Clinical nursing data management method and system based on cloud computing

Abstract

The invention relates to the technical field of data processing, in particular to a clinical nursing data management method and system based on cloud computing, and the method comprises the steps: collecting sample points; constructing a multi-dimensional sample space and a two-dimensional sample space according to the sample points; according to the two-dimensional direction vectors of the sample points in the multi-dimensional sample space and the two-dimensional sample space, obtaining a prominence degree; screening prominent signs according to the prominence degree of each dimension in the multi-dimensional sample space; obtaining a weakening degree according to the prominence degree of the prominent signs and the prominence degrees of all dimensions in the multi-dimensional sample space; correcting the protrusion degree according to the weakening degree to obtain a corrected protrusion degree; obtaining a correction redundancy range according to the correction prominence degree; and judging whether the clinical nursing data is abnormal or not according to the corrected redundancy range. According to the method, the accuracy of clinical nursing data abnormity judgment is improved by accurately and credibly correcting the redundancy range.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of data processing, in particular to a clinical nursing data management method and system based on cloud computing. BACKGROUND

[0002] The diseases of the department of gastroenterology are often closely related to the eating and living habits, and most of the diseases, such as diabetes, have long-term chronic effects. For the chronic diseases of the department of gastroenterology, the targeted management of patients needs to be strengthened, and the extended nursing is the key to improve the quality of nursing, which can effectively improve the compliance of patients with chronic diseases, effectively prevent and reduce the occurrence of complications, and improve the quality of life of patients. However, in the process of traditional extended nursing, due to the influence of time, region and environment, it is difficult to maintain and continuous the information of extended nursing after hospitalization.

[0003] With the rapid application and popularization of intelligent terminal devices and cloud computing service platforms, patients can share and access health data information through intelligent terminal devices and cloud computing platforms, provide real-time dynamic medical services, and meet the individualized extended nursing needs of patients with chronic diseases in the department of gastroenterology. However, in the existing cloud platform service technology, the difference of individual health information of patients is not considered, which leads to the problems of low nursing efficiency, low nursing quality and poor targeted nursing effect in the process of extended nursing after hospitalization. SUMMARY

[0004] The present application provides a kind of clinical nursing data management method and system based on cloud computing to solve the existing problems.

[0005] The clinical nursing data management method and system based on cloud computing of the present application adopt the following technical solutions: One embodiment of the present application provides a clinical nursing data management method based on cloud computing, which comprises the following steps: Collecting a plurality of patients' a plurality of sign data for each sign item, obtaining a plurality of sample points for each sign item of each patient; regarding all sample points of each sign item as a dimension, constructing a multi-dimensional sample space and a two-dimensional sample space of each dimension according to all sample points of all dimensions; According to the two-dimensional direction vector of each sample point in the multi-dimensional sample space and the two-dimensional sample space of each dimension, the prominence of each dimension in the multi-dimensional sample space is obtained; according to the prominence of each dimension in the multi-dimensional sample space, a plurality of prominent signs are selected from all dimensions, and the prominent signs contain signs of a plurality of patients; According to the prominence degree of each prominent sign of each patient and the prominence degree of all dimensions in the multi-dimensional sample space, the weakening degree of each prominent sign of each patient is obtained; and the prominence degree is corrected according to the weakening degree of each prominent sign of each patient, to obtain the corrected prominence degree of each prominent sign of each patient. According to the corrected prominence degree of each prominent sign of each patient, the corrected redundancy range of the prominent sign is obtained by adjusting the redundancy range of each prominent sign of each patient; and the abnormal clinical nursing data is obtained by applying the abnormal clinical nursing data to the clinical nursing intelligent management according to all the corrected redundancy ranges of each patient.

[0006] Further, the construction of the multi-dimensional sample space and the two-dimensional sample space of each dimension according to all the sample points of all dimensions includes the following specific steps: All sample points of each sign item of each patient constitute a dimension, and all dimensions of all patients constitute a multi-dimensional sample space; The two-dimensional sample space is constructed in the following manner: for any dimension of any patient in the multi-dimensional sample space, the sampling value of the sample point of the dimension is taken as the vertical coordinate, and the mean value of the sample point of the corresponding collection item of all dimensions of all patients except the dimension at each collection time is calculated, and the sample point mean value at each collection time is taken as the horizontal coordinate to construct the two-dimensional sample space of each dimension of each patient.

[0007] Further, the construction of the multi-dimensional sample space and the two-dimensional sample space of each dimension according to all the sample points of all dimensions includes the following specific steps: The multi-dimensional sample space is input into the PCA principal component analysis algorithm, and the PCA principal component analysis algorithm is used to obtain a plurality of principal component directions in the multi-dimensional sample space and the characteristic value and variance contribution rate of each principal component direction, all the characteristic values of the principal component directions are arranged in descending order to obtain a characteristic descending sequence; A preset characteristic accumulation threshold is set, and the first characteristic value of the characteristic descending sequence is sequentially accumulated from the back, and when the accumulated characteristic value is less than the characteristic accumulation threshold, the next characteristic value in the characteristic descending sequence is accumulated; and when the accumulated characteristic value is greater than or equal to the characteristic accumulation threshold, the principal component direction corresponding to all the accumulated characteristic values is recorded as the principal direction in the multi-dimensional sample space; Each principal direction is projected onto the two-dimensional sample space of each dimension to obtain the projected principal direction of the two-dimensional sample space of each dimension; the direction vector of each sample point in each dimension in the two-dimensional sample space is obtained, denoted as the two-dimensional direction vector of each sample point in each dimension; the direction vector of each sample point in each dimension in the multidimensional sample space is obtained, denoted as the multidimensional direction vector of each sample point in each dimension; based on the two-dimensional direction vector of each sample point in any dimension, the projected principal direction of the two-dimensional sample space of each dimension, the variance contribution rate of the principal direction, and the multidimensional direction vector of each sample point, the prominence of each dimension in the multidimensional sample space is obtained.

[0008] Furthermore, the specific calculation method for obtaining the prominence of each dimension in the multidimensional sample space based on the two-dimensional direction vector of each sample point in any dimension, the projection principal direction of the two-dimensional sample space in each dimension, the variance contribution rate of the principal direction, and the multidimensional direction vector of each sample point is as follows: in, For the first in the multidimensional sample space The degree of prominence of each dimension The total number of sample points for each dimension. The number of main directions, In the first In the dimension of the first Two-dimensional direction vector of each sample point In the first In the two-dimensional sample space of the nth dimension The principal projection directions of each principal component direction. In the first In the two-dimensional sample space of the nth dimension The variance contribution rate of the principal directions before projection of each principal direction. For the first Multidimensional direction vector of each sample point For the first in the multidimensional sample space The direction vectors of the principal directions For the first in the multidimensional sample space Variance contribution rate of each principal direction This is the sigmoid function.

[0009] Furthermore, the specific steps for obtaining the degree of attenuation of each prominent sign for each patient based on the prominence of each patient's prominent sign and the prominence of all dimensions in the multidimensional sample space are as follows: The degree of particularity of all dimensions of any patient is obtained by the ratio of the kurtosis of all dimensions of particularity and the number of prominent signs to the total number of signs. The degree of difference in the degree of particularity of each prominent sign of any patient is obtained by the degree of particularity of all dimensions of any patient and the degree of difference in the degree of particularity of each prominent sign. The degree of attenuation of each prominent sign of each patient is obtained by the degree of particularity of all dimensions of any patient and the degree of difference in the degree of particularity of each prominent sign.

[0010] Furthermore, the reduction degree of each prominent sign for each patient is obtained based on the specificity of all dimensions of any patient and the degree of difference in the prominence of each prominent sign. The specific calculation method includes the following: In the formula, For the first The first patient's The degree of weakening of a prominent physical sign, For the first The kurtosis of the prominence of all dimensions of a patient. For the first The number of prominent physical signs in each patient For the first The number of all vital signs for each patient For the first The first patient's The degree of prominence of each prominent physical sign For the first The first patient's The degree of prominence of each prominent physical sign For the sigmoid function, It is an absolute value function.

[0011] Furthermore, the specific steps involved in correcting the degree of prominence based on the weakening degree of each prominent sign for each patient to obtain the corrected degree of prominence for each prominent sign include the following: Calculate 1 and the first The first patient's The difference in the degree of attenuation of each prominent symptom, and the difference is compared with the first... The first patient's The product of the prominence of each prominent physical feature is denoted as the first. The first patient's The degree of prominence of the modified prominent physical signs.

[0012] Furthermore, the specific steps involved in adjusting the redundancy range of each prominent sign for each patient based on the corrected degree of prominence for each prominent sign to obtain the corrected redundancy range for the prominent sign are as follows: Preset correction highlight threshold, when the first The first patient's When the corrected prominence of a prominent feature is greater than or equal to the corrected prominence threshold, the first... The normal redundancy range of the first prominent symptom is used as the corrected redundancy range; when the first... The first patient's When the correction prominence of a prominent feature is less than the preset correction prominence threshold, the first prominent feature will be... The first patient's The corrected salience level of each prominent feature is multiplied by the lower boundary value of its normal redundancy range, and the result is used as the first... The first patient's The lower boundary value of the corrected redundancy range for the prominent symptom, and the value of 2 with the first The first patient's The difference in the corrected prominence of each prominent feature is used as a gain coefficient, and the gain coefficient is compared with the first... The first patient's Multiplying the upper boundary values ​​of the normal redundancy range of each prominent physical sign by the result, we obtain the first... The first patient's The upper boundary value of the corrected redundancy range for the first prominent symptom is obtained, thus yielding the first... The first patient's Correction redundancy range for prominent physical signs.

[0013] Furthermore, the specific steps for obtaining abnormal clinical nursing data by judging all corrected redundancy ranges for each patient include the following: Obtain the corrected redundancy range for each prominent sign of each patient. For any prominent sign of a patient, if the sample value of the sample point of the prominent sign is within the corrected redundancy range, the sample point of the prominent sign is judged as normal clinical nursing data; if there are sample points of the prominent sign that are not within the redundancy range, the sample points of the prominent sign are judged as abnormal clinical nursing data.

[0014] The present invention also proposes a cloud-based clinical nursing data management system, including a memory, a processor, and a computer program stored in the memory and executable on the processor. The processor executes the computer program stored in the memory to implement the steps of the aforementioned cloud-based clinical nursing data management method.

[0015] The beneficial effects of the technical solution of this invention are as follows: This invention acquires sample points of patients' vital signs and constructs a multidimensional sample space and a two-dimensional sample space; based on the two-dimensional direction vector of each sample point in the multidimensional sample space and the two-dimensional sample space, it obtains the degree of prominence of each dimension in the multidimensional sample space, and then filters out prominent vital signs; based on the degree of prominence of the patient's prominent vital signs and the degree of prominence of all dimensions in the multidimensional sample space, it obtains the degree of weakening of the prominent vital signs, and then corrects the degree of prominence to obtain the corrected degree of prominence; based on the corrected degree of prominence, it adjusts the redundancy range to obtain the corrected redundancy range, and then judges whether the clinical nursing data is abnormal, thereby realizing intelligent management of clinical nursing. Specifically, by utilizing the two-dimensional direction vectors of each sample point in both the multidimensional and two-dimensional sample spaces, the salience of each dimension in the multidimensional sample space is determined, and prominent signs are screened to narrow down the target range for sign judgment, making the anomaly judgment results of clinical nursing data more accurate. Based on the salience of the patient's prominent signs and the salience of all dimensions in the multidimensional sample space, the weakening degree of the prominent signs is obtained, accurately determining the degree of adjustment of the redundancy range, enabling the redundancy range to achieve adaptive adjustment. Based on the weakening degree of the prominent signs, the salience is corrected to obtain the corrected salience, which adjusts the redundancy range and judges the data anomalies in clinical nursing data, improving the accuracy of the anomaly judgment results of clinical nursing data. Attached Figure Description

[0016] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0017] Figure 1 This is a flowchart illustrating the steps of a cloud-based clinical nursing data management method according to the present invention. Detailed Implementation

[0018] To further illustrate the technical means and effects adopted by the present invention to achieve its intended purpose, the following, in conjunction with the accompanying drawings and preferred embodiments, details the specific implementation, structure, features, and effects of a cloud-based clinical nursing data management method and system proposed according to the present invention. In the following description, different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. Furthermore, specific features, structures, or characteristics in one or more embodiments can be combined in any suitable form.

[0019] Unless otherwise defined, 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 pertains.

[0020] The following description, in conjunction with the accompanying drawings, details a specific solution for a cloud-based clinical nursing data management method and system provided by this invention.

[0021] Please see Figure 1 The diagram illustrates a flowchart of a cloud-based clinical nursing data management method according to an embodiment of the present invention, which includes the following steps: Step S001: Collect several vital sign data for each vital sign item from several patients to obtain several sample points for each vital sign item for each patient; take all the sample points for each vital sign item as a dimension, and construct a multidimensional sample space and a two-dimensional sample space for each dimension based on all the sample points for all dimensions.

[0022] The purpose of this embodiment is to extract the patient's vital signs that show fluctuations that differ from the normal range, in order to avoid errors when cloud computing processes internal and gastrointestinal nursing data. Therefore, it is necessary to collect vital sign data from all patients first.

[0023] Specifically, gastroenterology nursing data for all patients is obtained from the gastroenterology department. This data includes several vital signs for each patient, such as heart rate, blood oxygen saturation, and blood pressure. Each vital sign contains several sample points. The sampling frequency for each vital sign is the same as the sampling frequency of the acquisition device. The vital sign with the highest sampling frequency is selected and its sampling frequency is recorded as the standard frequency. A linear interpolation algorithm is used to interpolate and fill in the missing values ​​of other vital signs according to the standard frequency, ensuring that each vital sign has a value at every acquisition time, thus obtaining the temporal sequence of each vital sign. Based on the above method, the temporal sequence of all vital signs for each patient is obtained. The linear interpolation algorithm is a well-known technique, and its specific method will not be described here.

[0024] Specifically, all sample points for each vital sign item of each patient constitute a dimension, and all dimensions of all patients constitute a multidimensional sample space. For any dimension of any patient in the multidimensional sample space, the sampled value of the sample points of that dimension is used as the vertical axis, and the mean value of the sample points of all dimensions of the corresponding collection items of all other patients is calculated at each collection time. The mean value of the sample points at each collection time is used as the horizontal axis to construct a two-dimensional sample space for each dimension of each patient.

[0025] Step S002: Based on the multidimensional sample space and the two-dimensional direction vector of each sample point in the two-dimensional sample space of each dimension, obtain the prominence of each dimension in the multidimensional sample space; based on the prominence of each dimension in the multidimensional sample space, select several prominent signs from all dimensions, and the prominent signs include signs of multiple patients.

[0026] It should be noted that due to the coordinated control of the human body, there are certain correlations between the vital signs of a patient, including positive and negative correlations. Therefore, in a multidimensional sample space composed of all dimensions of all patients, if the projection of each sample point in one dimension of a patient in the sample space composed of its own dimensions differs significantly from its projection in the multidimensional sample space, it indicates that the vital sign corresponding to that dimension is more likely to be a prominent sign. In a two-dimensional sample space, the distribution of sample points tends to approximate y=x, and the further away from this trend, the greater the difference in the sample points. Therefore, this embodiment obtains the prominence of each dimension in the multidimensional sample space based on the difference between the projection of each sample point in one dimension of a patient in the sample space composed of its own dimensions and its projection in the multidimensional sample space.

[0027] Specifically, the multidimensional sample space is input into the PCA principal component analysis algorithm. The PCA principal component analysis algorithm is used to obtain several principal component directions in the multidimensional sample space, as well as the eigenvalues ​​and variance contribution rate of each principal component direction. The eigenvalues ​​of all principal component directions are arranged in descending order to obtain the feature descending sequence. Furthermore, a preset feature accumulation threshold is set. This embodiment uses a feature accumulation threshold. The description is provided below; other values ​​may be used in other implementations, and this embodiment is not limited to them. The feature values ​​are accumulated sequentially from the first feature value in the descending sequence. When the accumulated feature value is less than the feature accumulation threshold... When the summed feature value is greater than or equal to the feature summing threshold, the next feature value in the descending feature sequence is accumulated. When all the accumulated eigenvalues ​​are recorded as principal component directions in the multidimensional sample space, each principal direction is projected onto the two-dimensional sample space of each dimension to obtain the projected principal direction of the two-dimensional sample space of each dimension.

[0028] Obtain the direction vector of each sample point in each dimension in the two-dimensional sample space, denoted as the two-dimensional direction vector of each sample point in each dimension; obtain the direction vector of each sample point in each dimension in the multi-dimensional sample space, denoted as the multi-dimensional direction vector of each sample point in each dimension. In the multidimensional sample space, the first The method for calculating the prominence of each dimension is as follows: in, For the first in the multidimensional sample space The degree of prominence of each dimension The total number of sample points for each dimension. The number of main directions, In the first In the dimension of the first Two-dimensional direction vector of each sample point In the first In the two-dimensional sample space of the nth dimension The principal projection directions of each principal component direction. In the first In the two-dimensional sample space of the nth dimension The variance contribution rate of the principal directions before projection of each principal direction. For the first Multidimensional direction vector of each sample point For the first in the multidimensional sample space The direction vectors of the principal directions For the first in the multidimensional sample space Variance contribution rate of each principal direction This is the sigmoid function.

[0029] Represented as the first The sample point at the th th In the two-dimensional sample space of the nth dimension, for the nth The projection value in the principal direction, the magnitude of the projection value indicates its relationship with the projection value in the first principal direction. In the two-dimensional sample space of the nth dimension, and the nth The directional differences of the main directions For the first The nth sample point in the multidimensional sample space is related to the nth sample point. The projection values ​​of the principal directions are used, and the variance contribution rate is used as the influence weight to correct for the differences. The more similar the projection values ​​of the two directions are, the better the projection values ​​of the principal directions are. The sample point at the th th The influence generated in the two-dimensional sample space of the first dimension and the second dimension The influence of each dimension on the multidimensional sample space is approximated, then the th dimension... The first dimension The sample point and the first The more similar the influence directions of the principal directions are, the better. The more likely the first dimension is to be a normal dimension, the greater the difference between the projected values ​​of the two dimensions, indicating that the first dimension is more likely to be a normal dimension. The sample point at the th th The greater the deviation between the influence generated in the two-dimensional sample space and its influence in the multi-dimensional sample space, the less it conforms to the direction or trend of changes in vital signs of all patients, indicating that the influence in the two-dimensional sample space is greater than its influence in the multi-dimensional sample space. The more likely the first dimension is to be an anomalous dimension, the more likely the second dimension is to be an anomalous dimension The prominence of each dimension The larger the value, the better.

[0030] Furthermore, a preset highlight threshold is defined. This embodiment emphasizes the threshold. The description is provided below; other values ​​may be set in other implementations, and this embodiment is not limited thereto. Based on the above method, the salience degree of each dimension in the multidimensional sample space is obtained. When the... The prominence of each dimension When the value exceeds the preset threshold w, the first... Each dimension is recorded as a prominent feature.

[0031] Step S003: Based on the degree of prominence of each patient's prominence and the degree of prominence of all dimensions in the multidimensional sample space, obtain the degree of weakening of each prominence of each patient; based on the degree of weakening of each prominence of each patient, correct the degree of prominence to obtain the corrected degree of prominence of each patient's prominence.

[0032] It should be noted that some patients have different physical conditions compared to other patients, resulting in most of their physical signs being prominent. Therefore, it is necessary to adjust the prominent signs according to the degree of prominence of each patient's physical signs. For a patient whose most of their physical signs are prominent, it is necessary to analyze the degree of weakening of the prominent signs to avoid misjudging normal signs as prominent signs due to differences in physical conditions between patients and other patients, i.e., normal signs appearing as abnormal signs.

[0033] Specifically, the first The first patient's The method for calculating the degree of weakening of a prominent physical sign is as follows: In the formula, For the first The first patient's The degree of weakening of a prominent physical sign, For the first The kurtosis of all dimensions of prominence for each patient is calculated using a well-known technique, the specific method of which will not be described here. For the first The number of prominent physical signs in each patient For the first The number of all vital signs for each patient For the first The first patient's The degree of prominence of each prominent physical sign For the first The first patient's The degree of prominence of each prominent physical sign For the sigmoid function, It is an absolute value function.

[0034] In the formula, The larger the value, the more likely it is to be the first. The greater the specificity of each patient across all dimensions, and the more specific the first patient, the more specific the first patient. The higher the proportion of each patient's total physical signs across all dimensions, the more likely it is that the patient is in a certain condition. If multiple signs in a patient are prominent signs, then the first... The first patient's The greater the value of the weakening of the most prominent symptom, the more likely it is due to the weakening of the first symptom. This is due to differences in the physical signs of individual patients; Indicates the first The first patient's The prominence of the first prominent physical sign is related to the degree of prominence of the second prominent physical sign. The degree of difference in the prominence of other physical signs among the patients; the larger the value, the more significant the difference. The first patient's The more likely a prominent physical sign is to be abnormal rather than caused by differences in physical signs, the smaller the value of its attenuation, i.e., the more likely the attenuation value is to be retained. The first patient's One prominent physical sign.

[0035] It should be noted that the greater the weakening of a patient's prominent sign, the greater the degree of correction required for that prominent sign. Therefore, the degree of correction for each prominent sign for each patient is calculated based on the degree of weakening of each prominent sign.

[0036] Specifically, calculate 1 and the... The first patient's The difference in the degree of attenuation of each prominent symptom, and the difference is compared with the first... The first patient's The product of the prominence of each prominent physical feature is denoted as the first. The first patient's The degree of prominence of the modified prominent physical signs.

[0037] Step S004: Based on the corrected prominence degree of each prominent sign for each patient, adjust the redundancy range of the corrected prominence degree of each prominent sign for each patient to obtain the corrected redundancy range of the prominent sign; based on all the corrected redundancy ranges for each patient, judge the clinical nursing data to obtain abnormal clinical nursing data, and apply the abnormal clinical nursing data to intelligent clinical nursing management.

[0038] Specifically, preset correction threshold This embodiment uses a modified protrusion threshold. The value is 0.4, but other values ​​can be set in other embodiments, and this embodiment is not limited thereto; when the first... The first patient's The correction prominence level of each prominent feature is greater than or equal to the preset correction prominence threshold. When, explain the first The first patient's If the first prominent physical sign is an abnormal sign, then no adjustment of the redundancy range is needed, and the second sign can be used. The normal redundancy range of the first prominent vital sign is used as the corrected redundancy range, where the normal redundancy range characterizes the normal numerical range of the vital sign data; when the first... The first patient's The correction prominence of each prominent feature is less than the preset correction prominence threshold. When, explain the first The first patient's The first prominent physical sign is a normal sign, which is due to individual differences among patients. The first patient's If one prominent physical sign is abnormal, then the first one will be... The normal redundancy range of the prominent symptom is expanded to obtain a redundancy range with a larger numerical range, which serves as the first... The correction redundancy range for the first prominent symptom, i.e., the first... The lower boundary value of the corrected redundancy range for a prominent symptom is smaller than the lower boundary value of the normal redundancy range. The upper boundary value of the corrected redundancy range for a prominent symptom is greater than the upper boundary value of the normal redundancy range. The interval between the lower boundary values ​​of the normal and corrected redundancy ranges, and the interval between the upper boundary values ​​of the normal and corrected redundancy ranges, can be fixed. In an exemplary embodiment, the first... The corrected salience level of each prominent feature is multiplied by the lower boundary value of its normal redundancy range, and the result is used as the first... The lower boundary value of the corrected redundancy range for the prominent symptom, and the value of 2 with the first The difference in the corrected prominence of the first prominent feature is used as a gain coefficient, which is then compared with the first... Multiplying the upper boundary values ​​of the normal redundancy range of each prominent physical sign by the result, we obtain the first... The upper boundary value of the corrected redundancy range of the first prominent feature is obtained, thus yielding the first... The redundancy range of the prominent physical characteristics is corrected. Through the above adjustment process, the first... The smaller the degree of prominence of a prominent symptom correction, the larger the numerical range of its correction redundancy range.

[0039] The system obtains the corrected redundancy range for each prominent sign of each patient. For any prominent sign of a patient, if the sample value of the sample point of the prominent sign is within the corrected redundancy range, the sample point of the prominent sign is judged as normal clinical nursing data; if there are sample points of the prominent sign that are not within the redundancy range, the sample points of the prominent sign are judged as abnormal clinical nursing data. The data of all sample points corresponding to the prominent sign are input into a cloud-based clinical nursing data management system, providing medical staff with more accurate and timely decision support.

[0040] The present invention also provides a cloud-based clinical nursing data management system, including a memory, a processor, and a computer program stored in the memory and executable on the processor. The processor executes the computer program stored in the memory to implement the steps of the aforementioned cloud-based clinical nursing data management method.

[0041] This invention is now complete.

[0042] 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 within the protection scope of the present invention.

Claims

1. A cloud-based clinical nursing data management method, characterized in that, The method includes the following steps: Collect several vital signs data for each vital sign item from several patients to obtain several sample points for each vital sign item for each patient; take all the sample points for each vital sign item as a dimension, and construct a multidimensional sample space and a two-dimensional sample space for each dimension based on all the sample points for all dimensions. Based on the multidimensional sample space and the two-dimensional direction vector of each sample point in the two-dimensional sample space of each dimension, the prominence of each dimension in the multidimensional sample space is obtained; based on the prominence of each dimension in the multidimensional sample space, several prominent signs are selected from all dimensions, and the prominent signs include signs of multiple patients. Based on the prominence of each patient's prominence and the prominence of all dimensions in the multidimensional sample space, the weakening degree of each prominence of each patient is obtained; based on the weakening degree of each prominence of each patient, the prominence is corrected to obtain the corrected prominence of each patient's prominence. Based on the corrected prominence level of each prominent sign for each patient, the redundancy range is adjusted to obtain the corrected redundancy range of the prominent sign; based on all corrected redundancy ranges for each patient, abnormal clinical nursing data is obtained from the clinical nursing data, and the abnormal clinical nursing data is applied to intelligent clinical nursing management.

2. The clinical nursing data management method based on cloud computing according to claim 1, characterized in that, The specific steps involved in constructing a multidimensional sample space and a two-dimensional sample space for each dimension based on all sample points across all dimensions are as follows: All sample points for each vital sign item of each patient constitute one dimension, and all dimensions of all patients constitute a multidimensional sample space. The specific construction method of the two-dimensional sample space is as follows: For any dimension of any patient in the multi-dimensional sample space, the sampled value of the sample point of the dimension is used as the vertical axis, and the mean value of the sample points of all dimensions of the collection items corresponding to the dimension of all other patients is calculated. The mean value of the sample points at each collection time is used as the horizontal axis to construct the two-dimensional sample space of each dimension of each patient.

3. The clinical nursing data management method based on cloud computing according to claim 1, characterized in that, The specific steps for obtaining the prominence of each dimension in the multidimensional sample space based on the multidimensional sample space and the two-dimensional direction vector of each sample point in the two-dimensional sample space of each dimension are as follows: The multidimensional sample space is input into the PCA principal component analysis algorithm. The PCA principal component analysis algorithm is used to obtain several principal component directions in the multidimensional sample space, as well as the eigenvalues ​​and variance contribution rate of each principal component direction. The eigenvalues ​​of all principal component directions are arranged in descending order to obtain the feature descending sequence. A preset feature accumulation threshold is set, and features are accumulated sequentially from the first feature value in the descending feature sequence. When the accumulated feature value is less than the feature accumulation threshold, the next feature value in the descending feature sequence is accumulated. When the accumulated feature value is greater than or equal to the feature accumulation threshold, the principal component directions corresponding to all accumulated feature values ​​are recorded as the principal directions in the multidimensional sample space. Each principal direction is projected onto the two-dimensional sample space of each dimension to obtain the projected principal direction of the two-dimensional sample space of each dimension; the direction vector of each sample point in each dimension in the two-dimensional sample space is obtained, denoted as the two-dimensional direction vector of each sample point in each dimension; the direction vector of each sample point in each dimension in the multidimensional sample space is obtained, denoted as the multidimensional direction vector of each sample point in each dimension; based on the two-dimensional direction vector of each sample point in any dimension, the projected principal direction of the two-dimensional sample space of each dimension, the variance contribution rate of the principal direction, and the multidimensional direction vector of each sample point, the prominence of each dimension in the multidimensional sample space is obtained.

4. The clinical nursing data management method based on cloud computing according to claim 3, characterized in that, The method for calculating the salience of each dimension in the multidimensional sample space based on the two-dimensional direction vector of each sample point in any dimension, the principal projection direction of the two-dimensional sample space in each dimension, the variance contribution rate of the principal direction, and the multidimensional direction vector of each sample point includes the following specific calculation method: in, For the first in the multidimensional sample space The degree of prominence of each dimension The total number of sample points for each dimension. The number of main directions, In the first In the dimension of the first Two-dimensional direction vector of each sample point In the first In the two-dimensional sample space of the nth dimension The principal projection directions of each principal component direction. In the first In the two-dimensional sample space of the nth dimension The variance contribution rate of the principal directions before projection of each principal direction. For the first Multidimensional direction vector of each sample point For the first in the multidimensional sample space The direction vectors of the principal directions For the first in the multidimensional sample space Variance contribution rate of each principal direction This is the sigmoid function.

5. The clinical nursing data management method based on cloud computing according to claim 1, characterized in that, The specific steps for determining the degree of attenuation of each prominent sign for each patient, based on the prominence of each patient's prominent sign and the prominence of all dimensions in the multidimensional sample space, are as follows: The degree of particularity of all dimensions of a patient is obtained by the ratio of the kurtosis of the prominence of all dimensions and the number of prominence signs to the total number of signs for any given patient. Based on the degree of prominence of each prominent sign in any patient, the degree of difference in the prominence of each prominent sign is obtained; Based on the specificity of all dimensions of any patient and the degree of difference in the prominence of each prominent sign, the degree of attenuation of each prominent sign for each patient is obtained.

6. The clinical nursing data management method based on cloud computing according to claim 5, characterized in that, The reduction degree of each prominent sign for each patient is obtained based on the specificity of all dimensions of any patient and the degree of difference in the prominence of each prominent sign. The specific calculation method is as follows: In the formula, For the first The first patient's The degree of weakening of a prominent physical sign, For the first The kurtosis of the prominence of all dimensions of a patient. For the first The number of prominent physical signs in each patient For the first The number of all vital signs for each patient For the first The first patient's The degree of prominence of each prominent physical sign For the first The first patient's The degree of prominence of each prominent physical sign For the sigmoid function, It is an absolute value function.

7. The clinical nursing data management method based on cloud computing according to claim 1, characterized in that, The specific steps involved in adjusting the degree of prominence based on the weakening degree of each prominent sign for each patient to obtain the adjusted degree of prominence for each prominent sign are as follows: Calculate 1 and the first The first patient's The difference in the degree of attenuation of each prominent symptom, and the difference is compared with the first... The first patient's The product of the prominence of each prominent physical feature is denoted as the first. The first patient's The degree of prominence of the modified prominent physical signs.

8. The clinical nursing data management method based on cloud computing according to claim 1, characterized in that, The specific steps involved in adjusting the redundancy range of each prominent sign for each patient based on the corrected degree of prominence for each prominent sign to obtain the corrected redundancy range for the prominent sign are as follows: Preset correction highlight threshold, when the first The first patient's When the corrected prominence of a prominent feature is greater than or equal to the corrected prominence threshold, the first... The normal redundancy range of the first prominent symptom is used as the corrected redundancy range; when the first... The first patient's When the correction prominence of a prominent feature is less than the preset correction prominence threshold, the first prominent feature will be... The first patient's The corrected salience level of each prominent feature is multiplied by the lower boundary value of its normal redundancy range, and the result is used as the first... The first patient's The lower boundary value of the corrected redundancy range for the prominent symptom, and the value of 2 with the first The first patient's The difference in the corrected prominence of each prominent feature is used as a gain coefficient, and the gain coefficient is compared with the first... The first patient's Multiplying the upper boundary values ​​of the normal redundancy range of each prominent physical sign by the result, we obtain the first... The first patient's The upper boundary value of the corrected redundancy range for the first prominent symptom is obtained, thus yielding the first... The first patient's Correction redundancy range for prominent physical signs.

9. The clinical nursing data management method based on cloud computing according to claim 1, characterized in that, The specific steps involved in obtaining abnormal clinical nursing data by judging the clinical nursing data based on all corrected redundancy ranges for each patient are as follows: Obtain the corrected redundancy range for each prominent sign of each patient. For any prominent sign of a patient, if the sample value of the sample point of the prominent sign is within the corrected redundancy range, the sample point of the prominent sign is judged as normal clinical nursing data. If a sample point in the prominent sign is not within the redundancy range, the sample point of the prominent sign is judged as abnormal clinical nursing data.

10. A cloud-based clinical nursing data management system, comprising a memory, a processor, and a computer program stored on the memory and executable on the processor, characterized in that, When the computer program is executed by the processor, it implements the steps of a cloud-based clinical nursing data management method as described in any one of claims 1-9.

Images