A personalized medication regimen generation method for rheumatoid eye involvement patients

Abstract

The present application relates to the technical field of medical data processing, in particular to a kind of individualized medication scheme generation method for rheumatoid eye involvement patient, comprising: obtaining the medication scheme of each patient and the data of each dimension before and after each patient medication, the patient is aggregated into multiple class clusters, and then the general drug of all patients in each class cluster is obtained;With the difference of the drug contained in the medication scheme of each patient, the membership of each patient belonging to each class cluster is obtained;With the difference of the change value of the data of each dimension before and after the medication of each patient and the patient in each class cluster, the general drug of all patients in each class cluster is the performance degree of the best drug of each user, and the individualized medication scheme of each patient is obtained.The present application aims to solve the problem of obtaining the medication scheme of the patient without considering the physical fitness of the patient when treating rheumatoid eye involvement patient.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of medical data processing, and particularly relates to a personalized medication scheme generation method for rheumatism eye involvement patients. BACKGROUND

[0002] Rheumatism is a group of multiple system diseases related to the pathogenesis and autoimmune function disorder, and its lesions involve various systems and organs of the whole body, and the eye is one of the common involved organs of rheumatism; many eye inflammations are related to rheumatic immune diseases, and the common inflammatory eye diseases can be divided into three categories in clinic, one of which is the inflammatory eye disease often accompanied or prompted by rheumatism, i.e. rheumatism causes eye diseases; therefore, in the treatment of rheumatism, the eye and other parts of the body are relieved to treat the eye diseases of rheumatism eye involvement patients.

[0003] Currently, when doctors treat rheumatism patients, the database determines the symptom degree of the patient according to the data of each dimension of the patient before treatment, and recommends a personalized medication scheme for the patient to the doctor for treatment; however, due to the different physical conditions of each patient from other patients, the absorption degree of different patients to different drugs is different, and patients with the same symptoms may have different medication schemes, but the current patients with the same symptoms are given the same medication scheme, which may not be the best medication scheme for the patient. SUMMARY

[0004] The present application provides a personalized medication scheme generation method for rheumatism eye involvement patients, to solve the problem that the existing treatment of rheumatism eye involvement patients does not consider the difference in physical quality of patients.

[0005] The personalized medication scheme generation method for rheumatism eye involvement patients provided by the present application adopts the following technical scheme:

[0006] The present application provides a personalized medication scheme generation method for rheumatism eye involvement patients, to solve the problem that the existing treatment of rheumatism eye involvement patients does not consider the difference in physical quality of patients.

[0007] Obtain the medication scheme of each patient and the data of each dimension before and after medication of each patient, and the standard value of the data of each dimension;

[0008] Based on the changes in data across various dimensions before and after medication for each patient, patients are clustered into multiple groups. The similarity of medication regimens among different patients within the same group is determined by the differences in these changes. The prevalence of each drug in each group's regimen is obtained based on the distribution of drugs in each patient's regimen within each group and the similarity of regimens among different patients within the same group, thus identifying universal drugs for all patients within each group. Finally, the membership degree of each patient to each group is determined by the difference between the universal drugs for all patients within each group and the drugs included in each patient's regimen.

[0009] Based on the differences in the changes of data in each dimension before and after medication for each patient and patients in each cluster, the final physical differences of all patients in each cluster are obtained; combined with the membership degree of each patient to each cluster, the performance of the universal drug for all patients in each cluster as the best drug for each user is obtained.

[0010] Based on the performance of the same universal drug for all patients within all clusters as the best drug for a user, the matching degree of each drug as the best drug for each patient is obtained, and then a personalized medication plan for each patient is generated and recommended.

[0011] Furthermore, the method for obtaining the similarity of medication regimens among different patients in the same cluster based on the differences in data changes across various dimensions before and after medication includes:

[0012] ;

[0013] In the formula, Indicates the first In the cluster, the th The patient and the first Similarity of medication regimens among patients; Indicates the first Standard values ​​for each dimension of data. Indicates the first All patients in each cluster before and after medication were on day 2. The mean of the absolute values ​​of the changes in data across each dimension Indicates the first In the cluster, the th The patients before and after medication were on the [number]th [day / month]. Changes in data across each dimension Indicates the first In the cluster, the th The patients before and after medication were on the [number]th [day / month]. Changes in data across each dimension Indicates the number of dimensions collected. Represents the absolute value function. To prevent hyperparameters with a denominator of 0, This represents the weight normalization function. This represents the sigmoid function.

[0014] Furthermore, the method for obtaining the prevalence of each drug in each cluster based on the distribution of drugs in the medication regimens of each patient within each cluster and the similarity of medication regimens among different patients within the same cluster includes the following specific methods:

[0015] According to the The specific formula for calculating the representativeness of the medication regimen for each patient in each cluster, based on the similarity of the medication regimens of each patient to other patients, is as follows:

[0016] ;

[0017] In the formula, Indicates the first In the cluster, the th The representativeness of the medication regimen for each patient Indicates the first The number of patients contained in each cluster Indicates the first In the cluster, the th The patient and the first The similarity of medication regimens among patients This represents the weight normalization function;

[0018] Based on the representativeness of each medication regimen in each cluster and the distribution of drugs in the medication regimen of each patient in each cluster, the prevalence of each drug in the medication regimen of each cluster is obtained.

[0019] Furthermore, the method for obtaining the prevalence of each drug in each cluster based on the representativeness of each medication regimen within each cluster and the distribution of drugs in the medication regimen of each patient within each cluster includes the following specific methods:

[0020] ;

[0021] In the formula, Indicates the first The drug in Popularity of medication regimens in each category Indicates the first The number of patients contained in each cluster Indicates the first The medication regimen within each cluster contains the first... The number of patients for each drug Indicates the first The medication regimen in this cluster does not include the [number]. The number of patients for each drug Indicates the first The medication regimen within each cluster contains the first... The first drug The representativeness of the medication regimen for each patient Indicates the first The medication regimen within each cluster contains the first... The first drug The patient and those not included in the first The first drug Similarity of medication regimens among patients.

[0022] Furthermore, the specific method for obtaining the universal drug for all patients within each cluster includes:

[0023] Preset popularity threshold If the first The drug in Popularity of medication regimens in each cluster Then the first The drug is the first A universal drug for patients within a certain cluster was obtained. A universal drug for all patients within a specific cluster.

[0024] Furthermore, the method for determining the membership degree of each patient to each cluster based on the differences between the common drugs for all patients within each cluster and the drugs included in each patient's medication regimen includes:

[0025] ;

[0026] In the formula, Indicates the first The patient belongs to the first... Membership degree of each cluster, Indicates the first The common medication for all patients within each cluster and the first The number of drug types included in a patient's medication regimen. Indicates that they appear simultaneously in the first... The common medication for all patients within each cluster and the first The number of drug types in a patient's medication regimen.

[0027] Furthermore, the method for obtaining the final physical differences of all patients within each cluster based on the differences in data changes across various dimensions before and after medication for each patient and patients within each cluster includes the following specific methods:

[0028] ;

[0029] In the formula, Indicates the first The patient and the first The final physical differences among all patients within each cluster Indicates the number of dimensions collected. Indicates the first The patients before and after medication were on the [number]th [day / month]. Changes in data across each dimension Indicates the first All patients in each cluster before and after medication were on day 2. The mean of the absolute values ​​of the changes in data across each dimension Indicates the first All patients in each cluster before and after medication were on day 2. The mean of the changes in the data across each dimension Indicates the first Standard values ​​for each dimension of data. Represents the absolute value function. This represents the weight normalization function.

[0030] Furthermore, the specific method for obtaining the performance of the universal drug for all patients within each cluster as the optimal drug for each user includes:

[0031] ;

[0032] In the formula, Indicates the first The universal medication for all patients within each cluster is the [number]. The best drug performance for each user Indicates the first The patient and the first The final physical differences among all patients within each cluster This indicates that the s-th patient belongs to the s-th patient. Membership degree of each cluster, To prevent hyperparameters with a denominator of 0.

[0033] Furthermore, the specific method for obtaining the best drug match for each patient based on the performance of the same universal drug for all patients within all clusters as the best drug for a user includes:

[0034] ;

[0035] In the formula, Indicates the first The drug is the first The optimal drug match for each patient This indicates that the generic medication for all patients contains the first... The number of drug clusters, This indicates that the generic medication for all patients contains the first... The first drug The universal medication for all patients within each cluster is the [number]. The best drug performance for each user This represents the sigmoid function.

[0036] Furthermore, the specific methods for generating personalized medication plans for each patient include:

[0037] Preset probability threshold ,like Then the first The drug is the first The best medicine for each patient;

[0038] The first All the best medications for the patient as the first The medications in the personalized medication plan for each patient are generated. Personalized medication plans for each patient.

[0039] The beneficial effects of the technical solution of this invention are as follows: When obtaining personalized medication plans for patients with rheumatic eye involvement, this invention, based on the differences in data changes in various dimensions before and after medication, groups patients with similar post-medication physical changes into one category. Based on the differences in post-medication physical changes between each patient and patients in different clusters, the differences in physical condition between each patient and patients in different clusters are obtained, solving the problem of not considering the differences in physical condition among different patients when obtaining customized medication plans for each patient. Based on the distribution of drugs in the medication plans of all patients within each cluster, a general medication plan for patients with similar post-medication physical changes is obtained. Based on the differences between each patient's medication plan and the general medication plan for patients within each cluster, the membership degree of each patient to each cluster is obtained from the perspective of drug analysis. Furthermore, based on the differences in physical condition between each patient and patients within each cluster, and considering the differences in physical condition between each patient and other patients, the membership degree is corrected. Combining the general medication plan for patients within each cluster, and considering the physical condition of each patient, a personalized medication plan recommendation is made for each patient. Attached Figure Description

[0040] 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.

[0041] Figure 1 This is a flowchart illustrating the steps of a method for generating an individualized medication regimen for patients with rheumatic eye involvement according to the present invention. Detailed Implementation

[0042] 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 method for generating individualized medication regimens for patients with rheumatic eye involvement based on 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.

[0043] 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.

[0044] The following describes in detail, with reference to the accompanying drawings, a specific scheme for generating an individualized medication regimen for patients with rheumatic eye involvement provided by the present invention.

[0045] Please see Figure 1 The diagram illustrates a flowchart of a method for generating an individualized medication regimen for patients with rheumatic eye involvement, according to an embodiment of the present invention. The method includes the following steps:

[0046] Step S001: Obtain the medication regimen for each patient, the data for each patient before and after medication, and the standard values ​​for each dimension of the data.

[0047] Specifically, data on each patient with rheumatic eye involvement before and after medication is obtained from the hospital's medical system. This data includes changes in each dimension before and after medication, the medications included in each patient's treatment plan, and the standard values ​​for each dimension. The number of data collection dimensions is preset in this embodiment. The dimensions collected are the OSDI score of the eye, tear film breakup time, Sjögren's syndrome, and facial margin and meibomian gland opening score; in other implementations, the number of dimensions and the selected dimensions can be set to other values.

[0048] Step S002: Based on the changes in data across various dimensions before and after medication for each patient, patients are clustered into multiple groups; based on the differences in the changes in data across various dimensions before and after medication for different patients within each group, the similarity of medication regimens among different patients in the same group is obtained; based on the distribution of drugs in the medication regimens of each patient within each group and the similarity of medication regimens among different patients in the same group, the prevalence of each drug in the medication regimens of each group is obtained, thereby obtaining the universal drugs for all patients within each group; based on the differences between the universal drugs for all patients within each group and the drugs included in the medication regimens of each patient, the membership degree of each patient belonging to each group is obtained.

[0049] It should be noted that, under current circumstances, when doctors prescribe medication, they only consider the numerical values ​​of various dimensions of the patient before medication, without taking into account the differences in the physical conditions of different patients. This means that the patient's medication plan may not be the optimal one. Therefore, obtaining the patient's medication plan is crucial.

[0050] It should be further noted that patients with rheumatic eye involvement and similar physical conditions tend to experience similar changes in their physical condition after medication. Therefore, patients are clustered based on changes in various dimensions of data before and after medication, grouping those with similar physical conditions into one category. Then, a common medication regimen is derived for all patients within that cluster based on their respective medication regimens.

[0051] It should be further noted that when obtaining the shared medication regimens for each patient within each cluster, the more frequently a drug appears in the medication regimens of patients within a cluster, the greater the likelihood that the drug is a shared medication regimen for patients within that cluster. Therefore, the probability that each drug is a shared medication regimen for patients in each cluster is calculated based on the frequency of each drug's appearance in the medication regimens of patients within each cluster.

[0052] It should be further explained that when obtaining the common medication regimens for patients in each cluster, since the physical changes of patients in each cluster before and after medication vary, and these changes are influenced by the drugs in the medication regimen, the similarity of the medication regimens of two patients is obtained based on the similarity of their physical changes before and after medication with those of other patients. The stronger the similarity between a patient's medication regimen and the medication regimens of other patients in a cluster, the more likely that the drugs in that patient's regimen are common drugs for all patients in that cluster. Therefore, based on the similarity between the medication regimens of each patient in each cluster and the medication regimens of other patients, the medication regimen of each patient in each cluster is obtained as the representative medication regimen for all patients in that cluster, i.e., the representativeness of each patient's medication regimen.

[0053] It should be further noted that the greater the similarity between the medication regimens of two patients, the greater the likelihood that their regimens are interchangeable. Therefore, based on the representativeness of the medication regimens of patients within a cluster who use the same drug, and the similarity of these patients' regimens to those of other patients, we can determine the probability that a drug is a universal drug for patients within a cluster, thus obtaining the universal drug for each patient within each cluster. Then, based on the differences between the universal drug for each patient within each cluster and the drugs included in the medication regimens of each patient within that cluster, we can determine the membership degree of each patient to each cluster.

[0054] Specifically, the changes in each dimension of the patient's data before and after medication are used as data axes to obtain a multidimensional sample space. All patients are mapped into this multidimensional sample space, resulting in multiple data points within the multidimensional sample. The Euclidean distance between each data point and other data points is used as a distance metric, and the DBSCAN clustering algorithm is used to cluster the data points in the multidimensional sample space into multiple clusters. The method for obtaining the Euclidean distance between two data points and the DBSCAN clustering algorithm are both existing well-known technologies and will not be elaborated upon in this embodiment.

[0055] Furthermore, the first In the cluster, the th The patient corresponding to the data point is denoted as the [number]th data point. In the cluster, the th The first patient. Obtain the first... In the cluster, the th The patient and the first The specific formula for calculating the similarity of medication regimens among patients is as follows:

[0056] ;

[0057] In the formula, Indicates the first In the cluster, the th The patient and the first Similarity of medication regimens among patients; Indicates the first Standard values ​​for each dimension of data. Indicates the first All patients in each cluster before and after medication were on day 2. The mean of the absolute values ​​of the changes in data across each dimension Indicates the first In the cluster, the th The patients before and after medication were on the [number]th [day / month]. Changes in data across each dimension Indicates the first In the cluster, the th The patients before and after medication were on the [number]th [day / month]. Changes in data across each dimension Indicates the number of dimensions collected. Represents the absolute value function. To prevent hyperparameters with a denominator of 0, this embodiment sets... Other values ​​can be set in other implementations; This represents the weight normalization function, whose input is the weight of each dimension. ; This represents the sigmoid function, which is used for normalization in this embodiment.

[0058] It should be noted that, The larger the value, the more significant the [value]. The first patient within each cluster The data for one dimension showed significant changes before and after medication, meaning the drug had a positive effect on the [missing data]. The first patient within the cluster The impact of the first dimension is significant, therefore the impact on the second dimension is relatively large. The data in the first dimension is given a larger weight; when the first dimension... Within the _ cluster _ The patient and the first The first patient's When the changes in data across the first dimension before and after medication are quite similar, it indicates that the medication regimens of these two patients are effective in controlling the second dimension. The similarity in the degree of treatment across several dimensions resulted in a high degree of similarity in the medication regimens for these two patients; through This is to address the issue of varying magnitudes and ranges of change in data across different dimensions. Taking blood glucose and heart rate as examples, a normal range for blood glucose is 3.9-6.0 mmol / L, and a change of 0.5 mmol / L is considered relatively large. Similarly, a normal range for heart rate is 60-100 beats / minute, and a change of 1 beat / minute is considered relatively small. Therefore, by… To address the issue of different units and varying degrees of change in data from different dimensions.

[0059] This allows us to obtain the similarity of medication regimens for each patient in each cluster to that of other patients.

[0060] Furthermore, according to the first The specific formula for calculating the representativeness of the medication regimen for each patient in each cluster, based on the similarity of the medication regimens of each patient to other patients, is as follows:

[0061] ;

[0062] In the formula, Indicates the first In the cluster, the th The representativeness of the medication regimen for each patient Indicates the first The number of patients contained in each cluster Indicates the first In the cluster, the th The patient and the first The similarity of medication regimens among patients This represents the weight normalization function, whose input object is the first... The sum of the representativeness of the medication regimens of each patient in each cluster and other patients.

[0063] It should be noted that if the first In the cluster, the th The stronger the similarity between the medication regimen of a particular patient and the medication regimens of other patients in that cluster, the better the indication that the medication regimen of the first patient is similar to that of the second patient. In the cluster, the th The medication regimen of one patient is more representative of the medication regimens of all patients in that cluster. The larger the value, the better.

[0064] Furthermore, to obtain the first The drug in The specific formula for calculating the prevalence of medication regimens for each cluster is as follows:

[0065] ;

[0066] In the formula, Indicates the first The drug in Popularity of medication regimens in each category Indicates the first The number of patients contained in each cluster Indicates the first The medication regimen within each cluster contains the first... The number of patients for each drug Indicates the first The medication regimen in this cluster does not include the [number]. The number of patients for each drug Indicates the first The medication regimen within each cluster contains the first... The first drug The representativeness of the medication regimen for each patient Indicates the first The medication regimen within each cluster contains the first... The first drug The patient and those not included in the first The first drug Similarity of medication regimens among patients.

[0067] It should be noted that if the first The drug in The higher the frequency of occurrence in the medication regimen of a particular patient cluster, the more likely it is to be a positive result. The drug has the effect on the first The stronger the treatment capability of patients in each cluster, the more it illustrates the importance of the first cluster. The more likely the drug is to be the first Common medications for patients in each cluster; therefore, according to the... The first cluster contains the first The representativeness of the medication regimen for patients using this drug and excluding the first drug. The medication regimen for patients with this drug and the medication regimen containing the first drug are as follows: The similarity of the medication regimens of patients using the drug was obtained. The drug in Popularity of medication regimens in each category.

[0068] Furthermore, a preset popularity threshold is set. If the first The drug in Popularity of medication regimens in each cluster Then the first The drug is the first A universal drug for patients within a certain cluster was obtained. A universal drug for all patients within a specific patient cluster. This embodiment includes a preset accessibility threshold. This example is used to illustrate the concept; other values ​​can be set in other implementations.

[0069] Furthermore, based on each patient's medication regimen and the common medications used by all patients within each cluster, the specific formula for calculating the membership degree of each patient to each cluster is as follows:

[0070] ;

[0071] In the formula, Indicates the first The patient belongs to the first... Membership degree of each cluster, Indicates the first The common medication for all patients within each cluster and the first The number of drug types included in a patient's medication regimen. Indicates that they appear simultaneously in the first... The common medication for all patients within each cluster and the first The number of drug types in a patient's medication regimen.

[0072] It should be noted that when the first The medication regimen for the first patient is the same as that for the second patient. The more common types of drugs are included in the general medications for all patients within a cluster, the higher the likelihood of clustering from a drug-based perspective. The patient belongs to the first... The stronger the probability of the nth cluster, that is, the higher the probability of the nth cluster. The common drugs in the first cluster include the first The higher the likelihood of a drug being included in the optimal medication regimen for a patient.

[0073] At this point, we have obtained the universal medication for all patients within each cluster, as well as the membership degree of each patient to each cluster.

[0074] Step S003: Based on the differences in the changes of data in each dimension before and after medication for each patient and patients in each cluster, obtain the final physical differences of all patients in each cluster; combine the membership degree of each patient to each cluster to obtain the performance of the universal drug for all patients in each cluster as the best drug for each user.

[0075] It should be noted that since the membership degree of each patient to each cluster is calculated only based on the difference between the drugs in each patient's medication regimen and the common drugs of all patients in each cluster, the physical differences between each patient and all patients in each cluster are not considered. Therefore, the physical differences between each patient and all patients in each cluster are calculated.

[0076] It should be further noted that the greater the difference in physical condition between two patients, the greater the difference in the changes in data across various dimensions before and after medication. Therefore, based on the changes in data across various dimensions before and after medication for different patients, the final physical difference between each patient and all patients within a cluster is obtained. Based on the final physical difference between each patient and all patients within a cluster, and the membership degree of each patient to each cluster, the performance of the universal drug for all patients within each cluster as the optimal drug for each user is obtained, thus determining the drugs in each user's personalized medication plan.

[0077] Specifically, based on the changes in data across various dimensions before and after medication for each patient and all patients within each cluster, the relationship between each patient and the first... The specific formula for calculating the final physical differences among all patients within a cluster is as follows:

[0078] ;

[0079] In the formula, Indicates the first The patient and the first The final physical differences among all patients within each cluster Indicates the number of dimensions collected. Indicates the first The patients before and after medication were on the [number]th [day / month]. Changes in data across each dimension Indicates the first All patients in each cluster before and after medication were on day 2. The mean of the absolute values ​​of the changes in data across each dimension Indicates the first All patients in each cluster before and after medication were on day 2. The mean of the changes in the data across each dimension Indicates the first Standard values ​​for each dimension of data. Represents the absolute value function. This represents the weight normalization function, whose input is the weight of each dimension. .

[0080] It should be noted that when the first The patient and the first The first patient within each cluster The more similar the changes in each dimension of the data before and after medication, the more likely it is that the changes in the first dimension are similar. The patient and the first Patients within each cluster in the first The more similar the differences in physical fitness across all dimensions, the better. The smaller the value, the more it indicates that the... The patient and the first The more similar the physical conditions of patients within a cluster; It is used to avoid the problem of different fluctuation ranges in data of different dimensions; when the first The patient and the first The first patient within each cluster The greater the change in the data of each dimension before and after medication, the more likely it is that the data of the first dimension has changed significantly. The patient and the first The first patient within each cluster The greater the impact of drugs on changes in each dimension of the data, therefore, according to Give Weights.

[0081] This yields the final physical differences between each patient and all patients within each cluster.

[0082] Furthermore, to obtain the first The universal medication for all patients within each cluster is the [number]. The specific formula for calculating the optimal drug performance for each user is as follows:

[0083] ;

[0084] In the formula, Indicates the first The universal medication for all patients within each cluster is the [number]. The best drug performance for each user Indicates the first The patient and the first The final physical differences among all patients within each cluster This indicates that the s-th patient belongs to the s-th patient. Membership degree of each cluster, To prevent hyperparameters with a denominator of 0, this embodiment sets... In other implementations, it can be set to other values.

[0085] It should be noted that when the first The patient and the first The greater the difference in physical condition among patients within a cluster, the more it indicates that the... The medications in the individualized medication regimen for each patient are the same as those in the first patient. The greater the difference in the universal medication among all patients within a cluster, i.e., the greater the difference in the number of clusters, the more likely it is to be included in the general list. The universal medication for all patients within each cluster is the [number]. The smaller the performance of the best drug for a user.

[0086] Step S004: Based on the performance of the same universal drug for all patients within all clusters as the best drug for a user, obtain the matching degree of each drug as the best drug for each patient, and then generate and recommend a personalized medication plan for each patient.

[0087] Specifically, the formula for calculating the optimal drug match for each patient is as follows:

[0088] ;

[0089] In the formula, Indicates the first The drug is the first The optimal drug match for each patient This indicates that the generic medication for all patients contains the first... The number of drug clusters, This indicates that the generic medication for all patients contains the first... The first drug The universal medication for all patients within each cluster is the [number]. The best drug performance for each user This represents the sigmoid function, which is used for normalization in this embodiment.

[0090] Furthermore, a preset probability threshold is set. ,like Then the first The drug is the first The best medication for each patient. (The first...) All the best medications for the patient as the first The medications in the personalized medication plan for each patient are generated. Personalized medication plans for each patient are then recommended to the physician. In this embodiment, a preset probability threshold is used. This example is used to illustrate the concept; other values ​​can be set in other implementations.

[0091] This concludes the embodiment.

[0092] In summary, the above descriptions in the embodiments of the present invention are merely preferred embodiments of the present invention and are 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 method for generating an individualized medication regimen for patients with rheumatic eye involvement, characterized in that, The method includes the following steps: Obtain each patient's medication regimen and data on each patient before and after medication, as well as the standard values ​​for each dimension of data; Based on the changes in data across various dimensions before and after medication for each patient, patients are clustered into multiple groups. The similarity of medication regimens among different patients within the same cluster is determined by the differences in these changes. Specific methods include: ； In the formula, Indicates the first In the cluster, the th The patient and the first Similarity of medication regimens among patients; Indicates the first Standard values ​​for each dimension of data. Indicates the first All patients in each cluster before and after medication were on day 2. The mean of the absolute values ​​of the changes in data across each dimension Indicates the first In the cluster, the th The patients before and after medication on the [number]th Changes in data across each dimension Indicates the first In the cluster, the th The patients before and after medication on the [number]th Changes in data across each dimension Indicates the number of dimensions collected. Represents the absolute value function. To prevent hyperparameters with a denominator of 0, This represents the weight normalization function. This represents the sigmoid function; based on the distribution of drugs in the medication regimens of each patient in each cluster and the similarity of medication regimens among different patients in the same cluster, the prevalence of each drug in the medication regimens of each patient in each cluster is obtained, and thus the universal drugs for all patients in each cluster are obtained; based on the difference between the universal drugs for all patients in each cluster and the drugs included in the medication regimens of each patient, the membership degree of each patient to each cluster is obtained. Based on the differences in the changes of data in each dimension before and after medication for each patient and patients in each cluster, the final physical differences of all patients in each cluster are obtained; combined with the membership degree of each patient to each cluster, the performance of the universal drug for all patients in each cluster as the best drug for each user is obtained. Based on the performance of the same universal drug for all patients within all clusters as the best drug for a user, the matching degree of each drug as the best drug for each patient is obtained, and then a personalized medication plan for each patient is generated and recommended.

2. The method for generating an individualized medication regimen for patients with rheumatic eye involvement according to claim 1, characterized in that, The method for determining the prevalence of each drug in each cluster based on the distribution of drugs in the medication regimens of each patient within each cluster and the similarity of medication regimens among different patients within the same cluster includes the following specific methods: According to the The specific formula for calculating the representativeness of the medication regimen for each patient in each cluster, based on the similarity of the medication regimens of each patient to other patients, is as follows: ； In the formula, Indicates the first In the cluster, the th The representativeness of the medication regimen for each patient Indicates the first The number of patients contained in each cluster Indicates the first In the cluster, the th The patient and the first The similarity of medication regimens among patients This represents the weight normalization function; Based on the representativeness of each medication regimen in each cluster and the distribution of drugs in the medication regimen of each patient in each cluster, the prevalence of each drug in the medication regimen of each cluster is obtained.

3. The method for generating an individualized medication regimen for patients with rheumatic eye involvement according to claim 2, characterized in that, The method for determining the prevalence of each drug in each cluster based on the representativeness of each medication regimen within each cluster and the distribution of drugs in the medication regimen for each patient within each cluster includes the following specific methods: ； In the formula, Indicates the first The drug in Popularity of medication regimens in each category Indicates the first The number of patients contained in each cluster Indicates the first The medication regimen within each cluster contains the first... The number of patients for each drug Indicates the first The medication regimen in this cluster does not include the [number]. The number of patients for each drug Indicates the first The medication regimen within each cluster contains the first... The first drug The representativeness of the medication regimen for each patient Indicates the first The medication regimen within each cluster contains the first... The first drug The patient and those not included in the first The first drug Similarity of medication regimens among patients.

4. The method for generating an individualized medication regimen for patients with rheumatic eye involvement according to claim 1, characterized in that, The specific method for obtaining the universal drug for all patients within each cluster is as follows: Preset popularity threshold If the first The drug in Popularity of medication regimens in each cluster Then the first The drug is the first A universal drug for patients within a certain cluster was obtained. A universal drug for all patients within a specific cluster.

5. The method for generating an individualized medication regimen for patients with rheumatic eye involvement according to claim 1, characterized in that, The method for determining the membership degree of each patient to each cluster based on the differences between the common medications for all patients within each cluster and the medications included in each patient's medication regimen includes: ； In the formula, Indicates the first The patient belongs to the first... Membership degree of each cluster, Indicates the first The common medication for all patients within each cluster and the first The number of drug types included in a patient's medication regimen. Indicates that they appear simultaneously in the first... The common medication for all patients within each cluster and the first The number of drug types in a patient's medication regimen.

6. The method for generating an individualized medication regimen for patients with rheumatic eye involvement according to claim 1, characterized in that, The method for obtaining the final physical differences of all patients within each cluster based on the differences in data changes across various dimensions before and after medication for each patient and patients within each cluster includes the following specific methods: ； In the formula, Indicates the first The patient and the first The final physical differences among all patients within each cluster Indicates the number of dimensions collected. Indicates the first The patients before and after medication on the [number]th Changes in data across each dimension Indicates the first All patients in each cluster before and after medication were on day 2. The mean of the absolute values ​​of the changes in data across each dimension Indicates the first All patients in each cluster before and after medication were on day 2. The mean of the changes in the data across each dimension Indicates the first Standard values ​​for each dimension of data. Represents the absolute value function. This represents the weight normalization function.

7. The method for generating an individualized medication regimen for patients with rheumatic eye involvement according to claim 1, characterized in that, The specific method for obtaining the performance of the universal drug for all patients within each cluster as the best drug for each user includes: ； In the formula, Indicates the first The universal medication for all patients within each cluster is the [number]. The best drug performance for each user Indicates the first The patient and the first The final physical differences among all patients within each cluster This indicates that the s-th patient belongs to the s-th patient. Membership degree of each cluster, To prevent hyperparameters with a denominator of 0.

8. The method for generating an individualized medication regimen for patients with rheumatic eye involvement according to claim 1, characterized in that, The method for determining the best drug match for each patient based on the performance of the same universal drug for all patients within all clusters as the best drug for a user includes: ； In the formula, Indicates the first The drug is the first The optimal drug match for each patient This indicates that the generic medication for all patients contains the first... The number of drug clusters, This indicates that the generic medication for all patients contains the first... The first drug The universal medication for all patients within each cluster is the [number]. The best drug performance for each user This represents the sigmoid function.

9. The method for generating an individualized medication regimen for patients with rheumatic eye involvement according to claim 1, characterized in that, The specific methods for generating personalized medication plans for each patient are as follows: Preset probability threshold ,like Then the first The drug is the first The best medicine for each patient; The first All the best medications for the first patient as the first The medications in the personalized medication plan for each patient are generated. Personalized medication plans for each patient.

Images