A method for setting initial parameters of a ventilator

Abstract

The present invention relates to a method for setting initial parameters of a ventilator, and to the technical field of intelligent medical equipment. In the method, the patient type and personal health medical data of a patient object are collected, and the personal health medical data is input into a cluster classification model corresponding to the patient type to obtain the cluster category to which the patient object belongs. Based on the initial parameter data of each ventilator of a historically treated patient in the cluster category to which the patient object belongs, the initial parameters of the ventilator of the patient object are set. Among them, the cluster classification model is obtained by clustering learning of the personal health medical data of historically treated patients under the patient type. In this way, through historical data, the initial parameters of the ventilator of the historically treated patient closest to the personal health condition of the patient object are obtained, and the initial parameters most suitable for the patient object are obtained for setting, while reducing the reliance on the experience of medical personnel, and the initial parameters with higher accuracy can be obtained based on the historical data, thereby improving the medical effect.

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent medical equipment, and in particular to a method for setting initial parameters of a ventilator. Background Art

[0002] In the modern medical field, ventilators are very important medical devices that assist or replace patients' breathing through mechanical ventilation. They are divided into two categories: non-invasive ventilators and invasive ventilators. Clinicians often choose different types of ventilators according to the type of respiratory failure and the severity of the disease. However, whether it is a non-invasive or invasive ventilator, the use of the machine requires precise initial parameter settings, which requires medical staff to have rich experience and expertise. For patients with acute coronary syndrome (ACS), because their condition changes rapidly and is complicated, most patients are prone to respiratory failure, so the initial parameter settings of the ventilator are particularly important.

[0003] At present, the initial parameters of the ventilator are usually set by the experience and judgment of medical staff. This method relies on the professional level and experience of medical staff. Although it can achieve a certain treatment effect, the accuracy of parameter setting is low, which has a great impact on the medical effect. Summary of the invention

[0004] The technical problem to be solved by the present application is to provide a method for setting the initial parameters of a ventilator, which has the characteristics of being able to more accurately set the initial parameters of the ventilator and improve the medical effect.

[0005] A method for setting initial parameters of a ventilator provided in one embodiment of the present application includes:

[0006] Collect patient type and personal health and medical data of patient subjects;

[0007] Inputting the personal health care data of the patient object into the cluster classification model corresponding to the patient type to obtain the cluster category to which the patient object belongs; the cluster classification model is obtained based on cluster learning of the personal health care data of historically treated patients under the patient type;

[0008] The initial ventilator parameters of the patient object are set based on the initial ventilator parameter data of each historically treated patient in the cluster category where the patient object belongs.

[0009] In one embodiment, the setting of the initial ventilator parameters of the patient object based on the initial ventilator parameter data of each patient historically treated in the cluster category to which the patient object belongs includes:

[0010] Obtaining each ventilator initial parameter data of each historically treated patient under the cluster category to which the patient object belongs; or calculating the similarity between the patient object and the historically treated patient under the cluster category to which the patient object belongs, and obtaining each ventilator initial parameter data of the historically treated patient under a set quantity threshold with the highest similarity;

[0011] The initial ventilator parameters of the patient object are set based on the respective ventilator initial parameter data.

[0012] In one embodiment, the calculating of the similarity between the patient object and the historically treated patients in the corresponding cluster category includes:

[0013] Calculate the average of the personal health care data of the patient object as a first average;

[0014] Calculate the average value of personal health care data of each historically treated patient under the cluster category to which the patient object belongs as the second average value;

[0015] Calculate the covariance of the first mean with each second mean;

[0016] Compute covariance computes the standard deviation of the first mean and each second mean;

[0017] Based on the covariance and standard deviation, the correlation coefficient between the patient object and each historically treated patient under the cluster category to which it belongs is calculated.

[0018] In one embodiment, the step of setting the initial parameters of the ventilator for the patient subject based on the initial parameter data of each ventilator includes:

[0019] Score the initial parameter data of each ventilator;

[0020] Based on the scoring result, initial parameters of the ventilator of the patient subject are set.

[0021] In one embodiment, the scoring of each ventilator initial parameter data includes:

[0022] For any ventilator initial parameter data, the patient's life status corresponding to the initial parameters under the preset first time threshold is judged. If the life status is death, a first score is given; if the life status is non-death, the next step is entered;

[0023] Determine whether the ventilator parameters are adjusted within a preset second time threshold, and if so, give a second score, and if not, proceed to the next step; wherein the second score is greater than the first score;

[0024] Determine whether the ventilator parameters are adjusted within a preset third time threshold, and if so, give a third score, and if not, proceed to the next step; wherein the third time threshold is greater than the second time threshold, and the third score is greater than the second score;

[0025] Determine whether the ventilator parameters are adjusted within a preset fourth time threshold. If yes, give a fourth score, if not, give a fifth score; wherein the fourth time threshold is greater than the third time threshold, the fourth score is greater than the third score, and the fifth score is greater than the fourth score.

[0026] In one embodiment, the setting of initial ventilator parameters for the patient subject based on the scoring result includes:

[0027] Based on the scores obtained, the ventilator initial parameter data with the highest score among the scores is set as the ventilator initial parameter of the patient object.

[0028] In one embodiment, a cluster analysis model is obtained based on cluster learning of personal health care data of historically treated patients under the patient type, including:

[0029] The personal health and medical data of each historically treated patient are formed into a first data set;

[0030] A clustering algorithm is applied to the first data set to obtain a clustering classification model through training and learning.

[0031] In one embodiment, the step of applying a clustering algorithm to the first data set to train and learn to obtain a clustering classification model includes:

[0032] Each data point in the first data set is used as an initial clustering center. The first data set can be expressed as:

[0033] ,

[0034] in, X represents the first data set, j Represents the index of data items in personal health and medical data, J Represents the total number of data items in personal health and medical data, 1≤ j ≤ J ; i represents the index of each historically treated patient, I Represents the total number of patients treated in each history, 1≤ i ≤ I ;

[0035] Based on each initial cluster center, iterative clustering is performed until the convergence condition is met to obtain a cluster classification model.

[0036] In one embodiment, the iterative clustering is performed based on each initial cluster center until a convergence condition is met to obtain a cluster classification model, including:

[0037] Based on any initial cluster center, a new cluster center is calculated based on the preset range after each iterative clustering;

[0038] It is determined whether the new cluster center meets the convergence conditions. If so, the data points that meet the preset conditions are classified into the corresponding clusters based on the final cluster center to obtain a cluster classification model.

[0039] In one embodiment, the calculation of a new cluster center based on any initial cluster center and a preset range after each iterative clustering includes:

[0040] For any cluster center after iterative clustering , based on the radius h The mean vector of all data points in the range is used as the new cluster center. ; Where t represents the index of the number of iterations;

[0041] The method of classifying the data points that meet the preset conditions into corresponding clusters based on the final cluster center includes:

[0042] For any data point, calculate the distance from the data point to each converged cluster center;

[0043] Assign each data point to the cluster to which the cluster center with the smallest distance belongs.

[0044] The beneficial effects of the present invention are:

[0045] A comprehensive analysis can be conducted based on the specific conditions of each patient, such as the type of disease, oxygen saturation, the presence or absence of underlying diseases, and the historical parameter settings of the ventilator, and finally personalized initial ventilator parameter settings can be achieved. This personalized setting can better meet the actual needs of patients and improve the treatment effect. In addition, through scientific and reasonable initial ventilator parameter settings, adverse consequences caused by improper parameter settings, such as ventilator-associated pneumonia, barotrauma, circulatory failure, etc., can be reduced, thereby reducing the patient's treatment risk. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] Figure 1 It is a schematic diagram of a cluster learning acquisition method of a cluster analysis model according to an embodiment of the present application;

[0047] Figure 2 This application Figure 1 A method flow diagram of an embodiment of step S200;

[0048] Figure 3 It is a schematic flow chart of a method for classifying data points that meet preset conditions into corresponding clusters based on final cluster centers according to an embodiment of the present application;

[0049] Figure 4 This is a flow chart of a method for setting initial parameters of a ventilator according to an embodiment of the present application;

[0050] Figure 5 This application Figure 4 A method flow diagram of an embodiment of step S30;

[0051] Figure 6 This application Figure 5 A method flow diagram of an embodiment of step S3012;

[0052] Figure 7 This application Figure 4 A schematic diagram of a method flow chart of another embodiment of step S30

[0053] Figure 8 The flowchart of a method for calculating the similarity between a patient object and a historically treated patient in a corresponding cluster category in an embodiment is shown. DETAILED DESCRIPTION

[0054] The present invention is further described in detail below by specific embodiments in conjunction with the accompanying drawings. Wherein similar elements in different embodiments adopt associated similar element numbers. In the following embodiments, many detailed descriptions are for making the present application better understood. However, those skilled in the art can easily recognize that some features can be omitted in different situations, or can be replaced by other elements, materials, methods. In some cases, some operations related to the present application are not shown or described in the specification, this is to avoid the core part of the present application being overwhelmed by too much description, and for those skilled in the art, it is not necessary to describe these related operations in detail, and they can fully understand the related operations according to the description in the specification and the general technical knowledge in the art.

[0055] In addition, the features, operations or characteristics described in the specification can be combined in any appropriate manner to form various implementations. At the same time, the steps or actions in the method description can also be interchanged or adjusted in a manner that is obvious to those skilled in the art. Therefore, the various sequences in the specification and the drawings are only for the purpose of clearly describing a certain embodiment and are not meant to be a required sequence, unless otherwise specified that a certain sequence must be followed.

[0056] The serial numbers assigned to the components in this article, such as "first", "second", etc., are only used to distinguish the objects described and do not have any order or technical meaning.

[0057] To facilitate the description of the inventive concept of the present application, the ventilator initial parameter setting technology is briefly described below.

[0058] The current initial parameter settings of ventilators rely on the experience of medical staff and are subject to uncertainty. However, the applicant found in the study that the initial parameter settings will be different for different patient types. Even for the same patient type, the initial parameter settings will be different for different patients with different physical conditions. It is difficult for medical staff to accurately determine the initial parameters that should be set. Moreover, the patient's condition changes rapidly, and medical staff may not have enough time to fully assess the patient's condition and set appropriate parameters. These are the technical difficulties in the initial parameter settings of ventilators.

[0059] In view of this, an embodiment of the present application provides a method for setting initial parameters of a ventilator. Based on this method, personal health and medical data of historically treated patients of the same type of patients are obtained, and the personal health and medical data of the patient object is added thereto for cluster analysis, thereby obtaining the cluster to which the patient object belongs, and then the initial parameters of the ventilator of the patient object are set based on the initial parameter data of the ventilator of the historically treated patients in the cluster to which it belongs. In this way, the initial parameters of the ventilator of the historically treated patients closest to the personal health conditions of the patient object are obtained through historical data, and the initial parameters that are most suitable for the patient object are obtained for setting. On the one hand, the reliance on the experience of medical personnel is reduced, and on the other hand, the initial parameters with higher accuracy can be obtained based on historical data, thereby further improving the medical effect.

[0060] In order to more clearly explain the method for setting the initial parameters of the ventilator, the clustering learning acquisition method based on the clustering analysis model is first explained. In one embodiment, please refer to Figure 1 ,include:

[0061] Step S100, personal health and medical data of each historically treated patient are combined into a first data set.

[0062] It should be noted that the historically treated patients here are historically treated patients under the same patient type. The following explanation is given by taking the patient type of acute coronary syndrome (ACS) as an example.

[0063] In one embodiment, personal health and medical data may include at least some of basic information, physical sign data, disease type, oxygen saturation, and whether there are underlying diseases such as chronic obstructive pulmonary disease (COPD), adult respiratory distress syndrome (ARDS), and obesity.

[0064] In one embodiment, the first data set can be expressed as:

[0065] ,

[0066] in, X represents the first data set, j Represents the index of data items in personal health and medical data, J Represents the total number of data items in personal health and medical data, 1≤ j ≤ J ; i represents the index of each historically treated patient, I Represents the total number of patients treated in each history, 1≤ i ≤ I .

[0067] In one embodiment, if I =1000, J =7, there are 7000 data points in the first data set.

[0068] Step S200: Apply a clustering algorithm to the first data set to obtain a clustering classification model through training and learning.

[0069] In one implementation, please refer to Figure 2 , step S200 may include:

[0070] Step S2001: taking each data point in the first data set as an initial cluster center.

[0071] In one embodiment, for any data point , whose initial cluster centers It can be expressed as . The superscript 0 indicates the number of iterations is 0.

[0072] Step S2002: Based on each initial cluster center, iterative clustering is performed until a convergence condition is met to obtain a cluster classification model.

[0073] In one embodiment, any initial cluster center For example, for the next iterative clustering, we can calculate The mean vector of all data points within the high-dimensional sphere with a radius of h as the center is taken as the new iterative clustering center. .

[0074] It can be seen that step S2002 may include:

[0075] First, based on any initial cluster center, a new cluster center is calculated based on the preset range after each iterative clustering.

[0076] Then for any cluster center after iterative clustering , based on the radius h The mean vector of all data points in the range is used as the new cluster center. ;in, t An index indicating the iteration number.

[0077] In one embodiment, for any cluster center after iterative clustering , get the new cluster center The specific methods may include: defining a kernel function , the kernel function can use Gaussian kernel function, then:

[0078] ,

[0079] in, d Represents the dimension of the data, represents the Euclidean norm, e Represents a natural number exponent.

[0080] The measurement radius is h Each data point in the range Cluster Center The contribution degree of each data point is obtained based on the contribution degree. , then the obtained weight vector It can be used to measure the contribution of data points to the cluster center. In one embodiment, the weight vector It can be expressed as:

[0081] .

[0082] The new cluster center It can be expressed as:

[0083] .

[0084] Secondly, it is determined whether the new cluster center meets the convergence conditions. If so, the data points that meet the preset conditions are classified into the corresponding clusters based on the final cluster center.

[0085] In one embodiment, determining whether the new cluster center meets the convergence condition includes: for any updated cluster center , calculate the distance from the cluster center before updating as the first distance; determine whether the first distance is less than a preset first distance threshold, if so, it is considered that the convergence condition is met.

[0086] In one embodiment, the Euclidean distance between the updated cluster center and the cluster center before the update is calculated as the first distance. If the first distance obtained for all cluster centers is less than the preset first distance threshold, it is considered that the convergence condition is met and the iteration is stopped, otherwise, the next iteration is continued.

[0087] In one embodiment, please refer to Figure 3 , based on the final cluster center, the data points that meet the preset conditions are classified into the corresponding clusters, which may include:

[0088] Step S1000: for any data point, calculate the distance from the data point to each converged cluster center.

[0089] Step S1000 calculates the distance from each data point to each converged cluster center as the second distance. For example, if there are 6 cluster centers, each data point has 6 corresponding second distances. In one embodiment, the second distance uses the Euclidean distance.

[0090] Step S2000, assigning each data point to the cluster to which the cluster center with the smallest distance belongs.

[0091] In one embodiment, for any data point, since the minimum distance can be obtained from the six obtained second distances, the cluster center corresponding to the minimum distance can be used as the cluster center of the cluster to which it belongs. Thus, the cluster to which each data point belongs can be obtained, and a cluster classification model can be obtained.

[0092] The model of the embodiment of the present application is trained based on the historical data of patients with acute coronary syndrome (ACS) and the historical data of initial ventilator parameter settings that have been saved by the hospital, so it is highly scientific and reasonable. Compared with the practice of relying on the experience and judgment of medical staff in the prior art, various factors can be considered more systematically, the influence of human factors can be reduced, and the error rate of parameter settings can be reduced.

[0093] In one embodiment of the present application, a method for setting initial parameters of a ventilator is provided. Figure 4 ,include:

[0094] Step S10, collecting the patient type and personal health and medical data of the patient object.

[0095] In order to solve the problem that the initial parameter settings for different patient types may be different, when collecting the personal health care data of the patient object, the patient type of the patient object is first determined.

[0096] Step S20: inputting the personal health care data of the patient object into the cluster classification model corresponding to the patient type to obtain the cluster category to which the patient object belongs.

[0097] The cluster classification model is obtained based on cluster learning of personal health care data of historically treated patients under patient types, and the acquisition method can refer to the methods of step S100 to step S200.

[0098] Step S30 , setting the initial ventilator parameters of the patient object based on the initial ventilator parameter data of each patient historically treated in the cluster category to which the patient object belongs.

[0099] Among them, for any set of initial parameter data, it can include at least some of ventilator type selection, mode selection, oxygen concentration, positive end-expiratory pressure (PEEP), inspiratory time, respiratory rate, inspiratory pressure or tidal volume, inspiratory support pressure, etc.

[0100] In one embodiment, please refer to Figure 5 , step S30 comprises:

[0101] Step S3011, obtaining initial parameter data of each ventilator of each historically treated patient under the corresponding cluster category of the patient object.

[0102] Step S3012, setting the initial parameters of the ventilator of the patient object based on the initial parameter data of each ventilator.

[0103] In one embodiment, step S3012 may include: scoring each ventilator initial parameter data, and setting the ventilator initial parameters of the patient object based on the scoring results.

[0104] In one embodiment, please refer to Figure 6 , the method for scoring the initial parameter data of each ventilator includes:

[0105] Step S30121, for any ventilator initial parameter data, judge the patient's life status corresponding to the initial parameters under the preset first time threshold. If the life status is death, give a first score; if the life status is non-death, proceed to the next step.

[0106] In one embodiment, the first time threshold may be set to 2 hours, and the first score may be 0.

[0107] Step S30122, determining whether the ventilator parameters are adjusted within the preset second time threshold, if yes, giving a second score, if no, proceeding to the next step, wherein the second score is greater than the first score.

[0108] In one embodiment, the second time threshold may be set to 40 minutes, and the second score may be 1.

[0109] Step S30123, determine whether the ventilator parameters are adjusted within the preset third time threshold, if yes, give a third score, if no, proceed to the next step. The third time threshold is greater than the second time threshold, and the third score is greater than the second score.

[0110] In one embodiment, the third time threshold may be set to 3 hours, and the third score may be 2.

[0111] Step S30124, determining whether the ventilator parameters are adjusted within the preset fourth time threshold, if yes, giving a fourth score, if no, giving a fifth score. The fourth time threshold is greater than the third time threshold, the fourth score is greater than the third score, and the fifth score is greater than the fourth score.

[0112] In one embodiment, the fourth time threshold may be set to 24 hours, the fourth score may be 3, and the fifth score may be 4.

[0113] Based on the scores obtained, the ventilator initial parameter data with the highest score among the scores is set as the ventilator initial parameter of the patient object.

[0114] In some embodiments, the doctor can improve the ventilator initial parameter data with the highest score according to actual needs, and set the improved ventilator initial parameter data as the ventilator initial parameter of the patient object. The doctor can also improve the ventilator initial parameter data with the second highest score according to actual needs, and set the improved ventilator initial parameter data as the ventilator initial parameter of the patient object.

[0115] In order to further improve the accuracy of the initial parameter setting of the ventilator, in another embodiment, please refer to Figure 7 , step S30 comprises:

[0116] Step S3021, calculating the similarity between the patient object and the historically treated patients in the corresponding cluster category, and obtaining the initial parameter data of each ventilator of the historically treated patients under the set quantity threshold with the highest similarity.

[0117] In one embodiment, please refer to Figure 8 In step S3021, calculating the similarity between the patient object and the historically treated patients in the corresponding cluster category may include:

[0118] Step S30211, calculating the average value of the personal health care data of the patient object as a first average value.

[0119] In one embodiment, the first average value can be expressed as:

[0120] ,

[0121] in, A Represents the personal health and medical data of a patient object, Any data item in the personal health care data of a patient object j data, Indicates the first average value.

[0122] Step S30212, calculating the average value of personal health care data of each historically treated patient under the cluster category to which the patient object belongs as the second average value.

[0123] In one embodiment, the second average value can be expressed as:

[0124] ,

[0125] in, B m Represents any historically treated patient under the cluster category to which the patient object belongs m of personal health and medical data, Indicates any historical treatment patient in the cluster category to which the patient object belongs m The corresponding second average value is Indicates m Any data item in the personal health care data of patients with historical treatment j data.

[0126] Step S30213, calculating the covariance between the first mean value and each second mean value.

[0127] In one embodiment, the covariance can be expressed as:

[0128] ,

[0129] in, Represents a patient object and any historically treated patient m The covariance of personal health care data.

[0130] Step S30214, calculating the covariance between the first mean value and each second mean value.

[0131] In one embodiment, the standard deviation of the first mean value can be expressed as:

[0132] ,

[0133] in, represents the standard deviation of the first mean.

[0134] The standard deviation of any second mean can be expressed as:

[0135] ,

[0136] in, represents the standard deviation of any second mean.

[0137] Step S30215, based on the covariance and standard deviation, calculate the correlation coefficient between the patient object and each historically treated patient in the corresponding cluster category.

[0138] Based on the above embodiment, the correlation coefficient between the patient object and any historically treated patient in the cluster category to which it belongs can be expressed as:

[0139] ,

[0140] in, Represents a patient object and any historically treated patient m The correlation coefficient of .

[0141] The larger the correlation coefficient is, the greater the similarity between the two.

[0142] In one embodiment, the obtained correlation coefficients are arranged from large to small, so that the initial parameter data of each ventilator of the top three historically treated patients with the highest similarity can be obtained.

[0143] Step S3022, setting the initial parameters of the ventilator of the patient object based on the initial parameter data of each ventilator.

[0144] The detailed scheme of step S3022 can refer to the description of step S3012.

[0145] In one embodiment, each ventilator initial parameter data of the first three historically treated patients can be scored and judged. For example, based on the scores obtained in step S3012, the ventilator initial parameter data with the highest score among the scores can be set as the ventilator initial parameters for the patient object. The doctor can improve the ventilator initial parameter data with the highest score according to actual needs, and set the improved ventilator initial parameter data as the ventilator initial parameters for the patient object. The doctor can also improve the ventilator initial parameter data with the second highest score according to actual needs, and set the improved ventilator initial parameter data as the ventilator initial parameters for the patient object.

[0146] In some embodiments, the doctor may also select any one of the ventilator initial parameter data of the top three historically treated patients with the highest similarity as the ventilator initial parameter for the patient object based on actual needs, or may improve any one of the ventilator initial parameter data of the top three historically treated patients with the highest similarity based on actual needs, and use the improved ventilator initial parameter data as the ventilator initial parameter for the patient object.

[0147] Based on the scheme in the above embodiment, a comprehensive analysis can be performed according to the specific conditions of each patient, such as the type of disease, oxygen saturation, the presence or absence of underlying diseases, etc., as well as the historical parameter settings of the ventilator, and finally a personalized ventilator initial parameter setting can be achieved. This personalized setting can better meet the actual needs of the patient and improve the treatment effect. In addition, through scientific and reasonable initial parameter settings of the ventilator, adverse consequences caused by improper parameter settings, such as ventilator-associated pneumonia, barotrauma, circulatory failure, etc., can be reduced, thereby reducing the patient's treatment risk. The automated ventilator parameter setting greatly improves the efficiency of the setting, which can save precious treatment time for patients and improve treatment efficiency.

[0148] The method in the embodiments of the present application can be used in various medical institutions, especially in areas where there is a lack of experienced medical personnel. It can help medical personnel better set the initial parameters of the ventilator and improve the quality of treatment. Its application can greatly improve the quality and efficiency of medical services, and will also bring huge economic and social benefits. It has broad application prospects and great application value.

[0149] In one embodiment of the present application, a computer-readable storage medium is provided, on which a program is stored. The stored program includes a method that can be loaded by a processor and process any one of the above embodiments.

[0150] Those skilled in the art will appreciate that all or part of the functions of the various methods in the above-mentioned embodiments can be implemented by hardware or by computer programs. When all or part of the functions in the above-mentioned embodiments are implemented by computer programs, the program can be stored in a computer-readable storage medium, and the storage medium can include: read-only memory, random access memory, disk, optical disk, hard disk, etc., and the program is executed by a computer to implement the above-mentioned functions. For example, the program is stored in the memory of the device, and when the program in the memory is executed by the processor, all or part of the above-mentioned functions can be implemented. In addition, when all or part of the functions in the above-mentioned embodiments are implemented by computer programs, the program can also be stored in a storage medium such as a server, another computer, disk, optical disk, flash disk or mobile hard disk, and can be downloaded or copied and saved in the memory of the local device, or the system of the local device is updated, and when the program in the memory is executed by the processor, all or part of the functions in the above-mentioned embodiments can be implemented.

[0151] The above specific examples are used to illustrate the present invention, which is only used to help understand the present invention and is not intended to limit the present invention. For those skilled in the art, according to the concept of the present invention, some simple deductions, modifications or substitutions can be made.

Claims

1. A method for setting initial parameters of a ventilator, characterized in that: include: Collecting the patient type and personal health and medical data of the patient; the personal health and medical data includes any one of basic information, physical sign data, disease type, oxygen saturation, and whether there are other symptom reactions, and the other symptom reactions include any one or more of chronic obstructive pulmonary disease, adult respiratory distress syndrome, and obesity; Inputting the personal health care data of the patient object into the cluster classification model corresponding to the patient type to obtain the cluster category to which the patient object belongs; The cluster classification model is obtained based on cluster learning of personal health care data of historically treated patients under the patient type; Setting the initial ventilator parameters of the patient object based on the initial ventilator parameter data of each historically treated patient in the cluster category where the patient object belongs; The step of setting the initial ventilator parameters for the patient object based on the initial ventilator parameter data of each patient historically treated in the cluster category to which the patient object belongs includes: Obtaining each ventilator initial parameter data of each historically treated patient under the cluster category to which the patient object belongs; or calculating the similarity between the patient object and the historically treated patient under the cluster category to which the patient object belongs, and obtaining each ventilator initial parameter data of the historically treated patient under a set quantity threshold with the highest similarity; The initial ventilator parameters of the patient object are set based on the initial parameter data of each ventilator; wherein any set of initial ventilator parameters includes any number of ventilator type selection, mode selection, oxygen concentration, positive end-expiratory pressure, inspiratory time, respiratory rate, inspiratory pressure or tidal volume, and inspiratory support pressure; The step of setting the initial parameters of the ventilator for the patient object based on the initial parameter data of each ventilator includes: Score the initial parameter data of each ventilator; Based on the scoring result, setting initial parameters of the ventilator for the patient subject; The scoring of the initial parameter data of each ventilator includes: For any ventilator initial parameter data, the patient's life status corresponding to the initial parameters under the preset first time threshold is judged. If the life status is death, a first score is given; if the life status is non-death, the next step is entered; Determine whether the ventilator parameters are adjusted within a preset second time threshold, and if so, give a second score, and if not, proceed to the next step; wherein the second score is greater than the first score; Determine whether the ventilator parameters are adjusted within a preset third time threshold, and if so, give a third score, and if not, proceed to the next step; wherein the third time threshold is greater than the second time threshold, and the third score is greater than the second score; Determine whether the ventilator parameters are adjusted within a preset fourth time threshold. If yes, give a fourth score, if not, give a fifth score; wherein the fourth time threshold is greater than the third time threshold, the fourth score is greater than the third score, and the fifth score is greater than the fourth score.

2. The method for setting initial parameters of a ventilator according to claim 1, wherein: This includes calculating the similarity of the patient object to the historically treated patients in the cluster category to which it belongs, including: Calculate the average of the personal health care data of the patient object as a first average; Calculate the average value of personal health care data of each historically treated patient under the cluster category to which the patient object belongs as the second average value; Calculate the covariance of the first mean with each second mean; Compute covariance computes the standard deviation of the first mean and each second mean; Based on the covariance and standard deviation, the correlation coefficient between the patient object and each historically treated patient under the cluster category to which it belongs is calculated.

3. The method for setting the initial parameters of a ventilator as claimed in claim 2, characterized in that: The step of setting the initial parameters of the ventilator for the patient subject based on the scoring result includes: Based on the scores obtained, the ventilator initial parameter data with the highest score among the scores is set as the ventilator initial parameter of the patient object.

4. The method for setting initial parameters of a ventilator according to claim 1, wherein: A cluster analysis model is obtained based on cluster learning of personal health care data of historically treated patients under the patient type, including: The personal health and medical data of each historically treated patient are formed into a first data set; A clustering algorithm is applied to the first data set to obtain a clustering classification model through training and learning.

5. The method for setting initial parameters of a ventilator as claimed in claim 4, characterized in that: The step of applying a clustering algorithm to the first data set and training and learning to obtain a clustering classification model includes: Each data point in the first data set is used as an initial clustering center. The first data set can be expressed as: ， Wherein, X represents the first data set, j represents the index of the data item in the personal health and medical data, J represents the total number of data items in the personal health and medical data, 1≤j≤J; i represents the index of each historically treated patient, I represents the total number of each historically treated patient, 1≤i≤I; Based on each initial cluster center, iterative clustering is performed until the convergence condition is met to obtain a cluster classification model.

6. The method for setting initial parameters of a ventilator as claimed in claim 5, characterized in that: The iterative clustering is performed based on each initial cluster center until the convergence condition is met to obtain a cluster classification model, including: Based on any initial cluster center, a new cluster center is calculated based on the preset range after each iterative clustering; It is determined whether the new cluster center meets the convergence conditions. If so, the data points that meet the preset conditions are classified into the corresponding clusters based on the final cluster center to obtain a cluster classification model.

7. The method for setting initial parameters of a ventilator according to claim 6, wherein: The calculation of a new cluster center based on any initial cluster center and a preset range after each iterative clustering includes: For any cluster center after iterative clustering , based on the mean vector of all data points within a radius of h, and using the mean vector as the new cluster center ; Where t represents the index of the number of iterations; The method of classifying the data points that meet the preset conditions into corresponding clusters based on the final cluster center includes: For any data point, calculate the distance from the data point to each converged cluster center; Assign each data point to the cluster to which the cluster center with the smallest distance belongs.

Images