the current airway pressure signal.

[0116] The beneficial effect of the above technical solution is: by determining the current chest cavity resonance parameters based on the body parameters of the target person and then correcting the collected airway pressure signal, it can be ensured that the corrected airway pressure signal is more consistent with the target person, thereby improving the data reference.

[0117] In this embodiment, the dynamic correction factor is determined according to the current chest cavity resonance parameter and the preset acquisition parameters of the current airway pressure signal, including:

[0118] Determining a set chest compression variation amplitude parameter based on preset acquisition parameters of the current airway pressure signal;

[0119] Determine the airway resistance index based on the chest compression variation parameter and the current chest resonance parameter;

[0120] Generate a respiratory signal sample set using a respiratory mechanics model based on the airway resistance index;

[0121] Determine the morphological parameter values ​​of each point on the airway of the target person according to the respiratory signal sample set;

[0122] Determining the force tensor range of the target person's airway through the morphological parameter values ​​of each point on the target person's airway;

[0123] Acquire characteristic parameters for maintaining steady-state airway performance of the target person, and acquire airway performance indicators of the target person based on the characteristic parameters;

[0124] Determine the force differentiation of the target person's airway based on the target person's airway performance index and the airway force tensor range;

[0125] Determine the standard data of chest compression force on the target person and the error range of the actual test data based on the force differentiation;

[0126] The dynamic correction factor of the current airway pressure signal is determined based on the standard data of the chest compression force of the target person, the error range of the actual detection data, and the detection pressure data corresponding to the detected current airway pressure signal.

[0127] In this embodiment, the set chest compression variation parameter is represented by a data parameter of the fluctuation variation of the target person's chest as the chest is compressed under set conditions;

[0128] In this embodiment, the airway resistance index is expressed as the ventilation resistance difficulty index of the target person's airway under the pressure of the body shape;

[0129] In this embodiment, the respiratory mechanics model is represented as a network model for performing respiratory mechanics simulation;

[0130] In this embodiment, the respiratory signal sample set is represented by a respiratory signal sample data set simulated by a respiratory mechanics model under an airway resistance index;

[0131] In this embodiment, the morphological parameter values ​​of each point at the target person's airway are expressed as parameter values ​​of the undulating morphology of each distributed sampling point at the target person's airway;

[0132] In this embodiment, the force tensor range is represented by the numerical range of the tensor of the target person's airway after being subjected to the compression force;

[0133] In this embodiment, the characteristic parameter for maintaining the steady-state performance of the target person's airway is represented by a relevant characteristic parameter for maintaining the steady-state performance of the target person's airway;

[0134] In this embodiment, the airway performance index represents an index of the airway maintaining good breathing performance;

[0135] In this embodiment, the force differentiation condition is represented by the differentiation and loss of feedback airflow in the target person's airway during chest compression.

[0136] The beneficial effect of the above technical solution is: by determining the airway obstruction parameters of the target person and then determining the force differentiation of the target person's airway during compression, the conversion relationship between the actual compression force and the force intensity can be accurately evaluated to determine the dynamic correction factor of the detection signal, thereby ensuring the accuracy and objectivity of the final factor.

[0137] In one embodiment, the current airway pressure signal is continuously analyzed to obtain compression frequency and compression phase for the target person and high-dimensional signal information, including:

[0138] Analyzing the proximal pressure signal using a time domain analysis algorithm to obtain a first analysis result;

[0139] Analyzing the proximal pressure signal using a frequency domain analysis algorithm to obtain a second analysis result;

[0140] Determining the compression frequency, compression interval, and single compression duration for the target person based on the first analysis result, and determining the compression phase based on the compression interval and single compression duration;

[0141] The signal frequency domain features are obtained according to the second analysis result, and the signal frequency domain features are deeply analyzed to obtain the signal high-dimensional information.

[0142] The beneficial effect of the above technical solution is that through double analysis, detailed high-dimensional information such as the pressing-related rule parameters and pressing intensity of the target person can be accurately analyzed, laying the foundation for subsequent model training and further improving practicality.

[0143] In one embodiment, analyzing the proximal pressure signal using a frequency domain analysis algorithm to obtain a second analysis result includes:

[0144] The proximal pressure signal is subjected to time-frequency conversion using a window function and Fourier transform to obtain a spectrum signal;

[0145] Mark the peak of the spectrum signal, and extract the area around the peak, the area of ​​the low-frequency area, and the area of ​​the high-frequency area;

[0146] Obtaining regional signals corresponding to the peak surrounding area, the low-frequency area, and the high-frequency area;

[0147] A second analysis result is obtained by performing geometric operations on the regional signals corresponding to the peak peripheral area, the low-frequency area, and the high-frequency area using a frequency domain analysis algorithm.

[0148] The beneficial effect of the above technical solution is that it can maximize the acquisition of frequency domain analysis results corresponding to the signal, ensuring the integrity and reliability of high-dimensional information.

[0149] In one embodiment, training the machine learning model by presetting compression parameters and ventilation rule parameters includes:

[0150] Generate compression pattern parameters based on preset compression parameters based on a preset machine learning algorithm;

[0151] generating a first training sample according to the compression rule parameters, and setting model learning parameters and model output parameters based on the ventilation rule parameters;

[0152] Based on the model learning parameters and the model output parameters, the machine learning model is trained through the first training sample training.

[0153] The beneficial effects of the above technical solution are: by generating training samples and setting model learning parameters and output parameters at the same time, stable convergence can be achieved during the model training process, ensuring the subsequent recognition stability of the model, and improving practicality and reliability.

[0154] In one embodiment, inputting compression frequency and compression phase of the target person, high-dimensional signal information, and mode setting parameters of the ventilation device into a machine learning model to determine the ventilation probability and ventilation time window of the target pressure signal, and controlling the ventilation device to automatically ventilate the target person according to the ventilation time window includes:

[0155] Obtain ventilation compression ratio, target frequency and ventilation threshold of different ventilation modes according to the mode setting parameters of the ventilation equipment;

[0156] Determine the target ventilation mode based on the compression rate for the target person and the ventilation compression ratio of different modes, target rate and ventilation threshold;

[0157] Obtain the ventilation-compression phase in the target ventilation mode, and match the ventilation-compression phase according to the compression phase of the target person using a dynamic threshold method;

[0158] Determine the ventilation probability of the target pressure signal based on the matching result, and determine the ventilation time window threshold based on the positional relationship between the ventilation-compression phase and the compression phase for the target person;

[0159] According to the ventilation time window, the ventilation equipment is controlled to automatically ventilate the target person according to the ventilation threshold.

[0160] The beneficial effects of the above technical solution are: it can grasp the ventilation timing and ventilation volume for the target person, further ensuring the safety of the target person and the probability of successful rescue.

[0161] In one embodiment, this embodiment further discloses a ventilator suitable for the above-mentioned cardiopulmonary resuscitation airway pressure automatic feedback ventilation method, the device comprising: a ventilator body 1, a ventilation circuit 2, an airflow and air pressure detection device 3, a ventilation mask 4 and a pressure plate 5;

[0162] The pressure plate is used to perform manual or machine chest compressions on the target person;

[0163] The ventilation mask is used to cover the target person's mouth and nose area;

[0164] The airflow and air pressure detection equipment is used to detect the airway pressure signal of the target person's heart during compression;

[0165] The ventilator body is used to analyze the airway pressure signal, obtain the target person's cardiopulmonary resuscitation and compression information, determine the ventilation time window based on the target person's cardiopulmonary resuscitation and compression information, issue ventilation commands and perform ventilation operations;

[0166] The ventilation tube is used to pass the gas to be ventilated in the ventilator body into the mouth and nose of the target person.

[0167] The working principle and beneficial effects of the above technical solution have been described in the method claims and will not be repeated here.

[0168] In one embodiment, as shown in FIG5 , the compression pressure and airflow data are analyzed to obtain an airway pressure curve during cardiac compression, as shown in FIG6 . First, the signal is preprocessed, as shown in FIG7 , using a low-pass filter to remove high-frequency noise (such as equipment vibration, etc.) and clutter caused by other factors (such as electromagnetic interference, etc.) contained in the original signal. Then, the baseline offset signal (such as the PEEP setting of the respiratory device, etc.) is eliminated by eliminating the DC signal to obtain a clean proximal pressure signal. The signal is then analyzed in the frequency domain and time domain, as shown in FIG8 and FIG9 . The obtained time domain features and frequency domain features are combined with the mode settings of the ventilation device and submitted to the machine learning model for judgment to obtain an instruction on whether ventilation is needed to control the ventilation device for ventilation. The ventilation timing of machine compression and manual compression is determined by the dynamic threshold method, as shown in FIG10 and FIG11 .

[0169] Other embodiments of the present disclosure will readily occur to those skilled in the art after considering the specification and practicing the disclosure herein. This application is intended to cover any variations, uses, or adaptations of the present disclosure that follow from the general principles of the present disclosure and include common knowledge or customary techniques in the art not disclosed herein. The description and examples are to be considered as exemplary only, with the true scope and spirit of the present disclosure being indicated by the following claims.

[0170] It should be understood that the present disclosure is not limited to the exact structures that have been described above and shown in the drawings, and that various modifications and changes can be made without departing from the scope thereof. The scope of the present disclosure is limited only by the appended claims.

Claims

1. A cardiopulmonary resuscitation airway pressure automatic feedback ventilation method, characterized in that: The following steps are involved: Acquire a current airway pressure signal when compressing the heart of a target person according to a preset sampling frequency; Dynamically correct the current airway pressure signal according to the body shape parameters of the target person to obtain the target airway pressure signal; Continuously analyze the target airway pressure signal to obtain the compression frequency and compression phase of the target person and high-dimensional signal information; Train the machine learning model by presetting compression parameters and ventilation rule parameters; The compression frequency and phase of the target person, as well as the high-dimensional information of the signal and the mode setting parameters of the ventilation equipment are input into the machine learning model to determine the ventilation probability and ventilation time window of the target pressure signal, and the ventilation equipment is controlled to automatically ventilate the target person according to the ventilation time window.

2. The cardiopulmonary resuscitation airway pressure automatic feedback ventilation method according to claim 1, characterized in that: The dynamically correcting the current airway pressure signal according to the body shape parameters of the target person to obtain the target airway pressure signal includes: Detecting the compression force and compression speed when compressing the heart of the target person; Set a preset sampling frequency according to the pressing force and pressing speed, obtain the area where the target person's mouth and nose are located and set it as the sampling area; The air pressure of the target person is sampled in the sampling area by using an air pressure sensor at a preset sampling frequency; The current airway pressure signal when compressing the heart of the target person is obtained according to the sampling result.

3. The cardiopulmonary resuscitation airway pressure automatic feedback ventilation method according to claim 1, characterized in that: Before continuously analyzing the current airway pressure signal to obtain the compression frequency and compression phase of the target person and the high-dimensional information of the signal, it also includes: confirming the current airway pressure signal as an initial signal; Using a low-frequency filter to remove high-frequency noise signals and other clutter signals in the initial signal to obtain the initial signal after the first processing; Eliminating the baseline offset signal in the initial signal after the first processing by eliminating the DC signal, and obtaining the initial signal after the second processing; A clean proximal pressure signal is generated based on the second processed initial signal.

4. The cardiopulmonary resuscitation airway pressure automatic feedback ventilation method according to claim 1, characterized in that: The dynamically correcting the current airway pressure signal according to the body shape parameters of the target person to obtain the target airway pressure signal includes: Determine the current chest resonance parameters of the target person according to the body shape parameters of the target person; Determining a dynamic correction factor according to a current chest cavity resonance parameter and a preset acquisition parameter of a current airway pressure signal; Dynamically correct the current airway pressure signal using a dynamic correction factor; The target airway pressure signal is obtained according to the correction result.

5. The cardiopulmonary resuscitation airway pressure automatic feedback ventilation method according to claim 3, characterized in that: Continuously analyze the current airway pressure signal to obtain the compression frequency and compression phase of the target person and high-dimensional signal information, including: Analyzing the proximal pressure signal using a time domain analysis algorithm to obtain a first analysis result; Analyzing the proximal pressure signal by using a frequency domain analysis algorithm to obtain a second analysis result; Determine the compression frequency, compression interval, and single compression duration for the target person according to the first analysis result, and determine the compression phase according to the compression interval and single compression duration; The signal frequency domain features are obtained according to the second analysis result, and the signal frequency domain features are deeply analyzed to obtain the signal high-dimensional information.

6. The cardiopulmonary resuscitation airway pressure automatic feedback ventilation method according to claim 5, characterized in that: The analyzing the proximal pressure signal by using a frequency domain analysis algorithm to obtain a second analysis result includes: The proximal pressure signal is subjected to time-frequency conversion using a window function and Fourier transform to obtain a spectrum signal; Mark the peak of the spectrum signal, and capture the area around the peak, the area of ​​the low-frequency area, and the area of ​​the high-frequency area; Obtaining regional signals corresponding to the peak peripheral area, the low-frequency area, and the high-frequency area; A frequency domain analysis algorithm is used to perform geometric operations on the regional signals corresponding to the peak peripheral area, the low-frequency area, and the high-frequency area to obtain a second analysis result.

7. The cardiopulmonary resuscitation airway pressure automatic feedback ventilation method according to claim 1, characterized in that: The method of training the machine learning model by presetting compression parameters and ventilation rule parameters includes: Generate compression pattern parameters based on preset compression parameters based on a preset machine learning algorithm; Generate a first training sample according to the compression rule parameters, and set a model learning parameter and a model output parameter based on the ventilation rule parameters; Based on the model learning parameters and the model output parameters, the machine learning model is trained through the first training sample.

8. The cardiopulmonary resuscitation airway pressure automatic feedback ventilation method according to claim 1, characterized in that: The method inputs the compression frequency and compression phase of the target person, the signal high-dimensional information, and the mode setting parameters of the ventilation device into the machine learning model to determine the ventilation probability and the ventilation time window of the target pressure signal, and controls the ventilation device to automatically ventilate the target person according to the ventilation time window, including: According to the mode setting parameters of the ventilation equipment, the ventilation compression ratio, target frequency and ventilation threshold of different ventilation modes are obtained; Determine the target ventilation mode according to the compression frequency of the target person and the ventilation compression ratio of different modes, the target frequency and the ventilation threshold; Obtain the ventilation-compression phase under the target ventilation mode, and match the ventilation-compression phase according to the compression phase of the target person using a dynamic threshold method; Determine the ventilation probability of the target pressure signal according to the matching result, and determine the ventilation time window threshold according to the positional relationship between the ventilation compression phase and the compression phase for the target person; According to the ventilation time window, the ventilation equipment is controlled to automatically ventilate the target person according to the ventilation threshold.

9. The cardiopulmonary resuscitation airway pressure automatic feedback ventilation method according to claim 4, characterized in that: The dynamic correction factor is determined according to the current chest cavity resonance parameter and the preset acquisition parameters of the current airway pressure signal, including: Determine the set chest compression variation amplitude parameter according to the preset acquisition parameter of the current airway pressure signal; Determine the airway resistance index according to the chest compression variation parameter and the current chest resonance parameter; Generate a respiratory signal sample set using a respiratory mechanics model based on an airway resistance index; Determine the morphological parameter values ​​of each point at the airway of the target person according to the respiratory signal sample set; Determine the force tensor range of the target person's airway through the morphological parameter values ​​of each point on the target person's airway; Acquire characteristic parameters for maintaining steady-state performance of the target person's airway, and acquire airway performance indicators of the target person according to the characteristic parameters; Determine the force differentiation of the target person's airway based on the target person's airway performance index and the airway force tensor range; Determine the standard data of chest compression force on the target person and the error range of actual test data based on the force differentiation; The dynamic correction factor of the current airway pressure signal is determined according to the standard data of the chest compression force of the target person, the error range of the actual detection data and the detection pressure data corresponding to the detected current airway pressure signal.

10. A ventilator, suitable for the cardiopulmonary resuscitation airway pressure automatic feedback ventilation method according to claim 1, characterized in that: The device includes: a ventilator body, a ventilation circuit, an air flow and air pressure detection device, a ventilation mask and a pressure plate; The pressure plate is used to perform manual or machine chest compressions on the target person; The ventilation mask is used to cover the target person's mouth and nose area; The airflow and air pressure detection equipment is used to detect the airway pressure signal of the target person's heart when compression is performed; The main body of the ventilator is used to analyze the airway pressure signal, obtain the information of the target person's cardiopulmonary resuscitation and compression, and obtain the ventilation time window based on the information of the target person's cardiopulmonary resuscitation and compression, issue ventilation instructions and perform ventilation operations; The ventilation circuit is used to pass the gas to be ventilated in the ventilator body into the mouth and nose of the target person.

Images