A data analysis and management system for the diagnosis of mental diseases

Abstract

The present invention relates to the technical field of medical information processing, and specifically, to a data analysis and management system for mental disease diagnosis materials, including: a label setting module, which is used to obtain the disease diagnosis data of patients, preliminarily classify the disease diagnosis data, and set data labels for the disease diagnosis data of each classification; a verification label module, which is used to check the missing situation of the data labels under each classification, identify and record the missing information as the target data for inspection; a symptom matching module, which is used to match the target data with the symptom model, obtain the disease score value corresponding to the target data, and determine the treatment plan for the patients under each classification according to the disease score value; analyze the coverage and missing ratio of the treatment plan under each classification to obtain the symptom matching result; realizing the scientific nature of the treatment plan and the rationality of resource allocation, and improving the overall quality of medical services.

Description

Technical Field

[0001] The present invention relates to the technical field of medical information processing, and specifically, to a data analysis and management system for mental disease diagnosis materials. Background Art

[0002] With the rapid development of mobile technology, electronic health records have gradually been combined with mobile devices. Patients can use mobile applications to access and update their health records, and doctors can view and manage patient information through mobile devices. However, the phenomenon often occurs that the electronic health records are missing information, resulting in doctors being unable to detect the changes in the patient's condition in a timely manner, especially the diagnostic data related to mental diseases; this part of the data is prone to missing labels during normal management, so it is necessary to adjust the processing of mental disease diagnostic data.

[0003] For example, Chinese Patent Publication No. CN116884592A discloses a method and system for screening medical information based on big data analysis. By online inputting the patient's visit time period and the patient's disease symptom data, the department type and the disease symptom level of the patient's disease symptoms are identified and matched according to the cooperation of the patient's disease symptom data and the hospital disease department type data, and the doctor data of the department corresponding to the professional level of patients with different disease symptom levels is output to finally form the patient's appointment registration and treatment order data, so as to realize the scientific and accurate matching of the patient's disease symptoms and the doctors of the department corresponding to the required professional level; the difference degree between the patient's appointment registration and treatment order data and the patient's disease diagnosis data input by the doctor is calculated, and the trust degree of the patient's appointment registration is analyzed.

[0004] However, when the prior art adjusts the appointment matching of different symptoms and corresponding departments, it is necessary to adjust the specific processing method according to the different labels corresponding to different patients, so that when the doctor diagnoses and treats the patient, a sufficiently comprehensive treatment plan can be obtained to improve the treatment effect of the patient. Summary of the Invention

[0005] In order to solve the above technical problems, the technical solution adopted by the present invention is: a data analysis and management system for mental disease diagnosis materials, including: a label setting module, which is used to obtain the disease diagnosis data of patients, and conduct a preliminary classification on the disease diagnosis data, and set data labels for the disease diagnosis data of each classification.

[0006] A verification label module, which is used to check the missing situation of the data labels under each classification, identify and record the missing information as the target data for inspection.

[0007] A symptom matching module is used to match the target data with a symptom model, obtain the disease score value corresponding to the target data, and determine the treatment plan for the patient in each classification according to the disease score value; analyze the coverage and missing ratio of the treatment plan in each classification to obtain the symptom matching result.

[0008] A resource allocation module is used to obtain the diagnosis matrix and symptom demand distribution of the current patient according to the symptom matching result; determine the treatment cycle and resource allocation of the current patient according to the diagnosis matrix and symptom demand distribution of the current patient, and output the corresponding treatment cycle and resource allocation as a resource allocation strategy.

[0009] An adjustment and optimization module is used to verify the qualification of the resource allocation strategy. When the resource allocation strategy is qualified, obtain the degree evaluation value related to the qualification. When the resource allocation strategy is unqualified, obtain the adjustment factor of the resource allocation strategy, and optimize the resource allocation strategy according to the adjustment factor of the resource allocation strategy.

[0010] The beneficial effects of the present invention are as follows: First, by obtaining the disease diagnosis data of the patient and performing preliminary classification, and setting labels for the data in each classification, the standardization and traceability of the data are ensured.

[0011] Second, the present invention matches the target data with the symptom model to obtain the disease score value, thereby determining the treatment plan for the patient in each classification. By analyzing the coverage and missing ratio of the treatment plan, the symptom matching result is obtained to ensure the scientificity and rationality of the treatment plan.

[0012] Third, according to the symptom matching result, the present invention obtains the diagnosis matrix and symptom demand distribution of the current patient, and determines the treatment cycle and resource allocation. The output resource allocation strategy can allocate medical resources more reasonably and improve the resource utilization efficiency.

[0013] Fourth, the present invention verifies the qualification of the resource allocation strategy. When it is unqualified, calculates the adjustment factor and optimizes the resource allocation strategy. Through the degree evaluation value and the adjustment factor, the resource allocation strategy can be dynamically adjusted to ensure its continuous optimization. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] The present invention will be further described below with reference to the drawings and embodiments.

[0015] Figure 1 It is a system framework diagram of a mental disease diagnosis data analysis and management system.

[0016] Figure 2 It is a system flow chart of a mental disease diagnosis data analysis and management system.

[0017] Figure 3It is a schematic flow diagram of the verification label module of a data analysis and management system for mental illness diagnosis data.

[0018] Figure 4 It is a schematic flow diagram of the matching process of the symptom matching module of a data analysis and management system for mental illness diagnosis data.

[0019] Figure 5 It is a schematic flow diagram of the similar retrieval process of the symptom matching module of a data analysis and management system for mental illness diagnosis data.

[0020] Figure 6 It is a schematic flow diagram of the disease score value of the symptom matching module of a data analysis and management system for mental illness diagnosis data. Detailed implementation manners

[0021] The embodiments of the present invention will be described in detail below. The embodiments described below are exemplary and are only used to explain the present invention, and should not be construed as a limitation to the present invention. For those not specified in the embodiments, the techniques or conditions described in the literature in the art or according to the product specifications are followed.

[0022] Refer to Figure 1 , a data analysis and management system for mental illness diagnosis data, including: a label setting module, a verification label module, a symptom matching module, a resource allocation module, and an adjustment and optimization module; the label setting module transmits the data with set labels to the verification label module; the verification label module transmits the corresponding verified data to the symptom matching module; after the symptom matching module completes the matching, it sends the corresponding data to the resource allocation module; the resource allocation module then determines the medical resources set at this time and transmits the corresponding data to the adjustment and optimization module. After the adjustment and optimization module completes the verification, it adjusts the corresponding medical resources to complete the allocation of resources.

[0023] In the present invention, as Figure 2 shown, the specific implementation manners of each module are as follows.

[0024] The label setting module is used to obtain the disease diagnosis data of the patient, perform a preliminary classification on the disease diagnosis data, and set data labels for the disease diagnosis data of each classification.

[0025] The verification label module is used to check the missing situation of the data labels under each classification, identify and record the missing information as the target data for inspection.

[0026] The symptom matching module is used to match the target data with the symptom model, obtain the disease score value corresponding to the target data, determine the treatment plan for the patient under each classification according to the disease score value; analyze the coverage and missing ratio of the treatment plan under each classification to obtain the symptom matching result.

[0027] A resource allocation module, which is used to obtain the diagnosis matrix and symptom demand distribution of the current patient according to the symptom matching result; determine the treatment cycle and resource allocation of the current patient according to the diagnosis matrix and symptom demand distribution of the current patient, and output the corresponding treatment cycle and resource allocation as a resource allocation strategy.

[0028] An adjustment and optimization module, which is used to verify the qualification of the resource allocation strategy. When the resource allocation strategy is qualified, obtain the degree evaluation value related to the qualification. When the resource allocation strategy is unqualified, obtain the adjustment factor of the resource allocation strategy, and optimize the resource allocation strategy according to the adjustment factor of the resource allocation strategy.

[0029] Judge the classification of the disease and corresponding screening through the obtained data. Use various collected data (such as medical history, drug use records, etc.) to accurately classify mental diseases and screen out cases that meet specific conditions, so as to facilitate the formulation of subsequent personalized treatment plans.

[0030] The disease diagnosis data used at this time includes medical history, drug use, treatment records, and physical data. Through these data, the classification and screening of the identified mental diseases are completed to facilitate the adjustment of the treatment plan and the corresponding process configuration.

[0031] The above-mentioned medical history includes the patient's mental disease history, family medical history, and other relevant health conditions, which are used to determine whether the patient has a relevant medical history and whether there are problems with the body.

[0032] The above-mentioned drug use is used to record the drug information used by the patient in the past and present, including dosage, medication time, etc.; if there is no drug use record for the patient at this time, the flag is empty, so as to identify whether the patient is using drug treatment for the first time or multiple times.

[0033] The above-mentioned treatment records are used to record the psychological or physical treatment methods received by the patient in the past and their effects. The effects recorded here are mostly expressed as the doctor's evaluation of the patient's state after treatment, or the data changes in the scales or other scoring forms between every two treatments, so as to quantify the changes in the patient's own treatment and track the specific situation of the patient.

[0034] The above-mentioned physical data includes physiological indicators such as physical examination results and laboratory test reports, such as height, weight, and the situation of corresponding genes or chromosomes under specific disease records. At this time, genes and chromosomes are mainly used to observe whether the current patient has a hereditary mental disease, and compare the patient's own chromosomes with the chromosomes of the inheritable unit to determine whether there are changes.

[0035] In the present invention, when initially classifying disease diagnosis data and setting data labels for the disease diagnosis data of each classification, first, patients can be classified into anxiety disorders, mood disorders (such as depression), schizophrenia spectrum and other psychotic disorders, etc. according to a standard medical classification system, such as the mental disorder classification in ICD-10-CM or DSM-5.

[0036] At the same time, classification can also be carried out according to the main symptoms found in the disease diagnosis data, such as anxiety symptom clusters, depressive symptom clusters, cognitive dysfunction, etc.; this way can quickly identify the main problems of the patients for a preliminary assessment of the patients.

[0037] At this time, it is mainly to verify whether the classification can fully represent the corresponding symptoms for each patient, and adjust the classification for patients with unclear classification or symptoms, and whether there is a tendency for the patients' data to be missing under the corresponding classification. This missing situation is more inclined to incomplete data, resulting in missing and incomplete data after classification. The data with such missing parts is regarded as the target data to be tested at this time.

[0038] For example, when initially classifying in the present invention, it is divided into multiple categories according to the symptoms recorded in the disease diagnosis data, and data labels are set for the disease diagnosis data of the current patient according to the medical history, drug use, treatment records, and physical data existing in the disease diagnosis data.

[0039] Suppose a patient is initially classified as "severe depression", then the data labels may include: disease category: severe depression; symptom severity: severe; treatment status: receiving treatment; drug use: sertraline 50mg / day; type of psychotherapy: cognitive behavioral therapy; past medical history: having a history of mild depression; family medical history: mother has depression; social environmental factors: unemployed, tense family relationship; physical examination results: normal thyroid function; follow-up record: psychological counseling every two weeks, drug efficacy evaluation once a month.

[0040] According to these set labels, the data diagnosed for the current patient can be managed, so as to facilitate subsequent monitoring of the corresponding conditions of the patient after multiple treatment processes, verify whether the information obtained under the initial classification meets the expectation, and whether the data obtained under this expectation can represent the integrity of the current patient's data.

[0041] In the present invention, the verification label module mainly examines the data of the data labels under each classification, verifies the data ratio, corresponding data values and the logic before and after the data of the data labels to determine whether there is a missing situation at this time, and marks the data that may be identified as missing to obtain the target data that needs to be focused on processing.

[0042] For example, such asFigure 3 As shown, the implementation of checking for missing data tags under each classification can be as follows.

[0043] Determine the data ratio of the disease diagnosis data under each classification as the first ratio.

[0044] Extract comparison data from the disease diagnosis data under each classification, and obtain the data ratio of the comparison data under each classification as the second ratio.

[0045] According to the obtained first ratio and second ratio, calculate the missing rate of the corresponding fields of the disease diagnosis data under each classification.

[0046] Perform secondary classification on the disease diagnosis data after the initial classification according to the obtained missing rate, and output the part of the disease diagnosis data after the secondary classification where the data value is greater than the preset standard compared with the predicted value as the target data.

[0047] The first ratio mentioned above refers to the data volume ratio value corresponding to each classification of the disease diagnosis data after the initial classification. This value represents the ratio of this classification in the overall data. When calculating this ratio, the data will be verified first to determine the null data and duplicate data in the data, and this part of the data will be compared with the overall data. At this time, the ratio should be the same as the expected divided value. If it is different, it means that there are corresponding problems in the divided data at this time. For example, if the ratio value is large, it means that there is some unprocessed data, and if the ratio value is small, it means that some normal data has been processed. At this time, it is necessary to find the part of the data with misidentification.

[0048] The above comparison data is mainly used to supplement the corresponding content in the current disease diagnosis data through the content recorded by the corresponding doctor and the symptom description to generate comparison data, and determine the data ratio that needs to be supplemented at this time according to the ratio of this part of the corresponding data to the overall data to compare whether there are corresponding problems. The second ratio is the ratio value of these comparison data to the total data under each classification. For example, select a set of keyword fields (such as drug use, treatment history, etc.) as comparison data, calculate the number of records of these comparison data under each classification, and obtain their ratio in this classification.

[0049] When calculating the missing rate of the corresponding fields of disease diagnosis data under each classification based on the obtained first ratio and second ratio, the classification corresponding to the disease diagnosis data at this time and the corresponding data under the classification are mainly selected through the data corresponding to the first ratio and the second ratio. For each classification, the number of missing values of each field is calculated, and the missing rate is calculated using the number of missing values and the total amount of data in the classification; the missing rate is obtained by traversing the fields of the disease diagnosis data corresponding to the first ratio and the second ratio. If a field is missing, it is recorded as 1, and if it is not missing, it is recorded as 0. After counting all fields, the value obtained at this time is divided by the total number of fields to obtain the missing rate.

[0050] At this time, the preset standard is expressed as the missing rate exceeding a certain threshold. When the threshold is exceeded, the target data identified can clearly represent the corresponding data situation, so this part of the obviously missing data is selected as the target data.

[0051] When performing secondary classification, it is mainly based on the obtained missing rate to classify the corresponding disease diagnosis data. Each value of the missing rate represents a classification, and multiple categories are obtained in sequence.

[0052] In the present invention, as Figure 4 shown, the implementation method of matching the target data with the symptom model is as follows: preprocess the target data, extract the symptom features of the target data, perform a similarity search on the symptom features and the symptom model to obtain a similarity search result; the preprocessing method for the target data can be to impute the missing values in the target data by means of mean imputation; mean imputation is to supplement the missing part with the average value of the corresponding data to make the overall data more balanced.

[0053] Retrieve from the database according to the main symptoms of the current patient, and select the symptom model corresponding to the main symptoms. The symptom model is a model based on different symptom components in advance, and each symptom model contains a description of the symptoms and a treatment plan for the symptoms.

[0054] As Figure 5As shown in the figure, the method for performing similarity retrieval is as follows: Obtain the first keyword set that the symptom characteristics are expected to view and the second keyword set composed of related words of the first keyword set; the first keyword set is used to select the words in the symptom characteristics that are expected to be searched through this vocabulary to determine the symptom model corresponding to the current patient. For example, when querying mania, the vocabulary that is expected to be viewed at this time is mania. The first keyword set mainly consists of these words directly related to the symptoms. At this time, it is expected to complete the retrieval of the symptom model by searching for these words; the second keyword set is mainly composed of words related to the first keyword set. For example, when querying mania, words such as mood changes can be used as the query vocabulary to find related symptoms; the second keyword set can query more symptoms during the query, and a more accurate similarity retrieval result can be obtained after comparing the first keyword set with the second keyword set.

[0055] Calculate the matching degree between the first keyword set, the second keyword set and each symptom in the symptom model in turn; at this time, the calculation method of the matching degree is to divide the intersection of the first keyword set and each symptom in the symptom model by the union of the first keyword set and each symptom in the symptom model to obtain the matching degree between the first keyword set and each symptom in the symptom model. The calculation method for the second keyword set is the same as that for the first keyword.

[0056] Perform weighted summation of the matching degrees of the first keyword set and the second keyword set in turn according to each symptom in the symptom model to obtain the comprehensive matching degree; when the comprehensive matching degree is greater than the preset threshold, the symptoms and treatment plans corresponding to the symptom model are used as the similarity retrieval results.

[0057] At this time, the weighted summation of the matching degrees is to calculate the comprehensive matching degree by weighting each symptom separately to find the calculation result required for the similarity retrieval at this time; the output similarity retrieval result is a list of symptoms and treatment plans related to the current patient selected according to the matching degree. At this time, it only needs to be judged that the search can reach the minimum content, and the preset threshold is set to 0.55.

[0058] When the symptom characteristics are verified to be accurate, calculate the score of the similarity retrieval result to obtain the disease score value; use the treatment plan corresponding to the maximum value of the disease score value in each classification as the output treatment plan.

[0059] The method for verifying the accuracy of the symptom characteristics is to compare the vocabulary of the symptom characteristics with the pre-set vocabulary. If the vocabulary describing the symptom characteristics does not exist in the pre-set vocabulary, it is considered that the symptom characteristics are inaccurate, and the score of the corresponding similarity retrieval result is not calculated.

[0060] Such as Figure 6As shown in the figure, the method for calculating the scores of similar search results is as follows: Obtain the standard symptom list of the symptom model and determine the initial score of each symptom in the standard symptom list.

[0061] Find the symptoms output from the similar retrieval results and the corresponding initial scores, adjust the incentives according to the patients corresponding to the symptoms, and output the scores after incentive adjustment as the disease score values.

[0062] At this time, the symptom model set is to define the standard symptom list of each mental illness according to medical knowledge and clinical guidelines; for example, the symptoms of severe depression may include persistent sad mood, loss of interest, sleep disorders, etc.; a weight or score is set for each symptom in the standard symptom list to facilitate subsequent quantification of the corresponding symptoms existing in the target data.

[0063] The corresponding incentive adjustment method is to obtain the feedback score and reputation score of the patient, and adjust the initial score through the feedback score and reputation score to obtain the disease score value.

[0064] The method for obtaining the feedback score is as follows.

[0065] DFS i = f i a i Q - x i p i ; where DFS i represents the feedback score at the i-th feedback, a i represents the degree of need of the patient for symptom query at the i-th feedback. The degree of need represents the priority or weight of the corresponding symptom query at this time. The degree of need at this time is expressed as a value in the range of 0 to 1, which is used to quantify the weight of the corresponding symptom query at each patient feedback; x i represents the index value of the data volume of the patient's symptom query at the i-th feedback. At this time, the data volume of the symptom query represents the ratio of the number of queries for a specific symptom to the number of queries traversed during normal queries, which is used to determine the number of times spent on querying a specific symptom at this time; f i represents the incentive amount for data query at the i-th feedback. The incentive amount at this time is set to the same value in the range of 0 to 1 as the degree of need, which is mainly used to verify the cost during query. When the symptom to be verified is queried, a corresponding value will be given to adjust the overall setting value of the feedback score; Q represents the index value of the total demand for data query by the patient during feedback. The total demand here represents the ratio of the total data volume of multiple symptoms queried during query to the standard data volume; p i represents the index value of the cost amount of data query at the i-th feedback. The cost amount at this time mainly represents the ratio sum of the time taken to complete the retrieval, the system memory, the preset time, and the preset system memory.

[0066] The reputation score is obtained by getting the feedback score at each feedback and the time difference between each feedback, and the reputation score dynamically adjusts the feedback score by considering the factor of time to determine the change of the feedback score.

[0067] Among them, RI represents the reputation score, and DFS i represents the feedback score at the i-th feedback, e represents the exponential constant, λ represents the decay factor used to control the speed of time decay, and t i represents the time difference between the time of the i-th feedback and the current time, n represents the number of feedbacks, and i = 1, 2,..., n.

[0068] The reputation score is mainly used to verify whether the way of obtaining the feedback score is reliable when verifying the overall feedback, and whether the data retrieved each time meets the standards.

[0069] Take the product of the feedback score, reputation score, and initial score under each symptom as the disease score value; at this time, there will be multiple disease score values according to the number of feedbacks and different retrieved symptoms. Output the part with the largest disease score value at this time as the main symptom identified at this time, and use the treatment plan for this symptom as the treatment plan for each patient under each classification at this time.

[0070] Analyze the coverage and missing ratio of the treatment plan under each classification to obtain the implementation method of the symptom matching result. Obtain the total number of symptoms corresponding to the disease diagnosis data under each classification, the number of symptoms covered by the treatment plan under each classification, and the number of symptoms not covered by the treatment plan. The coverage of the treatment plan is expressed as the ratio of the number of symptoms covered by the treatment plan to the total number of symptoms, and the missing ratio is expressed as the ratio of the number of symptoms not covered to the total number of symptoms. Take the coverage and missing ratio identified at this time as the output symptom matching result.

[0071] The symptom matching result is mainly to understand the coverage of the treatment plan under each plan after selecting the corresponding treatment plan, so as to better perform symptom matching and resource allocation.

[0072] In the present invention, the diagnosis matrix is a matrix representing the symptoms of the patient. Each row represents a symptom, and each column represents an examination item. At this time, the symptoms in the symptom assignment result are extracted to obtain the examination items and corresponding symptoms corresponding to the patient during diagnosis.

[0073] For example, the diagnosis matrix can be represented as follows.

[0074] At this time, the diagnosis matrix is a matrix of m×k. For the diagnosis matrix, there is also a symptom demand distribution, and the symptom demand distribution at this time is represented as a vector as follows.

[0075] The number of symptom demand distributions is the same as the number of rows of the diagnosis matrix.

[0076] When the diagnosis matrix of the current patient in the resource allocation module is compared with the symptom demand distribution at this time, it is also necessary to verify whether the diagnosis matrix and the symptom demand distribution meet the requirements, and use the diagnosis matrix and symptom demand distribution that meet the requirements as the basis for selecting the resource allocation strategy.

[0077] The implementation method for verifying the diagnosis matrix and the symptom demand distribution is as follows.

[0078] Among them, e represents the exponential constant, d1 represents the first indicator function when the symptom occurs; d2 represents the second indicator function when the symptom occurs; s j represents the value of the symptom demand distribution of the j-th symptom, d1 j represents the value of the first indicator function related to the j-th symptom, d z represents the value of the diagnosis matrix related to the z-th examination item, dj,z represents the value of the diagnosis matrix related to the j-th symptom and the z-th examination item, d2 j represents the value of the second indicator function related to the j-th symptom, j = 1, 2,..., m, z = 1, 2,..., k; the reason for using commas in the above formula is to compare the first half and the second half according to the equation. At this time, the results finally calculated by the two parts should be the same to verify whether the matrix set at this time is reasonable.

[0079] The first indicator function when the symptom occurs can be expressed as: For the first indicator function, it is inclined to calculate the sum of products of all elements in the diagnosis matrix and the symptom demand distribution; then d1 j is the sum of products corresponding to the selected value of j.

[0080] The second indicator function when the symptom occurs can be expressed as: Among them, d m-j+1，k-z+1 represents the value of the diagnosis matrix under the (m - j + 1)-th symptom and the (k - z + 1)-th examination item; at this time, the second indicator function is used to obtain the sum of products of the symptom demand distribution and all elements in the diagnosis matrix after the diagnosis matrix is symmetrically changed according to the main diagonal. At this time, it is mainly to verify the rationality of the values set in the diagnosis matrix to prevent inconsistent values from occurring; the above d2 j The calculation result of is the sum of products corresponding to the selected j.

[0081] The value of the diagnosis matrix related to the z-th examination item can be expressed as At this time, the value of the corresponding diagnosis matrix in the z-th examination item is calculated, that is, the value of the z-th column in the diagnosis matrix.

[0082] When determining the treatment cycle and resource allocation for the current patient, obtain the standard treatment duration, which is represented as a vector at this time, and the size of the vector is the same as the number of rows of the diagnostic matrix. Obtain the treatment cycle based on the standard treatment duration, symptom demand distribution, and diagnostic matrix.

[0083] The calculation method of the treatment cycle is shown as follows.

[0084] Among them, T represents the treatment cycle, m represents the number of symptoms, j = 1, 2,..., m, k represents the number of examination items, z = 1, 2,..., k, s j represents the value of the symptom demand distribution of the j-th symptom, duration j represents the value of the standard treatment duration corresponding to the j-th symptom, d j，z represents the value of the diagnostic matrix related to the j-th symptom and the z-th examination item, w z represents the weight of the z-th examination item, e represents the exponential constant, min(w z ) represents the minimum value of the weights of the examination items, max(w z ) represents the maximum value of the weights of the examination items.

[0085] The implementation method for resource allocation is as follows. Obtain the standard resource allocation factor, and the resource allocation factor is the same as the number of rows of the diagnostic matrix. Obtain the resource allocation based on the resource allocation factor and the treatment cycle.

[0086] R j = s j ×T×SA j ; Among them, R j represents the value of the resource allocation for the j-th symptom, s j represents the value of the symptom demand distribution of the j-th symptom, T represents the treatment cycle, SA i represents the value of the resource allocation factor for the j-th symptom; Output the treatment cycle and resource allocation at this time as the resource allocation strategy.

[0087] In the present invention, the degree evaluation value in the adjustment and optimization module can be represented as follows.

[0088] Among them, G represents the degree evaluation value, T represents the treatment cycle, T' represents the standard value of the treatment cycle, R tota , represents the sum of the resource allocations, R' represents the standard value of the resource allocation, MS represents the missing ratio of the treatment plan, MS' represents the standard value of the missing ratio of the treatment plan, RS represents the disease score value, RS' represents the standard value of the disease score value, e represents the exponential constant.

[0089] At this time, a comprehensive evaluation will be carried out on the selected treatment plan and the configuration of this treatment plan to determine whether the currently selected plan is suitable for direct use; the degree evaluation value obtained at this time will be compared with the preset evaluation value, and whether the verification resource allocation strategy is qualified will be judged according to the difference between the degree evaluation value and the preset evaluation value. When it is unqualified, the degree evaluation value at this time will be adjusted, and the resource allocation strategy set for the previous processed data will be adjusted according to the adjusted value.

[0090] For example, the implementation method of the adjustment and optimization module also includes: when the difference between the degree evaluation value and the preset evaluation value is less than or equal to the first threshold, the resource allocation strategy is regarded as qualified and the corresponding resource allocation strategy is output; when the difference between the degree evaluation value and the preset evaluation value is greater than the first threshold, the first adjustment factor is used to adjust the resource allocation strategy; when the difference between the degree evaluation value and the preset evaluation value is greater than the first threshold and less than the second threshold, the second adjustment factor is used to adjust the resource allocation strategy; when the difference between the degree evaluation value and the preset evaluation value is greater than or equal to the second threshold, the resource allocation strategy is re-obtained until the difference between the degree evaluation value and the preset evaluation value is less than the second threshold; at this time, the first threshold is less than the second threshold.

[0091] The selected first adjustment factor can be expressed as follows.

[0092] Among them, λ1 represents the first adjustment factor, G represents the degree evaluation value; G0 represents the initial value of the degree evaluation value. At this time, the selected initial value is mainly used to adjust the setting of the adjustment factor to control the size of the first adjustment factor; e represents the exponential constant.

[0093] The second adjustment factor can be expressed as follows.

[0094] Among them, λ2 represents the second adjustment factor.

[0095] After obtaining the first adjustment factor and the second adjustment factor, the way to adjust the resource allocation strategy is to multiply the first adjustment factor and the second adjustment factor by the treatment cycle and the resource allocation, and use the product value calculated at this time as the adjusted resource allocation strategy, so as to complete the adjustment of the resource allocation.

[0096] Although the embodiments of the present invention have been shown and described above, it can be understood that the above embodiments are exemplary and should not be construed as limiting the present invention. Those of ordinary skill in the art can make changes, modifications, substitutions and variations to the above embodiments within the scope of the present invention, and still be covered by the protection scope of the present invention.

Claims

1. A data analysis and management system for diagnosis of mental illnesses, characterized in that: include: The label setting module obtains the patient's disease diagnosis data, performs preliminary classification of the disease diagnosis data, and sets a data label for each classified disease diagnosis data; Verify label module, check the missing data labels under each category, identify and record the missing information as the target data for verification; The symptom matching module matches the target data with the symptom model, obtains the disease score value corresponding to the target data, and determines the treatment plan for patients in each category based on the disease score value; analyzes the coverage and missing ratio of the treatment plan in each category to obtain the symptom matching result; The resource allocation module obtains the current patient's diagnosis matrix and symptom demand distribution according to the symptom matching results; determines the current patient's treatment cycle and resource allocation according to the current patient's diagnosis matrix and symptom demand distribution, and outputs the corresponding treatment cycle and resource allocation as a resource allocation strategy; The adjustment and optimization module verifies the eligibility of the resource allocation strategy. When the resource allocation strategy is qualified, the eligibility-related degree evaluation value is obtained. When the resource allocation strategy is unqualified, the adjustment factor of the resource allocation strategy is obtained. According to the adjustment factor of the resource allocation strategy, the resource allocation strategy is optimized; The implementation method of matching the target data with the symptom model is: preprocessing the target data, extracting the symptom features of the target data, performing similarity retrieval on the symptom features and the symptom model, and obtaining similarity retrieval results; When the symptom characteristics are verified to be accurate, similar search results are scored and calculated to obtain the disease score value; the treatment plan corresponding to the maximum disease score value under each category is used as the output treatment plan; The method of calculating the score of similar search results is as follows: obtaining a standard symptom list of the symptom model, and determining an initial score for each symptom in the standard symptom list; Find the output symptoms and the initial scores corresponding to the symptoms from the similar search results, make incentive adjustments according to the patients corresponding to the symptoms, and output the incentive-adjusted scores as disease scores; The incentive adjustment method is to obtain patient feedback scores and reputation scores , through feedback rating and The initial score is adjusted to obtain the disease score value; Feedback ratings are obtained as follows: ； in, represents the feedback score at the i-th feedback, Indicates the degree of demand for symptom query by the patient at the i-th feedback, represents the index value of the amount of data on the patient's symptom query at the i-th feedback, represents the incentive amount of data query at the i-th feedback; An indicator value representing the total demand for data query by patients at the time of feedback; The indicator value representing the cost of data query at the i-th feedback; ； in, represents the feedback score at the i-th feedback, represents the exponential constant, Represents the attenuation factor, which is used to control the speed of time decay. It represents the time difference between the time of the ith feedback and the current time. Indicates the number of feedbacks. ; The product of the feedback score, reputation score and initial score under each symptom is taken as the disease score value.

2. A system for analyzing and managing the diagnosis data of mental illnesses according to claim 1, characterized in that: The implementation of checking the missing data labels under each category is as follows: Determine the data ratio of the disease diagnosis data under each category as the first ratio; Extracting comparative data from the disease diagnosis data under each category, and obtaining the data ratio of the comparative data under each category as a second ratio; According to the obtained first ratio and second ratio, the missing rate of the corresponding field of disease diagnosis data under each category is counted; The disease diagnosis data after the initial classification is reclassified according to the obtained missing rate, and the part of the disease diagnosis data after the secondary classification whose data value and predicted value are greater than the preset standard is output as the target data.

3. A system for analyzing and managing the diagnosis data of mental illnesses according to claim 1, characterized in that: The method of performing similar search is: obtaining a first keyword set of symptoms and characteristics that are expected to be viewed and a second keyword set of related word combinations of the first keyword set; Calculate the matching degree of the first keyword set, the second keyword set and each symptom in the symptom model in turn; The matching degree of the first keyword set and the second keyword set is weighted and summed according to each symptom in the symptom model in turn to obtain a comprehensive matching degree; When the comprehensive matching degree is greater than the preset threshold, the symptoms and treatment plans corresponding to the symptom model are taken as similar retrieval results.

4. A system for analyzing and managing the diagnosis data of mental illnesses according to claim 1, characterized in that: When the diagnosis matrix and symptom demand distribution of the current patient in the resource allocation module are checked, it is also necessary to verify whether the diagnosis matrix and symptom demand distribution meet the requirements, and use the diagnosis matrix and symptom demand distribution that meet the requirements as the basis for selecting the resource allocation strategy; The implementation of the validation diagnosis matrix and symptom requirement distribution is as follows: ； in, represents the exponential constant, The first indicator function representing when the symptom occurs; a second indicator function representing when symptoms occur; represents the value of the symptom demand distribution of the jth symptom, represents the value of the first indicator function associated with the jth symptom, Represents the value of the diagnostic matrix related to the zth examination item, represents the value of the diagnostic matrix associated with the jth symptom and the zth examination item, represents the value of the second indicator function associated with the jth symptom, , .

5. A system for analyzing and managing the diagnosis data of mental illnesses according to claim 1, characterized in that: When determining the treatment cycle and resource allocation of the current patient, obtain the standard treatment duration, and obtain the treatment cycle based on the standard treatment duration, symptom demand distribution and diagnosis matrix; The calculation method of the treatment cycle is as follows: ； in, Represents the treatment cycle, Indicates the number of symptoms, , Indicates the number of inspection items. , represents the value of the symptom demand distribution of the jth symptom, represents the value of the standard treatment duration corresponding to the jth symptom, represents the value of the diagnostic matrix associated with the jth symptom and the zth examination item, represents the weight of the zth inspection item, represents the exponential constant, It is expressed as the minimum value of the weight of the inspection item. It is expressed as the maximum value of the weight of the inspection items; The implementation method of resource configuration is as follows: obtain the standard resource configuration factor, and obtain the resource configuration according to the resource configuration factor and treatment cycle: ； in, represents the value of the resource configuration for the jth symptom, represents the value of the symptom demand distribution of the jth symptom, Represents the treatment cycle, Represents the value of the resource allocation factor for the jth symptom; the treatment cycle and resource allocation at this time are output as the resource allocation strategy.

6. A system for analyzing and managing the diagnosis data of mental illnesses according to claim 1, characterized in that: The degree evaluation value in the adjustment optimization module can be expressed as follows: ； in, Indicates the degree evaluation value, Represents the treatment cycle, represents the standard value of the treatment cycle, represents the sum of resource allocations, Indicates the standard value of resource configuration, represents the missing proportion of treatment options, represents the standard value of the missing proportion of treatment options, represents the disease score value, represents the standard value of the disease score value, Represents an exponential constant.

7. A system for analyzing and managing the diagnosis data of mental illnesses according to claim 1, characterized in that: The implementation method of the adjustment and optimization module also includes: when the difference between the degree evaluation value and the preset evaluation value is less than or equal to the first threshold, the resource allocation strategy is regarded as qualified and the corresponding resource allocation strategy is output; when the difference between the degree evaluation value and the preset evaluation value is greater than the first threshold, the resource allocation strategy is adjusted using the first adjustment factor; when the difference between the degree evaluation value and the preset evaluation value is greater than the first threshold and less than the second threshold, the resource allocation strategy is adjusted using the second adjustment factor; when the difference between the degree evaluation value and the preset evaluation value is greater than or equal to the second threshold, the resource allocation strategy is reacquired until the difference between the degree evaluation value and the preset evaluation value is less than the second threshold.

Images