CN118356169B - Automatic monitoring system for medical care - Google Patents

Abstract

The invention relates to the technical field of data analysis, in particular to an automatic medical care monitoring system, which comprises: the data acquisition module is used for acquiring nursing data of all dimensions of a patient at each moment; the multidimensional analysis module is used for acquiring the similarity degree of the trend of each dimension and the direction vector of the main component according to the nursing data of all dimensions of the patient at each moment; the single-dimension module is used for acquiring the characteristic expression degree of each dimension according to the amplitude value and the acquisition time of the nursing data of each dimension and combining the similarity degree of the trend of each dimension and the direction vector of the main component; and the monitoring module is used for acquiring the core dimension according to the characteristic expression of each dimension, and judging the nursing data acquired in real time by utilizing the nursing data in the core dimension. According to the method, the dimension of the multidimensional nursing data is reduced, so that the dimension disaster is avoided, and whether abnormal data exist in the nursing data acquired in real time or not is accurately judged.

Description

A medical care automatic monitoring system

Technical Field

The present invention relates to the technical field of data analysis, and in particular to an automatic medical care monitoring system.

Background Art

Medical hospitals conduct clinical assessments of patients, develop treatment plans, and carry out necessary examinations and tests, combined with the collaborative work of doctors and nursing staff, to ensure that patients receive the best medical care, that is, it is necessary to monitor the nursing data collected in real time, and the traditional method of automatically monitoring the patient's nursing data is mainly to determine whether the patient's nursing data has abnormalities through preset thresholds; however, this method is limited by the value of the preset threshold, and due to the differences in patients' physical characteristics, the accuracy of monitoring whether the nursing data has abnormalities through threshold judgment is low; therefore, this application proposes a method based on the SOM neural network algorithm to determine whether the nursing data is abnormal, but because of the excessive dimensions of the nursing data in the process of training the SOM neural network, the high-dimensional data increases the distance between samples, making it difficult to capture the structure and similarity between the data, which causes a dimensionality disaster, resulting in the inability to accurately determine whether there are abnormal data in the nursing data collected in real time directly through the SOM neural network algorithm.

Summary of the invention

The present invention provides an automatic monitoring system for medical care to solve the existing problem that it is impossible to accurately judge whether there is abnormal data in the nursing data collected in real time by setting a threshold and directly using a SOM neural network algorithm.

A medical care automatic monitoring system of the present invention adopts the following technical solution:

Includes the following modules:

Data collection module, used to collect nursing data of all dimensions of patients at every moment;

The multi-dimensional analysis module is used to obtain several principal components of the nursing data of all dimensions, the direction vector of each principal component and the variance contribution rate of each principal component according to the nursing data of all dimensions of the patient at each moment; construct a sample space according to the nursing data of each dimension of the patient, and obtain the similarity between the trend of each dimension and the direction vector of the principal component according to the position of the data point in the sample space, combined with the direction vector of each principal component and the variance contribution rate of each principal component;

The single-dimension module is used to obtain the increment of each nursing data in each dimension according to the amplitude and the time of collection of the nursing data in each dimension, and classify each nursing data in each dimension to obtain the classification result of each nursing data in each dimension; according to the increment and classification result of each nursing data in each dimension, combined with the trend of each dimension and the similarity of the direction vector of the main component, the characteristic expression degree of each dimension is obtained;

The monitoring module is used to obtain the core dimension according to the characteristic expression of each dimension, train the neural network model with the nursing data in the core dimension, and monitor the nursing data collected in real time according to the neural network model.

Preferably, the method of obtaining several principal components of the nursing data of all dimensions, the direction vector of each principal component and the variance contribution rate of each principal component according to the nursing data of all dimensions of the patient at each moment includes the following specific methods:

The nursing data of all dimensions of the patient at each moment are input into the PCA principal component analysis algorithm to obtain several principal components of the nursing data of all dimensions, the direction vector of each principal component and the variance contribution rate of each principal component.

Preferably, the sample space is constructed according to the nursing data of each dimension of the patient, and according to the position of the data point in the sample space, the direction vector of each principal component and the variance contribution rate of each principal component are combined to obtain the similarity between the trend of each dimension and the direction vector of the principal component, including the specific method of:

Get the number of dimensions of the patient's nursing data in all dimensions at each moment, recorded as , construct a dimensional coordinate system, placing the nursing data of all dimensions of the patient at each moment into dimensional coordinate system, and obtain the sample space;

For Dimensions and principal component, and obtain the The direction vector of the principal component is dimensions, obtain the projection vector of the direction vector of each principal component on all dimensions; obtain the projection of all data points in the sample space on all dimensions, recorded as the projection point on all dimensions, for the The first dimension projection point, get the The first dimension The projection point, The direction vector of the principal component is The projection vector on the projection vector in the dimension is denoted as The first dimension The projection point projection vectors, get all the projection vectors of all the projection points in each dimension;

According to the direction vector of each principal component in The projection vector on the dimension, All projection vectors of all projection points in the dimension, the variance contribution rate of each principal component, and the The number of projection points on the dimension, get the The similarity between the trend of the dimension and the direction vector of the principal component.

Preferably, the obtaining The similarity between the trend of each dimension and the direction vector of the principal component includes the specific calculation formula:

in, Indicates The similarity between the trend of the dimension and the direction vector of the principal component, Indicates The number of projection points in dimensions, Indicates The variance contribution of the principal components is Indicates The direction vector of the principal component is The projection vector in dimensions, Indicates The first dimension The projection point projection vector; Represents the number of principal components.

Preferably, the method of obtaining the increment of each nursing data in each dimension according to the amplitude of the nursing data in each dimension and the time of collection, and classifying each nursing data in each dimension to obtain the classification result of each nursing data in each dimension includes the following specific methods:

For In the dimension The nursing data collected at each moment will be In the dimension The nursing data collected at the moment minus the In the dimension The difference in nursing data collected at the moment is recorded as In the dimension The increment of nursing data collected at each moment, and obtain the The increment of all nursing data in the dimension will be The incremental mean of all nursing data in the dimension is recorded as the benchmark value. The nursing data with increments less than the benchmark value in the dimension are recorded as the first category of data. Nursing data with increments greater than the benchmark value in each dimension are recorded as the second type of data.

Preferably, the method of obtaining the characteristic expression degree of each dimension according to the increment and classification result of each nursing data in each dimension and combining the trend of each dimension with the similarity of the direction vector of the principal component includes the following specific methods:

For dimensions, according to The increment of all nursing data in the dimension is obtained. The kurtosis of the increments of all nursing data in the dimension, combined with the The similarity between the trend of the dimension and the direction vector of the principal component, the number of the first type of data and the number of the second type of data, and obtain the The characteristic expression of the dimension.

Preferably, the obtaining The characteristic expression of each dimension includes the following specific calculation formula:

In the formula, Indicates The characteristic expression of each dimension; Indicates The similarity between the trend of each dimension and the direction vector of the principal component; Indicates The kurtosis of the increments of all nursing data in the dimension; Indicates the number of the first category of data; Indicates the number of the second type of data; represents the normalization function; Represents absolute value operation; Represents an exponential function with a natural constant as base.

Preferably, the obtaining of the core dimension according to the characteristic expression of each dimension includes the following specific methods:

Preset a performance threshold , for the The characteristic expression of the dimension The characteristic expression of the dimension is greater than , then Dimensions are core dimensions, and all core dimensions are obtained.

Preferably, the method of training the neural network model using the nursing data in the core dimension includes the following specific methods:

The nursing data in the core dimension is used as the input layer of the SOM neural network model, and the output layer is a A square topology network is constructed. training, and obtaining a SOM neural network model; and are the preset square side length and number of training times respectively.

Preferably, the monitoring of the nursing data collected in real time according to the neural network model includes the following specific methods:

After obtaining the SOM neural network model, the nursing data of each dimension collected in real time are input into the SOM neural network model. When the output of the SOM neural network model is an abnormal neuron, the nursing data collected in real time are abnormal data. When the output of the SOM neural network model is a normal neuron, the nursing data collected in real time are normal data.

The beneficial effect of the technical solution of the present invention is as follows: the present invention collects nursing data of all dimensions of the patient at each moment; according to the nursing data of all dimensions of the patient at each moment, obtains the similarity between the trend of each dimension and the direction vector of the principal component; since the closer the trend of a single dimension is to the direction vector of the principal component, the more this dimension can represent the characteristics presented by the patient's vital sign data, so as to prepare data for the subsequent acquisition of the core dimension;

According to the amplitude of the nursing data of each dimension and the time of collection, combined with the similarity between the trend of each dimension and the direction vector of the principal component, the characteristic expression of each dimension is obtained. The higher the characteristic expression of the dimension, the more it should be used as the dimension for training the SOM neural network model. Therefore, the core dimension can be obtained according to the characteristic expression of each dimension, and the nursing data in the core dimension can be used to judge the nursing data collected in real time, so as to finally accurately judge whether there is abnormal data in the nursing data collected in real time.

BRIEF DESCRIPTION OF THE DRAWINGS

In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

FIG1 is a block diagram of a medical care automatic monitoring system according to the present invention;

FIG. 2 is a flow chart of the present invention for determining whether there is abnormal data in the nursing data collected in real time.

DETAILED DESCRIPTION

In order to further explain the technical means and effects adopted by the present invention to achieve the predetermined invention purpose, the following is a detailed description of the specific implementation, structure, features and effects of a medical care automatic monitoring system proposed according to the present invention in combination with the accompanying drawings and preferred embodiments. In the following description, different "one embodiment" or "another embodiment" does not necessarily refer to the same embodiment. In addition, specific features, structures or characteristics in one or more embodiments may be combined in any suitable form.

Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.

The specific scheme of the automatic medical care monitoring system provided by the present invention is described in detail below with reference to the accompanying drawings.

Please refer to FIG1 , which shows a structural block diagram of a medical care automatic monitoring system provided by an embodiment of the present invention. The system includes the following modules:

The data collection module 101 is used to collect nursing data of all dimensions of the patient at every moment.

It should be noted that medical hospitals conduct clinical assessments on patients, formulate treatment plans, conduct necessary examinations and tests, and work together with doctors and nursing staff to ensure that patients receive the best medical care. As a medical care automatic monitoring system, the ultimate goal of this embodiment is to analyze the patient's nursing data and conduct a comprehensive evaluation of the patient's nursing data. Therefore, the patient's nursing data needs to be collected first.

Specifically, various nursing data of the patient at each moment are collected through electronic blood pressure measuring instruments, blood oximeters, dynamic electrocardiogram recorders and other equipment, and different nursing data are recorded as nursing data of different dimensions, where the sampling time is set to 1 second.

At this point, the nursing data of all dimensions of the patient at every moment is obtained.

The multidimensional analysis module 102 is used to obtain several principal components of the nursing data of all dimensions, the direction vector of each principal component and the variance contribution rate of each principal component based on the nursing data of all dimensions of the patient at each moment; construct a sample space based on the nursing data of each dimension of the patient, and obtain the similarity between the trend of each dimension and the direction vector of the principal component based on the position of the data point in the sample space, combined with the direction vector of each principal component and the variance contribution rate of each principal component.

It should be noted that the nursing data for automatic monitoring of the patient's nursing data is real-time data, and the traditional method of automatically monitoring the patient's nursing data is mainly through a preset threshold, by monitoring whether the nursing data exceeds the threshold, so as to determine whether the patient's nursing data has abnormalities; however, this method is limited by the value of the preset threshold, and due to the differences in the patient's physical characteristics, the accuracy of monitoring whether the nursing data has abnormalities through threshold judgment is low; therefore, this embodiment proposes a method based on the SOM neural network algorithm to determine whether the nursing data is abnormal, but because in the process of training the SOM neural network, due to the excessive dimensions of the nursing data, the high-dimensional data increases the distance between samples, and it is difficult to capture the structure and similarity between the data, which causes a dimensionality disaster, resulting in the inability to accurately determine whether there are abnormal data in the real-time collected nursing data directly through the SOM neural network algorithm.

It should be further explained that in order to improve the accuracy of the training model, this embodiment reduces the dimensionality of multidimensional data, obtains the feature difference of each dimension compared with the main component, obtains the feature expression degree of each dimension, and then obtains the core dimension, and constructs a SOM neural network model according to the core dimension to avoid causing dimensionality disaster, thereby accurately judging whether there is abnormal data in the nursing data collected in real time.

Specifically, the nursing data of all dimensions of the patient at each moment are input into the PCA principal component analysis algorithm to obtain each principal component of the nursing data of all dimensions, the direction vector of each principal component, and the variance contribution rate of each principal component; since the PCA principal component analysis algorithm is a well-known prior art, the principal component, the principal component vector, and its variance contribution rate can be obtained by the PCA principal component analysis algorithm, which will not be described in detail in this embodiment;

Then, the number of dimensions of the nursing data of all dimensions of the patient at each moment is recorded as , construct a dimensional coordinate system, placing the nursing data of all dimensions of the patient at each moment into dimensional coordinate system, and obtain the sample space;

Finally, for the Dimensions and principal component, and obtain the The direction vector of the principal component is dimensions, obtain the projection vector of the direction vector of each principal component on all dimensions; obtain the projection of all data points in the sample space on all dimensions, recorded as the projection point on all dimensions, for the The first dimension projection point, get the The first dimension The projection point, The direction vector of the principal component is The projection vector on the projection vector in the dimension is denoted as The first dimension The projection point projection vectors, get all the projection vectors of all the projection points in each dimension.

It should be noted that in the sample space composed of nursing data of multiple dimensions, the overall characteristics of the overall trend of each data in all dimensions in its dimension are the direction vector of the principal component. Then, in multidimensional data, the projection size of each data on the direction vector of the principal component determines the contribution relationship of the data value to the direction vector of the principal component. The larger the projection, the closer the data is to the direction vector of the principal component, that is, the overall trend of the data. Therefore, for the data monitored in the medical care process, it is necessary to evaluate the abnormal data according to the monitoring. The closer the trend of a single dimension is to the direction vector of the principal component, the more this dimension can represent the characteristics presented by the patient's vital sign data. When training the SOM neural network model, the models trained by these dimensions can better reflect the characteristics of normality, thereby isolating the characteristics of abnormal data. Therefore, it is necessary to obtain the similarity between the trend of each dimension and the direction vector of the principal component.

Specifically, for dimensions, the direction vector of each principal component is in The projection vector on the dimension, All projection vectors of all projection points in the dimension, the variance contribution rate of each principal component, and the The number of projection points on the dimension, get the The specific calculation formula is:

in, Indicates The similarity between the trend of the dimension and the direction vector of the principal component, Indicates The number of projection points in dimensions, Indicates The variance contribution of the principal components is Indicates The direction vector of the principal component is The projection vector in dimensions, Indicates The first dimension The projection point projection vector; Represents the number of principal components.

It should be noted that It means the The first dimension The projection point projection vector, and The direction vector of the principal component is The ratio of the projection vectors in dimensions, so The larger the value of The first dimension The closer the data point in the sample space corresponding to the projection point is to the principal components, so The larger the value of The first principal component The first dimension The higher the component content of the data points in the sample space corresponding to the projection point, that is, The larger the value of The more similar the trend of each dimension is to the direction vector of the principal component.

At this point, the similarity between the trend of each dimension and the direction vector of the principal component is obtained.

The single-dimensional analysis module 103 is used to obtain the increment of each nursing data in each dimension according to the amplitude of the nursing data in each dimension and the time of collection, and classify each nursing data in each dimension to obtain the classification result of each nursing data in each dimension; according to the increment and classification result of each nursing data in each dimension, combined with the trend of each dimension and the similarity of the direction vector of the main component, the characteristic expression degree of each dimension is obtained.

It should be noted that this embodiment is based on the SOM neural network algorithm to determine whether there are abnormalities in the patient's nursing data. When training the SOM neural network model, the selected dimensions should be mostly stable and normal, with a few fluctuating or abnormal. When training the model with data from these dimensions, the easier it is to distribute stable data to one or multiple neurons clustered together, then when optimizing all dimensional data, the dimensions with the same directional trend as the main component are further optimized, and the dimensions that can clearly reflect the differences are screened.

Specifically, for In the dimension The nursing data collected at each moment will be In the dimension The nursing data collected at the moment minus the In the dimension The difference in nursing data collected at the moment is recorded as In the dimension The increment of nursing data collected at each moment and The increment of nursing data collected at the first moment in the dimension is equal to The increment of nursing data collected at the second moment in the dimension is obtained The increment of all nursing data in the dimension will be The incremental mean of all nursing data in the dimension is recorded as the benchmark value. The nursing data with increments less than the benchmark value in the dimension are recorded as the first category of data. The nursing data with increments greater than the benchmark value in each dimension are recorded as the second type of data;

According to The increment of all nursing data in the dimension is obtained. The kurtosis of the increments of all nursing data in the dimension. Since the specific process of obtaining the kurtosis is a well-known prior art, it will not be described in detail in this embodiment. The kurtosis of the increments of all nursing data in the dimension, The similarity between the trend of the dimension and the direction vector of the principal component, the number of the first type of data and the number of the second type of data, and obtain the The specific calculation formula for the characteristic expression of each dimension is:

In the formula, Indicates The characteristic expression of each dimension; Indicates The similarity between the trend of each dimension and the direction vector of the principal component; Indicates The kurtosis of the increments of all nursing data in the dimension; Indicates the number of the first category of data; Indicates the number of the second type of data; Represents a normalization function, the normalized object is all dimensions ; Represents absolute value operation; represents an exponential function with a natural constant as the base. In this embodiment, The model shows an inverse proportional relationship. As the input of the model, the implementer can set the inverse proportional function according to the actual situation.

It should be noted that It means the The kurtosis of the increments of all nursing data in the dimension, The larger the value, the The more concentrated the increment of all nursing data in the dimension, the The overall change range of all nursing data in the dimension is within a stable range, that is, dimensions satisfy the characteristics of being stationary and normal, so The larger the value of The more dimensions should be used as the dimensions for training the SOM neural network model; It represents the ratio of the first type of data to the second type of data in terms of quantity. The closer the value is to 1, the closer the first type of data is to the second type of data in quantity. The closer the overall upward trend of the dimension is to the downward trend, that is, dimensions satisfy the characteristics of being stable and normal, and It means the The similarity between the trend of a dimension and the direction vector of the principal component is that the closer the trend of a single dimension is to the direction vector of the principal component, the more this dimension can represent the characteristics presented by the patient's physical sign data; therefore The larger the value of The more dimensions there are, the more they should be used as the dimensions for training the SOM neural network model.

So far, the feature expression of all dimensions.

The monitoring module 104 is used to obtain the core dimension according to the characteristic expression of each dimension, train the neural network model using the nursing data in the core dimension, and monitor the nursing data collected in real time according to the neural network model.

It should be noted that after the feature expression of all dimensions is obtained through the single-dimensional analysis module 103, the dimension for training the SOM neural network model can be obtained according to the feature expression of the dimension.

Specifically, a performance threshold is preset , The specific value of can be set according to the actual situation. This embodiment does not make a hard requirement. To narrate, for The characteristic expression of the dimension The characteristic expression of the dimension is greater than , then Dimensions are core dimensions, and all core dimensions are obtained.

It should be noted that the core dimension is the dimension for training the SOM neural network model. After obtaining the core dimension, the SOM neural network model can be obtained by training according to the core dimension.

Specifically, the nursing data in the core dimension is used as the input layer of the SOM neural network model, and the output layer is a A square topology network is constructed. training, and obtaining a SOM neural network model; and are the preset square side length and number of training times, and The specific value of can be set according to the actual situation. This embodiment does not make a hard requirement. , The specific training process of the SOM neural network model is a well-known prior art, so it will not be described in detail in this embodiment;

After obtaining the SOM neural network model, the nursing data of each dimension collected in real time is input into the SOM neural network model. When the output of the SOM neural network model is an abnormal neuron, the nursing data collected in real time is abnormal data. When the output of the SOM neural network model is a normal neuron, the nursing data collected in real time is normal data. Finally, it is possible to determine whether there is abnormal data in the nursing data collected in real time.

A flowchart of determining whether there is an abnormality in the nursing data collected in real time in this embodiment is shown in FIG2 .

At this point, this embodiment is completed.

The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the principles of the present invention should be included in the protection scope of the present invention.

Claims (4)

1. A medical care automatic monitoring system, characterized in that the system includes the following modules: The data collection module is used to collect nursing data of all dimensions at every moment of the patient; specifically, through various devices such as electronic blood pressure measuring instruments, blood oximeters, and dynamic electrocardiogram recorders, various nursing data of the patient at every moment are collected, and different nursing data are recorded as nursing data of different dimensions; The multi-dimensional analysis module is used to obtain several principal components of the nursing data of all dimensions, the direction vector of each principal component and the variance contribution rate of each principal component according to the nursing data of all dimensions of the patient at each moment; construct a sample space according to the nursing data of each dimension of the patient, and obtain the similarity between the trend of each dimension and the direction vector of the principal component according to the position of the data point in the sample space, combined with the direction vector of each principal component and the variance contribution rate of each principal component; The single-dimension module is used to obtain the increment of each nursing data in each dimension according to the amplitude and the time of collection of the nursing data in each dimension, and classify each nursing data in each dimension to obtain the classification result of each nursing data in each dimension; according to the increment and classification result of each nursing data in each dimension, combined with the trend of each dimension and the similarity of the direction vector of the main component, the characteristic expression degree of each dimension is obtained; The monitoring module is used to obtain the core dimension according to the characteristic expression of each dimension, train the neural network model with the nursing data in the core dimension, and monitor the nursing data collected in real time according to the neural network model; The method of constructing a sample space based on the nursing data of each dimension of the patient, and obtaining the similarity between the trend of each dimension and the direction vector of the principal component according to the position of the data point in the sample space and the direction vector of each principal component and the variance contribution rate of each principal component includes the following specific methods: Get the number of dimensions of the patient's nursing data in all dimensions at each moment, recorded as , construct a dimensional coordinate system, placing the nursing data of all dimensions of the patient at each moment into dimensional coordinate system, and obtain the sample space; For Dimensions and principal component, and obtain the The direction vector of the principal component is dimensions, obtain the projection vector of the direction vector of each principal component on all dimensions; obtain the projection of all data points in the sample space on all dimensions, recorded as the projection point on all dimensions, for the The first dimension projection point, get the The first dimension The projection point, The direction vector of the principal component is The projection vector on the projection vector in the dimension is denoted as The first dimension The projection point projection vectors, get all the projection vectors of all the projection points in each dimension; According to the direction vector of each principal component in The projection vector on the dimension, All projection vectors of all projection points in the dimension, the variance contribution rate of each principal component, and the The number of projection points on the dimension, get the The similarity between the trend of each dimension and the direction vector of the principal component; The acquisition The similarity between the trend of each dimension and the direction vector of the principal component includes the specific calculation formula: in, Indicates The similarity between the trend of the dimension and the direction vector of the principal component, Indicates The number of projection points in dimensions, Indicates The variance contribution of the principal components is Indicates The direction vector of the principal component is The projection vector in dimensions, Indicates The first dimension The projection point projection vector; represents the number of principal components; The method of obtaining the increment of each nursing data in each dimension according to the amplitude of the nursing data in each dimension and the time of collection, and classifying each nursing data in each dimension to obtain the classification result of each nursing data in each dimension includes the following specific methods: For In the dimension The nursing data collected at each moment will be In the dimension The nursing data collected at the moment minus the In the dimension The difference in nursing data collected at the moment is recorded as In the dimension The increment of nursing data collected at each moment, and obtain the The increment of all nursing data in the dimension will be The incremental mean of all nursing data in the dimension is recorded as the benchmark value. The nursing data with increments less than the benchmark value in the dimension are recorded as the first category of data. The nursing data with increments greater than the benchmark value in each dimension are recorded as the second type of data; The method of obtaining the characteristic expression of each dimension according to the increment and classification result of each nursing data in each dimension and combining the trend of each dimension with the similarity of the direction vector of the principal component includes the following specific methods: For dimensions, according to The increment of all nursing data in the dimension is obtained. The kurtosis of the increments of all nursing data in the dimension, combined with the The similarity between the trend of the dimension and the direction vector of the principal component, the number of the first type of data and the number of the second type of data, and obtain the The characteristic expression of each dimension; The acquisition The characteristic expression of each dimension includes the following specific calculation formula: In the formula, Indicates The characteristic expression of each dimension; Indicates The similarity between the trend of each dimension and the direction vector of the principal component; Indicates The kurtosis of the increments of all nursing data in the dimension; Indicates the number of the first category of data; Indicates the number of the second type of data; represents the normalization function; Represents absolute value operation; represents an exponential function with a natural constant as base; The specific method of obtaining the core dimension according to the characteristic expression of each dimension includes: Preset a performance threshold , for the The characteristic expression of the dimension The characteristic expression of the dimension is greater than , then Dimensions are core dimensions, and all core dimensions are obtained. 2. According to claim 1, a medical care automatic monitoring system is characterized in that the method of obtaining a plurality of principal components of the nursing data of all dimensions, the direction vector of each principal component and the variance contribution rate of each principal component according to the nursing data of all dimensions of the patient at each moment includes the following specific methods: The nursing data of all dimensions of the patient at each moment are input into the PCA principal component analysis algorithm to obtain several principal components of the nursing data of all dimensions, the direction vector of each principal component and the variance contribution rate of each principal component. 3. According to claim 1, a medical care automatic monitoring system is characterized in that the use of nursing data in the core dimension to train the neural network model includes the following specific methods: The nursing data in the core dimension is used as the input layer of the SOM neural network model, and the output layer is a A square topology network is constructed. training, and obtaining a SOM neural network model; and are the preset square side length and number of training times respectively. 4. According to claim 1, a medical care automatic monitoring system is characterized in that the monitoring of the real-time collected nursing data according to the neural network model includes the following specific methods: After obtaining the SOM neural network model, the nursing data of each dimension collected in real time are input into the SOM neural network model. When the output of the SOM neural network model is an abnormal neuron, the nursing data collected in real time are abnormal data. When the output of the SOM neural network model is a normal neuron, the nursing data collected in real time are normal data.