Method for automatically classifying and collecting doctor seeing information for performance assessment management of medical system

Abstract

The invention relates to the technical field of data processing, in particular to an automatic treatment information classification and collection method for medical system performance assessment management, and the method comprises the steps: obtaining a plurality of dimension data in the treatment information of each patient, taking any department as a target department, and taking the treatment information of the target department as target treatment information; determining an association degree between each piece of target treatment information and the target department, so as to screen out a main diagnosis and treatment direction of the target department from each type of disease in the target treatment information, thereby determining a performance assessment participation degree of each piece of dimension data of each piece of reference treatment information; the method is used for dividing all dimension data of each piece of reference treatment information into a plurality of classes. According to the method, the classification result is more accurate and credible by analyzing the participation degree of the data of each dimension of the doctor seeing information in the performance assessment.

Description

Technical Field

[0001] The present invention relates to the technical field of data processing, and in particular to a method for automatically classifying and collecting medical information for performance evaluation management of a medical system. Background Art

[0002] In the modern medical system, patients' medical information is usually used as a direct reference for performance appraisal. Through data processing and analysis of medical information, it can fully reflect a hospital's service efficiency, response speed, medical quality and medical capabilities. However, with the increase in the number of patients, the amount of medical information recorded by the hospital is also gradually increasing, which leads to increased pressure on the classification and statistics of patients' medical information. The traditional manual data processing method is too slow and has a high error rate in manual operations, and is gradually replaced by fast and efficient automated data classification and processing methods.

[0003] Existing methods mainly use supervised learning algorithms, such as the random forest algorithm, to achieve automatic classification of patient medical information. That is, according to different performance evaluation data requirements, different characteristic attributes and data values ​​of medical information samples are classified, and finally the classified medical information is collected.

[0004] Existing problems: When actually conducting performance appraisal management, different departments of the medical system have different standards for performance appraisal management, and due to differences in professional directions, departments of the same category have different standards for performance appraisal. This leads to different departments having different references to different medical information and data in the medical information. When automatically classifying and collecting, useless data may be collected and referenced, that is, the classification effect of medical information is poor, which leads to reduced authenticity and reliability of performance appraisal. Summary of the invention

[0005] The present invention provides an automatic classification and collection method for medical information used in performance evaluation management of a medical system to solve the existing problems.

[0006] The method for automatically classifying and collecting medical information for performance evaluation management of a medical system of the present invention adopts the following technical solutions: An embodiment of the present invention provides a method for automatically classifying and collecting medical information for performance evaluation management of a medical system, the method comprising the following steps: Acquire several dimensional data from the medical information of each patient; the dimensional data include: several departments involved in the treatment, several types of diseases suffered, the cure results of each type of disease, and the number of recurrences of each type of disease; Taking any department as the target department; taking the medical consultation information existing in the target department as the target medical consultation information; and determining the degree of association between each target medical consultation information and the target department according to the number of departments in the target medical consultation information; According to the degree of association between each target medical information and the target department, the cure result and the number of recurrences of each type of disease in the target medical information, the main diagnosis and treatment direction of the target department is screened out from each type of disease in the target medical information; According to the main diagnosis and treatment direction of the target department, reference medical information is screened out from all target medical information; according to the cure results of the main diagnosis and treatment direction of the target department in the reference medical information and the difference of the same dimensional data, the performance assessment participation of each dimensional data of each reference medical information is determined; Several level ranges are preset, and for any reference medical information, all dimensional data whose performance appraisal participation falls within any preset level range are classified into one category.

[0007] Furthermore, the step of determining the degree of association between each target medical consultation information and the target department includes the following specific steps: When When the number of departments in the target medical information is equal to the preset constant, the The target medical information is recorded as the medical information of a single department; When When the number of departments in the target medical information is greater than the preset constant, the The target medical information is recorded as multi-department medical information; Among all target medical information, the set consisting of all single-department medical information is recorded as the main set, and the set consisting of all multi-department medical information is recorded as the sub-set; Setting the correlation degree between each medical consultation information in the main set and the target department as a preset correlation degree parameter; According to the main set and the subset, the degree of association between each medical consultation information in the subset and the target department is determined.

[0008] Furthermore, the step of determining the degree of association between each medical consultation information in the subset and the target department includes the following specific steps: Use a random forest model to determine whether each medical consultation information in the subset belongs to the main set, and obtain the determination result of each medical consultation information in the subset in each decision tree; the determination result includes: belonging to the main set or not belonging to the main set; In the subset The number of decision trees in the main set is S1, and the ratio of S1 to the number of all decision trees is used as the first decision tree in the subset. The degree of correlation between the medical information and the target department.

[0009] Furthermore, the method of screening out the main diagnosis and treatment direction of the target department from each type of disease in the target medical information includes the following specific steps: According to the degree of correlation between each target medical consultation information and the target department, highly correlated medical consultation information is screened out from all target medical consultation information; Each type of disease in all highly correlated medical information is recorded as the target type of disease; Among all the target medical information, there are The number of target medical consultation information of the target type of disease S2 is calculated, and the ratio of S2 to the number of all target medical consultation information S3 is used as the first The frequency of occurrence of the target type of disease in the target department; In the existence Among all target visit information of target type diseases, according to The cure results and recurrence times of the target type of disease are determined The treatment capability of the target type of disease in the target department; the cure results include: cured and uncured; The said The treatment capacity of the target type of disease in the target department is the same as the first The normalized value of the product of the occurrence frequencies of the target type of disease in the target department is recorded as The diagnosis and treatment attention of the target type of disease in the target department; According to the diagnosis and treatment attention of all target types of diseases in the target departments, the main diagnosis and treatment directions of the target departments are screened out from all target types of diseases.

[0010] Furthermore, the step of selecting highly relevant medical information from all target medical information includes the following specific steps: When When the correlation between the target medical information and the target department is greater than the preset threshold, the The target medical information is recorded as high-correlation medical information.

[0011] Furthermore, the determination The specific steps involved in the treatment of the target type of disease in the target department are as follows: In the existence Among all the target medical consultation information of the target type of disease, the first The number of target medical consultation information with the cure result of the target type of disease is S4, and the statistics are The recurrence number of the target type of disease is greater than the target number of consultation information S5 preset as the second constant, and the ratio of S4 to S2 is recorded as The cure rate of the target type of disease in the target department is recorded as the ratio of S5 to S2. The recurrence rate of the target type of disease in the target department; The said The cure rate of the target type of disease in the target department is the same as that of the first The ratio of the recurrence rates of the target type of disease in the target department is recorded as The treatment capacity of the target type of disease in the target department.

[0012] Furthermore, the specific steps of screening out the main diagnosis and treatment directions of the target department from all target type diseases include the following: Among all target type diseases with regard to diagnosis and treatment attention in the target department, the target type disease corresponding to the maximum diagnosis and treatment attention is recorded as the main diagnosis and treatment direction of the target department.

[0013] Furthermore, the specific steps of selecting reference medical information from all target medical information are as follows: The target medical information of the main diagnosis and treatment direction of the target department is recorded as reference medical information.

[0014] Furthermore, the specific steps of determining the performance appraisal participation of each dimension data of each reference medical information include the following: Among all the reference medical information, all the reference medical information whose cure result in the main diagnosis and treatment direction of the target department is cured constitutes the cured category, and all the reference medical information whose cure result in the main diagnosis and treatment direction of the target department is uncured constitutes the uncured category; In the cured category, according to the difference of the same dimensional data in different reference medical information, the degree of response of each dimensional data of each reference medical information in the cured category to the patient's diagnosis and treatment results is determined; According to the method of obtaining the degree of response of each dimension data of each reference medical information in the cured category to the patient's medical treatment result, obtain the degree of response of each dimension data of each reference medical information in the uncured category to the patient's medical treatment result; The normalized value of the product of the treatment attention of the main treatment direction of the target department in the target department and the reaction degree of each dimension data of each reference medical information to the patient's treatment results is recorded as the performance assessment participation of each dimension data of each reference medical information.

[0015] Furthermore, the step of determining the degree of response of each dimension data of each reference medical information in the cured category to the patient's medical treatment result includes the following specific steps: In the cure category, calculate the The first The dimensional data is related to the first The difference in the data of the first dimension The first The dimension data is compared with the first dimension data of all reference medical information. The inverse normalized value of the mean of the difference in the dimension data is recorded as the first The first reference medical information The degree to which the dimensional data responds to the patient's diagnosis and treatment results.

[0016] The beneficial effects of the technical solution of the present invention are: In an embodiment of the present invention, several dimensional data in the medical information of each patient are obtained, and any department is taken as the target department. The medical information of the target department is taken as the target medical information, and the degree of association between each target medical information and the target department is determined, thereby determining whether the target department is the main treatment department during the patient's treatment process, thereby ensuring the accuracy of subsequent classification. Then, the main diagnosis and treatment direction of the target department is screened out from each type of disease in the target medical information, thereby determining the main type of disease treated by the target department, and further ensuring the accuracy of subsequent classification. The performance appraisal participation of each dimensional data of each reference medical information is obtained, so as to divide all dimensional data of each reference medical information into several categories, and assign a grade to each category, so as to make the process of automatic classification and collection more reasonable, so that the dimensional data with high participation is easier to collect. So far, the present invention analyzes the participation of each dimensional data of the medical information in the performance appraisal, so that the classification effect is more accurate and reliable, and the score of each dimensional data is adjusted according to the performance appraisal participation of each dimensional data, so as to improve the authenticity and reliability of performance appraisal management. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0018] Figure 1 It is a flowchart of the steps of the method for automatically classifying and collecting medical information used for performance evaluation management of a medical system according to the present invention; Figure 2 This is a flow chart for obtaining the comprehensive score of each department in the hospital in the present invention. DETAILED DESCRIPTION

[0019] In order to further explain the technical means and effects adopted by the present invention to achieve the predetermined invention purpose, the following is a detailed description of the specific implementation method, structure, features and effects of the automatic classification and collection method of medical information for performance appraisal management of the medical system proposed by the present invention in combination with the accompanying drawings and preferred embodiments. In the following description, different "one embodiment" or "another embodiment" does not necessarily refer to the same embodiment. In addition, specific features, structures or characteristics in one or more embodiments may be combined in any suitable form.

[0020] Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.

[0021] The specific scheme of the method for automatically classifying and collecting medical information for performance evaluation management of a medical system provided by the present invention is described in detail below with reference to the accompanying drawings.

[0022] See also Figure 1 , which shows a flowchart of a method for automatically classifying and collecting medical information for medical system performance evaluation management provided by an embodiment of the present invention, the method comprising the following steps: Step S001: Acquire several dimensional data in the medical information of each patient; the dimensional data include: several departments involved in the treatment, several types of diseases suffered from, the cure results of each type of disease and the number of recurrences of each type of disease.

[0023] The purpose of this embodiment is to analyze the relationship between medical information and departments, as well as the relationship between the data in the medical information and performance appraisal, so as to achieve participation marking of different data in performance appraisal, and classify the levels according to the participation labels, so as to make the results of automatic classification and collection more accurate and effective, and improve the authenticity and reliability of performance appraisal management.

[0024] In the database of any hospital, several dimensional data are collected from the medical information of each patient. The dimensional data include: several departments involved in the treatment, several types of diseases suffered, the cure results of each type of disease, and the number of recurrences of each type of disease. The cure results include: cured and uncured.

[0025] It should be noted that the patients in this embodiment are inpatients. The medical information of each patient is the medical information from the most recent hospitalization to discharge. The dimensional data also includes: the patient's personal information (name, ID number, etc.), the patient's medication information (drug number, name, category, dosage, etc.), and the doctor and nurse information treating the patient (doctor and nurse number, name, department number, etc.). Some patients have multiple types of diseases at the same time, and doctors from multiple departments are required to coordinate treatment during treatment. For the cured and uncured marks, when the patient's condition is effectively treated, the condition is stable or completely recovered, the doctor will record it as cured and discharged, but in some cases, the patient may be discharged from the hospital when the condition is not completely cured for various reasons, which may include the patient's active request for discharge, medical resource limitations, worsening of the condition and the need for transfer to another hospital for treatment. In this case, the doctor will record it as uncured discharge. The number of recurrences of each type of disease is obtained as follows: taking disease B in the medical information of patient A as an example, combined with the historical medical information, the number of times patient A was diagnosed with disease B in all the times he visited the hospital for treatment is counted, which is taken as the number of recurrences of disease B of patient A, that is, the number of recurrences is at least 1. When the number of recurrences is greater than 1, it means that disease B has been diagnosed in the historical medical information of patient A. Therefore, disease B in the most recent medical information is a recurring disease. In this embodiment, whether disease B in the historical medical information is cured is not considered.

[0026] It should be further explained that each dimension data is standardized to make the dimension uniform. For numerical dimensions, the minimum and maximum standardization method is used to normalize the data of each dimension to between 0 and 1. This is a well-known technology. For text dimensions, a pre-trained word embedding model (such as Word2Vec, a deep learning model for generating word vectors) is used to convert Chinese words into vector form. This is a well-known technology.

[0027] Step S002: taking any department as the target department; taking the medical consultation information existing in the target department as the target medical consultation information; and determining the degree of association between each target medical consultation information and the target department according to the number of departments in the target medical consultation information.

[0028] When a patient is receiving treatment in a hospital, there may be only one department involved in the treatment, or there may be multiple departments cooperating in the treatment at the same time. Therefore, a piece of medical information in the database may be referred to by multiple departments when conducting performance appraisal. However, due to the different degrees of participation of different departments in the treatment process, the degree of participation of medical information in performance appraisal is also different. It is necessary to analyze the degree of relevance of each department to each piece of medical information. The higher the degree of relevance between the medical information and the department, the higher the degree of participation of the department in the performance appraisal.

[0029] Preferably, in one embodiment of the present invention, the method for obtaining the degree of association between each target medical consultation information and the target department includes: Take any department as the target department. Take the medical information of the target department in the medical information as the target medical information, that is, the medical information corresponding to the target department in the hospital database.

[0030] The preset constant is 1, and this is used as an example for description.

[0031] When When the number of departments in the target medical information is equal to the preset constant (only the target department exists), The target medical information is recorded as single department medical information.

[0032] When When the number of departments in the target medical information is greater than the preset constant (there are departments other than the target department), The target medical treatment information is recorded as multi-department medical treatment information.

[0033] According to the above method, all target medical information is divided into single-department medical information and multi-department medical information.

[0034] The set consisting of all single-department medical records is recorded as the main set C. The set consisting of all multi-department medical records is recorded as the sub-set D.

[0035] A random forest model is used to determine whether each medical consultation information in the diversity set D belongs to the main set C, and the determination result of each medical consultation information in the diversity set D in each decision tree is obtained. The determination results include: belonging to the main set C or not belonging to the main set C.

[0036] It should be noted that the random forest model is a well-known technology. The random forest model improves prediction performance by integrating multiple decision trees. Each decision tree independently classifies the data and outputs a classification result. The specific process of classification is: assign a label to each medical information in the main set C (for example, the label is 1, indicating that it belongs to C), and then assign another label to each medical information in the subset D (for example, the label is 0, indicating that it does not belong to C), merge the main set C and the subset D into a new set E, and randomly divide a part of the new set E as a training set and another part as a test set. Make sure that both the training set and the test set contain medical information from C and D, use the training set data to train the random forest classifier, use the test set data to evaluate the performance of the random forest model, and finally use each decision tree in the model to determine whether each medical information in C belongs to D. This is a well-known operation.

[0037] In the diversity set D The number of decision trees in the main set C is S1. The ratio of S1 to the number of all decision trees is used as the first decision tree in the diversity set D. The degree of correlation between the medical information and the target department.

[0038] According to the above method, the correlation degree between each medical consultation information in the diversity set D and the target department is obtained.

[0039] The preset correlation parameter is 1, and this is used as an example for description.

[0040] The degree of association between each medical consultation information in the main set C and the target department is set as a preset association parameter.

[0041] It should be noted that the department in each medical information in the main set C only has the target department, so in this embodiment, the degree of association between each medical information in the main set C and the target department is set to 1. However, each medical information in the diverse set D contains the target department and other departments, so when the ratio of S1 to the number of all decision trees is closer to 1, it means that the first medical information in the diverse set D is The more likely the target department in the medical information is the patient's primary treatment department, the more likely the first The greater the correlation between the medical information and the target department.

[0042] Thus, the correlation degree between each target medical consultation information and the target department is obtained.

[0043] According to the above method, the correlation degree between each medical consultation information in each department and each department is obtained.

[0044] Step S003: According to the degree of association between each target medical information and the target department, the cure results and the number of recurrences of each type of disease in the target medical information, the main diagnosis and treatment direction of the target department is screened out from each type of disease in the target medical information.

[0045] In the medical system, in order to make medical services more professional and efficient, and to provide more accurate treatment plans for different case conditions, different departments usually have their main diagnosis and treatment directions, which are based on the division of medical specialties and the demand for specialized treatment of diseases. By analyzing the medical information of different departments, we can obtain the diagnosis and treatment capabilities of each department for each type of disease, and then obtain the attention paid to each type of disease, so as to determine the main diagnosis and treatment direction of the department.

[0046] Preferably, in one embodiment of the present invention, the method for acquiring the main diagnosis and treatment direction of the target department includes: The preset threshold is 0.5, and this is used as an example for description.

[0047] Still taking the target department and the target medical information corresponding to the target department as an example, when When the correlation between the target medical information and the target department is greater than the preset threshold, the first The target medical information is recorded as high-correlation medical information.

[0048] According to the above method, it is determined whether each target medical consultation information is highly correlated medical consultation information.

[0049] It should be noted that the higher the degree of association, the more it indicates that the main diagnosis and treatment direction of the target department is a certain disease suffered by the patient corresponding to the target medical information.

[0050] Each type of disease in all highly correlated medical information is recorded as a target type of disease.

[0051] Among all the target medical information, there are The number of target medical consultation information of the target type of disease S2 is taken as the ratio of S2 to the number of all target medical consultation information S3. The frequency of occurrence of the target type of disease in the target department.

[0052] What needs to be explained is: The greater the frequency of occurrence of the target type of disease in the target department, the higher the frequency of occurrence of the target type of disease in the target department. The greater the demand for diagnosis and treatment of a target type of disease, the more it will be the key development direction of the target department.

[0053] The second constant is preset to 1, and this is taken as an example for description.

[0054] In the existence Among all the target medical consultation information of the target type of disease, the first The number of target medical consultation information with the cure result of the target type of disease is S4, and the statistics are The number of recurrences of the target type of disease is greater than the target number of consultation information S5 preset as the second constant 1, and the ratio of S4 to S2 is recorded as The cure rate of the target type of disease in the target department is recorded as the ratio of S5 to S2. The recurrence rate of the target type of disease in the target department.

[0055] The first The cure rate of the target type of disease in the target department is the same as that of the first The ratio of the recurrence rates of the target type of disease in the target department is recorded as The treatment capacity of the target type of disease in the target department.

[0056] The first The treatment capacity of the target type of disease in the target department is different from that of the first The normalized value of the product of the occurrence frequencies of the target type of disease in the target department is recorded as The diagnosis and treatment attention of the target type of disease in the target department.

[0057] What needs to be explained is that a higher cure rate and a lower recurrence rate can indicate that the target department is The target type of disease has a stronger treatment ability, so the larger the ratio, the stronger the treatment ability. The greater the frequency of occurrence of the target type of disease and the stronger the treatment ability, the higher the The more target type of disease in the target department needs to be diagnosed and treated. The normalized value of the product of the frequency of occurrence is used in this embodiment. The linear normalization function is used to normalize the product of the occurrence frequency to between 0 and 1, and this is used as an example for description.

[0058] According to the above method, the diagnosis and treatment attention of each target type of disease in the target department is obtained.

[0059] Among all target type diseases with regard to diagnosis and treatment attention in the target department, the target type disease corresponding to the maximum diagnosis and treatment attention is recorded as the main diagnosis and treatment direction of the target department.

[0060] It should be noted that when there are multiple maximum diagnosis and treatment concerns, it means that the target department has multiple main diagnosis and treatment directions. Select one of the maximum diagnosis and treatment concerns as an example for subsequent analysis.

[0061] Step S004: Filter out reference medical information from all target medical information based on the main diagnosis and treatment direction of the target department; determine the performance appraisal participation of each dimensional data of each reference medical information based on the cure results of the main diagnosis and treatment direction of the target department in the reference medical information and the differences in the same dimensional data.

[0062] Due to the differences in the main diagnosis and treatment directions of different departments, even if two departments belong to the same category of diseases, the same medical information data will have different participation levels in performance appraisal. For example, the time and money spent in the treatment process of chronic diseases and acute diseases have different participation levels in performance appraisal. In summary, combined with the diagnosis and treatment directions that each department is good at, the performance appraisal participation level of each dimension of medical information data is analyzed.

[0063] Preferably, in one embodiment of the present invention, a method for obtaining the performance appraisal participation of each dimension data of each reference medical information includes: Still taking the target department and the target medical information corresponding to the target department as an example, the target medical information of the main diagnosis and treatment direction of the target department is recorded as the reference medical information.

[0064] It should be noted that the target medical information that does not exist in the main diagnosis and treatment direction of the target department will not be analyzed subsequently. This is because these target medical information generally correspond to patients who are assisted in treatment by the target department, not patients who are the main treatment of the target department. Therefore, they are generally not included in the performance appraisal, so there is no need to classify the information.

[0065] Among all the reference medical information, all the reference medical information whose cure result in the main diagnosis and treatment direction of the target department is cured constitutes the cured class, and all the reference medical information whose cure result in the main diagnosis and treatment direction of the target department is uncured constitutes the uncured class.

[0066] In the healing category, Take the reference medical information as an example and calculate the The first The dimensional data is related to the first The difference in the data of the first dimension The first The dimension data is compared with the first dimension data of all reference medical information. The inverse normalized value of the mean of the difference in the dimension data is recorded as the first The first reference medical information The degree to which the dimensional data responds to the patient's diagnosis and treatment results.

[0067] It should be noted that: in the same category, the smaller the difference between the same dimension data in different medical information of the same type of disease, the more consistent the treatment effect, so the greater the degree of response of this dimension data to the patient's diagnosis and treatment results. When the dimension data is numerical data, in the cure category, the The first The dimensional data is related to the first The absolute value of the difference between the dimensions is taken as the The first The dimensional data is related to the first The difference in data of each dimension. When the dimension data is text data, in the cure category, calculate the The first The text vector of the dimension data is related to the first The normalized value of the cosine similarity of the text vector of the dimension data is obtained, and the difference between 1 and the normalized value of the cosine similarity is taken as the first The first The dimensional data is related to the first The difference in data dimensions.

[0068] It should be further explained that cosine similarity is a well-known calculation, and the value range of cosine similarity is between -1 and 1. The closer it is to 1, the more similar the two text vectors are. Therefore, the smaller the cosine similarity, the greater the difference between the two text vectors. The linear normalization function is used to normalize the cosine similarity to between 0 and 1. The first The dimension data is compared with the first dimension data of all reference medical information. The mean of the difference of the dimension data is recorded as In this embodiment, To present The inverse proportional relationship and normalization processing, This is an exponential function with a natural constant as its base, and we will use this as an example to explain it.

[0069] According to the above method, the response degree of each dimension data of each reference medical information in the cured category to the patient's diagnosis and treatment results is obtained.

[0070] According to the method of obtaining the degree of response of each dimension data of each reference medical information in the cured category to the patient's diagnosis and treatment results, the degree of response of each dimension data of each reference medical information in the uncured category to the patient's diagnosis and treatment results is obtained.

[0071] Thus, the degree to which each dimension data of each reference medical information responds to the patient's medical treatment results is obtained.

[0072] The normalized value of the product of the treatment attention of the target department's main treatment direction in the target department and the degree of response of each dimensional data of each reference medical information to the patient's treatment results is recorded as the performance assessment participation of each dimensional data of each reference medical information.

[0073] It should be noted that the normalized value of the product of the reaction degree is used in this embodiment. The linear normalization function is used to normalize the product of the reaction degree to between 0 and 1, and this example is used for description. Among them, the greater the attention to diagnosis and treatment, and the greater the degree of reaction to the patient's diagnosis and treatment results, the more important the data of this dimension is to the performance appraisal, that is, the greater the participation in performance appraisal.

[0074] Step S005: Preset a number of level ranges, and for any reference medical information, classify all dimensional data whose performance appraisal participation degree belongs to any preset level range into one category.

[0075] The preset ranges are: Preset first level range , preset second level range , preset third level range , Preset fourth level range And the preset fifth level range , taking this as an example for description.

[0076] For any reference medical information, all dimension data whose performance appraisal participation falls within any preset level range are classified into one category.

[0077] Thus, all dimensional data in each reference medical information are divided into several categories, and the level corresponding to each category is determined, and the higher the level, the more important the dimensional data. Thus, the classified collection of all dimensional data in each reference medical information is completed.

[0078] According to the above method, all dimensional data in the medical information corresponding to each department and the main diagnosis and treatment direction of each department are classified and collected.

[0079] It should be noted that hospitals will use some standardized evaluation tools and methods, such as the HCAHPS (Hospital Consumer Assessment of Healthcare Providers and Systems) survey, which is a widely used patient satisfaction survey tool. This is a well-known technology. For all reference medical information corresponding to the target department, the HCAHPS survey is used to obtain the score of each dimension data of each reference medical information. The performance appraisal participation of each dimension data of each reference medical information is used as the weight, and the scores of all dimension data of each reference medical information are weighted and summed to obtain the comprehensive score of each reference medical information. The average of the comprehensive scores of all reference medical information is used as the comprehensive score of the target department. Among them, the dimensional data that is not scored does not participate in the calculation of the comprehensive score of each reference medical information. According to the above method, the comprehensive score of each department in the hospital can be obtained to complete the performance appraisal management. Among them, the flowchart for obtaining the comprehensive score of each department in the hospital is as follows: Figure 2 shown.

[0080] It should be noted that: in this embodiment, when calculating the ratio of two data values, when the denominator in the ratio is 0, the denominator is set to 0.1 to ensure that the ratio holds, and this example is used for description.

[0081] So far, the present invention is completed.

[0082] In summary, in an embodiment of the present invention, several dimensional data in the medical information of each patient are obtained, and any department is taken as the target department. The medical information of the target department is taken as the target medical information, and the degree of association between each target medical information and the target department is determined, so as to screen out the main diagnosis and treatment direction of the target department from each type of disease in the target medical information, thereby determining the performance appraisal participation of each dimensional data of each reference medical information, so as to divide all dimensional data of each reference medical information into several categories. The present invention makes the classification effect more accurate and reliable by analyzing the participation of each dimensional data of the medical information in the performance appraisal.

[0083] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A method for automatically classifying and collecting medical information for performance evaluation management of a medical system, characterized in that: The method comprises the following steps: Acquire several dimensional data from the medical information of each patient; the dimensional data include: several departments involved in the treatment, several types of diseases suffered, the cure results of each type of disease, and the number of recurrences of each type of disease; Taking any department as the target department; taking the medical consultation information existing in the target department as the target medical consultation information; and determining the degree of association between each target medical consultation information and the target department according to the number of departments in the target medical consultation information; According to the degree of association between each target medical information and the target department, the cure result and the number of recurrences of each type of disease in the target medical information, the main diagnosis and treatment direction of the target department is screened out from each type of disease in the target medical information; According to the main diagnosis and treatment direction of the target department, reference medical information is screened out from all target medical information; according to the cure results of the main diagnosis and treatment direction of the target department in the reference medical information and the difference of the same dimensional data, the performance assessment participation of each dimensional data of each reference medical information is determined; Several level ranges are preset, and for any reference medical information, all dimensional data whose performance appraisal participation falls within any preset level range are classified into one category.

2. The method for automatically classifying and collecting medical information for performance evaluation management of a medical system according to claim 1, characterized in that: The specific steps of determining the correlation between each target medical consultation information and the target department are as follows: When When the number of departments in the target medical information is equal to the preset constant, the The target medical information is recorded as the medical information of a single department; When When the number of departments in the target medical information is greater than the preset constant, the The target medical information is recorded as multi-department medical information; Among all target medical information, the set consisting of all single-department medical information is recorded as the main set, and the set consisting of all multi-department medical information is recorded as the sub-set; Setting the correlation degree between each medical consultation information in the main set and the target department as a preset correlation degree parameter; According to the main set and the subset, the degree of association between each medical consultation information in the subset and the target department is determined.

3. The method for automatically classifying and collecting medical information for performance evaluation management of a medical system according to claim 2 is characterized in that: The step of determining the degree of association between each medical consultation information in the subset and the target department includes the following specific steps: Use a random forest model to determine whether each medical consultation information in the subset belongs to the main set, and obtain the determination result of each medical consultation information in the subset in each decision tree; The determination results include: belonging to the main set and not belonging to the main set; In the subset The number of decision trees in the main set is S1, and the ratio of S1 to the number of all decision trees is used as the first decision tree in the subset. The degree of correlation between the medical information and the target department.

4. The method for automatically classifying and collecting medical information for performance evaluation management of a medical system according to claim 1, characterized in that: The specific steps of screening out the main diagnosis and treatment direction of the target department from each type of disease in the target medical information are as follows: According to the degree of correlation between each target medical consultation information and the target department, highly correlated medical consultation information is screened out from all target medical consultation information; Each type of disease in all highly correlated medical information is recorded as the target type of disease; Among all the target medical information, there are The number of target medical consultation information of the target type of disease S2 is calculated, and the ratio of S2 to the number of all target medical consultation information S3 is used as the first The frequency of occurrence of the target type of disease in the target department; In the existence Among all target visit information of target type diseases, according to The cure results and recurrence times of the target type of disease are determined The treatment capability of the target type of disease in the target department; the cure results include: cured and uncured; The said The treatment capacity of the target type of disease in the target department is the same as the first The normalized value of the product of the occurrence frequencies of the target type of disease in the target department is recorded as The diagnosis and treatment attention of the target type of disease in the target department; According to the diagnosis and treatment attention of all target types of diseases in the target departments, the main diagnosis and treatment directions of the target departments are screened out from all target types of diseases.

5. The method for automatically classifying and collecting medical information for performance evaluation management of a medical system according to claim 4, characterized in that: The specific steps of screening out highly relevant medical information from all target medical information are as follows: When When the correlation between the target medical information and the target department is greater than the preset threshold, the The target medical information is recorded as high-correlation medical information.

6. The method for automatically classifying and collecting medical information for performance evaluation management of a medical system according to claim 4, characterized in that: The determination The specific steps involved in the treatment of the target type of disease in the target department are as follows: In the existence Among all the target medical consultation information of the target type of disease, the first The number of target medical consultation information with the cure result of the target type of disease is S4, and the statistics are The recurrence number of the target type of disease is greater than the target number of consultation information S5 preset as the second constant, and the ratio of S4 to S2 is recorded as The cure rate of the target type of disease in the target department is recorded as the ratio of S5 to S2. The recurrence rate of the target type of disease in the target department; The said The cure rate of the target type of disease in the target department is the same as that of the first The ratio of the recurrence rates of the target type of disease in the target department is recorded as The treatment capacity of the target type of disease in the target department.

7. The method for automatically classifying and collecting medical information for performance evaluation management of a medical system according to claim 4, characterized in that: The specific steps of screening out the main diagnosis and treatment directions of the target department from all target type diseases are as follows: Among all target type diseases with regard to diagnosis and treatment attention in the target department, the target type disease corresponding to the maximum diagnosis and treatment attention is recorded as the main diagnosis and treatment direction of the target department.

8. The method for automatically classifying and collecting medical information for performance evaluation management of a medical system according to claim 1, characterized in that: The specific steps of selecting reference medical information from all target medical information are as follows: The target medical information of the main diagnosis and treatment direction of the target department is recorded as reference medical information.

9. The method for automatically classifying and collecting medical information for performance evaluation management of a medical system according to claim 4, characterized in that: The specific steps of determining the performance assessment participation of each dimension data of each reference medical information are as follows: Among all the reference medical information, all the reference medical information whose cure result of the main diagnosis and treatment direction of the target department is cured constitutes the cured category, and all the reference medical information whose cure result of the main diagnosis and treatment direction of the target department is uncured constitutes the uncured category; In the cured category, according to the difference of the same dimensional data in different reference medical information, the degree of response of each dimensional data of each reference medical information in the cured category to the patient's diagnosis and treatment results is determined; According to the method of obtaining the degree of response of each dimension data of each reference medical information in the cured category to the patient's medical treatment result, obtain the degree of response of each dimension data of each reference medical information in the uncured category to the patient's medical treatment result; The normalized value of the product of the treatment attention of the main treatment direction of the target department in the target department and the reaction degree of each dimension data of each reference medical information to the patient's treatment results is recorded as the performance assessment participation of each dimension data of each reference medical information.

10. The method for automatically classifying and collecting medical information for performance evaluation management of a medical system according to claim 9, characterized in that: The specific steps of determining the degree of response of each dimension data of each reference medical information in the cured category to the patient's medical treatment result are as follows: In the cure category, calculate the The first The dimensional data is related to the first The difference in the data of the first dimension The first The dimension data is compared with the first dimension data of all reference medical information. The inverse normalized value of the mean of the difference in the dimension data is recorded as the first The first reference medical information The degree to which each dimension of data responds to the patient's diagnosis and treatment results; Among them, when When the dimension data is numerical data, in the cure category, the The first The dimensional data is related to the first The absolute value of the difference between the dimensions is taken as the The first The dimensional data is related to the first Differences in data across dimensions; When When the dimension data is text data, in the cure category, calculate the The first The text vector of the dimension data is related to the first The normalized value of the cosine similarity of the text vector of the dimension data is obtained, and the difference between 1 and the normalized value of the cosine similarity is taken as the first The first The dimensional data is related to the first The difference in data dimensions.

Images