Patient-centered medical path optimization system

Abstract

The invention relates to a patient-centered medical path optimization system. The system comprises a data storage module, a diagnosis group determination module, a medical path optimization module and a display module, wherein the data storage module is used for storing medical data and standard data of each diagnosis group; the medical data are data involved in a hospital diagnosis and treatment process, and the standard data are regularly published by related personnel; the diagnosis group determination module is used for determining a target diagnosis group to be subjected to medical path optimization based on the standard data and the medical data of each diagnosis group; the medical path optimization module is used for optimizing the medical path of the target diagnosis group based on the patients of the target diagnosis group; and the display module is used for displaying the medical data, the standard data and the medical path of each diagnosis group. According to the system, the medical path of the target diagnosis group can be optimized based on the patients of the target diagnosis group, and then the medical path is displayed, so that the medical path difference of the same diagnosis group in different departments and / or diagnosis and treatment groups is reduced.

Description

Technical Field

[0001] The present invention relates to the field of medical technology, and particularly to a patient-centered medical path optimization system. Background Art

[0002] Patients can use medical insurance funds based on the medical insurance system during the medical treatment process.

[0003] Through DRG (Diagnosis Related Groups) / DIP (Disease Severity and Intervention Complexity Points), on the one hand, the utilization efficiency of medical insurance funds can be improved, and on the other hand, the allocation of medical resources can be reasonably regulated.

[0004] At present, a major problem is that there are significant differences in the medical paths of different departments for diagnostic groups. Summary of the Invention

[0005] The present invention provides a patient-centered medical path optimization system, which includes: a data storage module, a diagnostic group determination module, a medical path optimization module, and a display module;

[0006] Among them, the data storage module is used to store medical data and the standard data of each diagnostic group; the medical data is the data involved in the hospital diagnosis and treatment process, and the standard data is regularly released by relevant personnel;

[0007] The diagnostic group determination module is used to determine the target diagnostic group to be optimized for the medical path based on the standard data and medical data of each diagnostic group;

[0008] The medical path optimization module is used to optimize the medical path of the target diagnostic group based on the patients of the target diagnostic group;

[0009] The display module is used to display medical data, standard data, and the medical paths of each diagnostic group.

[0010] Optionally, determining the target diagnostic group to be optimized for the medical path based on the standard data and medical data of each diagnostic group includes:

[0011] Classify the medical data by diagnostic group to obtain the medical data of each diagnostic group;

[0012] Determine the medical data belonging to each diagnostic group in each department according to the diagnostic group to which the medical data of each department belongs;

[0013] Determine the target patients for each diagnostic group in each department and the target doctors for each diagnostic group in each department; among them, for any diagnostic group i in any department j, the target patients are the patients corresponding to each medical data belonging to diagnostic group i in department j, and the target doctor for diagnostic group i in department j is the doctor corresponding to each medical data belonging to diagnostic group i in department j;

[0014] Determine the probability of each diagnostic group for each department according to the medical data belonging to each diagnostic group in each department, the target patients and target doctors for each diagnostic group in each department;

[0015] Determine the diagnosis and treatment optimization degree of each diagnostic group for each department according to the standard data of each diagnostic group, the medical data belonging to each diagnostic group in each department, the target patients and target doctors for each diagnostic group in each department;

[0016] Determine the diagnosis and treatment optimization degree of each diagnostic group according to the probability of each diagnostic group for each department and the diagnosis and treatment optimization degree of each diagnostic group for each department; among them, the diagnosis and treatment optimization degree ODi of any diagnostic group i = ∑j(ODij × Pij) / J; where J is the total number of departments, ODij is the diagnosis and treatment optimization degree of diagnostic group i for department j, and Pij is the probability of diagnostic group i for department j;

[0017] Determine the diagnostic groups with the diagnosis and treatment optimization degree less than the preset threshold as the target diagnostic groups.

[0018] Optionally, determining the probability of each diagnostic group for each department according to the medical data belonging to each diagnostic group in each department, the target patients and target doctors for each diagnostic group in each department includes:

[0019] For any diagnostic group i and any department j, determine the probability of diagnostic group i for department j through the following steps:

[0020] Determine the number of target patients NPAij of diagnostic group i in department j and the number of target doctors NDOij of diagnostic group i in department j;

[0021] Based on the medical data, determine the total number of diagnoses and treatments of each target patient of diagnostic group i in department j, the number of diagnoses and treatments of each target patient of diagnostic group i in department j in department j, and the number of diagnoses and treatments belonging to diagnostic group i in the diagnoses and treatments of each target patient of diagnostic group i in department j in department j;

[0022] Determine the weight of each target patient of diagnostic group i in department j for department j; among them, the weight Wuj of any target patient u of diagnostic group i in department j for department j = the number of diagnoses and treatments of target patient u in department j / the total number of diagnoses and treatments of target patient u;

[0023] Determine the weights of each target patient in diagnosis group i of department j for diagnosis group i in department j; where, for any target patient u in diagnosis group i of department j, the weight Wuij of patient u for diagnosis group i in department j = the number of times patient u belongs to diagnosis group i in the diagnosis and treatment of department j / the total number of diagnosis and treatment times of patient u;

[0024] Determine the patient weight WU = ∑u(Wuij) / ∑u(Wuj);

[0025] Determine the patient ratio Rij(1) = WU × [NPAij / the total number of all patients diagnosed and treated in department j];

[0026] Determine the professional titles of each target doctor in diagnosis group i of department j;

[0027] Based on the medical data, determine the total number of diagnosis and treatment times of each target doctor in diagnosis group i of department j, the number of diagnosis and treatment times of each target doctor in diagnosis group i in department j, and the number of diagnosis and treatment times of each target doctor in diagnosis group i in department j that belong to diagnosis group i;

[0028] Determine the weights of each target doctor in diagnosis group i of department j for department j; where, for any target doctor v in diagnosis group i of department j, the weight Wvj of doctor v for department j = the number of diagnosis and treatment times of doctor v in department j / the total number of diagnosis and treatment times of doctor v;

[0029] Determine the weights of each target doctor in diagnosis group i of department j for diagnosis group i in department j; where, for any target doctor v in diagnosis group i of department j, the weight Wvij of doctor v for diagnosis group i in department j = the number of diagnosis and treatment times of doctor v in department j that belong to diagnosis group i / the total number of diagnosis and treatment times of doctor v;

[0030] Determine the doctor weight WD = [∑v(Wvij) / ∑v(Wvj)];

[0031] Determine the professional title weight WT = the highest professional title among all target doctors in diagnosis group i of department j / the highest professional title among all doctors in department j;

[0032] Determine the doctor ratio Rij(2) = [NDOij / the total number of doctors in department j] × (WD + WT) / 2;

[0033] Determine the probability Pij of diagnosis group i for department j = [Rij(1) + Rij(2)] / 2.

[0034] Optionally, according to the standard data of each diagnosis group, the medical data belonging to each diagnosis group in each department, the target patients and target doctors for each diagnosis group in each department, determine the diagnosis and treatment optimization degrees of each diagnosis group for each department, including:

[0035] For any diagnosis group i and any department j, the diagnosis and treatment optimization degree of diagnosis group i for department j is determined through the following steps:

[0036] Determine the matching degree between the medical data of each target patient of diagnosis group i for department j and the standard data of diagnosis group i;

[0037] Determine the minimum value Simu(min) and the standard deviation Simu(sd) among the matching degrees between the medical data of each target patient of diagnosis group i for department j and the standard data of diagnosis group i;

[0038] Determine the matching degree between the medical data of each target doctor of diagnosis group i for department j and the standard data of diagnosis group i;

[0039] Determine the minimum value Simv(min) and the standard deviation Simv(sd) among the matching degrees between the medical data of each target doctor of diagnosis group i for department j and the standard data of diagnosis group i;

[0040] Determine the diagnosis and treatment optimization degree ODij of diagnosis group i for department j = min{Simu(min) / patient matching degree threshold, Simv(min) / doctor matching degree threshold, 1 - [Simu(sd) / standard deviation threshold], 1 - [Simv(sd) / standard deviation threshold]}; where, min{} is the minimum value function.

[0041] Optionally, based on the patients of the target diagnosis group, optimize the medical path of the target diagnosis group, including:

[0042] Determine the medical data of the target diagnosis group as the target medical data;

[0043] According to the target medical data, determine the medical paths of each optimized doctor for each optimized patient and the standard degree of each medical path; where, the optimized doctor is the doctor involved in the target medical data, and the optimized patient is the patient involved in the medical data of the target diagnosis group;

[0044] Obtain the optimized medical path of the target diagnosis group according to the medical paths whose standard degree is not less than the standard degree threshold.

[0045] Among them, the medical path is composed of medical nodes arranged in the order of diagnosis and treatment;

[0046] Obtain the optimized medical path of the target diagnosis group according to the medical paths whose standard degree is not less than the standard degree threshold, including:

[0047] Construct a path diagram of the target disease according to all target medical paths; wherein, the target medical path is a medical path with a standard degree not less than the standard degree threshold, the points in the path diagram correspond one-to-one with the medical nodes in all medical paths, and the edges between the nodes in the path diagram correspond to the diagnosis and treatment order between the medical nodes in the medical path; any point in the path diagram has a point attribute, and the point attribute value of any node is the number of medical paths containing any node; any edge in the path diagram has an edge attribute, and the edge attribute value of any edge is the number of medical paths with the execution order between the two points connected by any edge;

[0048] Among the points in the path diagram corresponding to the first medical node of all target medical paths, take the point with the largest point attribute value as the starting node;

[0049] Among the points in the path diagram corresponding to the last medical node of all target medical paths, take the point with the largest point attribute value as the ending node;

[0050] Determine the target link in the path diagram; wherein, the starting point of any target link is the starting node, and the ending point is the ending node;

[0051] Determine the link value of each target link according to the point attribute value and edge attribute value of the target link;

[0052] Obtain the medical path of the optimized target diagnosis group according to the link value.

[0053] Optionally, determining the link value of each target link according to the point attribute value and edge attribute value of the target link includes:

[0054] For any target link l, determine its link value through the following steps:

[0055] For any edge of the target link l, determine its edge value as the edge attribute value of any edge × (the average value of the point attribute values of the two points connected by any edge);

[0056] Determine the average value Eavg(l) and standard deviation Esd(l) of the edge values of all edges of the target link l;

[0057] Determine the link value of the target link l as Esd(l) / Eavg(l) × the typicality of the target link l; wherein, the typicality of the target link l = [(the professional title of the doctor corresponding to the target link l / the highest professional title of all doctors corresponding to the target link) + the matching degree between the medical data corresponding to the target link l and the standard data of the diagnosis group i] / 2.

[0058] Optionally, obtaining the medical path of the optimized target diagnosis group according to the link value includes:

[0059] Determine the patients corresponding to each target link;

[0060] Cluster the corresponding patients according to patient attributes;

[0061] Determine the standard deviation of the link values of all target links in each category;

[0062] For the category with a standard deviation not greater than the standard deviation threshold, determine the target link with the largest link value in this category as the medical path of the optimized target diagnosis group;

[0063] For the category with a standard deviation greater than the standard deviation threshold, determine the target link with the largest typicality in this category as the medical path of the optimized target diagnosis group.

[0064] Optionally, the computer program involved in the patient-centered medical path optimization system is stored in the memory;

[0065] The memory is located inside the electronic device, and the electronic device further includes: a processor;

[0066] The processor is used to execute the computer program involved in the patient-centered medical path optimization system to implement the patient-centered medical path optimization system.

[0067] Optionally, the computer program involved in the patient-centered medical path optimization system is stored in a computer-readable storage medium;

[0068] The computer program involved in the patient-centered medical path optimization system is executed by the processor to implement the patient-centered medical path optimization system.

[0069] The present invention relates to a patient-centered medical path optimization system, which includes: a data storage module, a diagnosis group determination module, a medical path optimization module, and a display module; wherein, the data storage module is used to store medical data and standard data of each diagnosis group; the medical data is data involved in the hospital diagnosis and treatment process, and the standard data is regularly released by relevant personnel; the diagnosis group determination module is used to determine the target diagnosis group to be optimized for the medical path based on the standard data and medical data of each diagnosis group; the medical path optimization module is used to optimize the medical path of the target diagnosis group based on the patients in the target diagnosis group; the display module is used to display medical data, standard data, and the medical paths of each diagnosis group. The medical path optimization module of the system of the present invention can optimize the medical path of the target diagnosis group based on the patients in the target diagnosis group, and then display it through the display module, reducing the difference in medical paths of the same diagnosis group among different departments and / or diagnosis and treatment groups. Description of the Drawings

[0070] Figure 1 It is a schematic structural diagram of a patient-centered medical path optimization system provided by an embodiment of the present application;

[0071] Figure 2 A schematic diagram of an admission record provided by an embodiment of the present application;

[0072] Figure 3 A schematic diagram showing the distribution of medical data of the diagnosis group "ES2 pulmonary mycosis" in departments provided by an embodiment of the present application;

[0073] Figure 4 A schematic diagram of a medical path provided by an embodiment of the present application;

[0074] Figure 5 A schematic diagram of a roadmap provided by an embodiment of the present application;

[0075] Figure 6 A schematic diagram showing the distribution of patients provided by an embodiment of the present application;

[0076] Figure 7 A schematic diagram of the case situation in medical data provided by an embodiment of the present application;

[0077] Figure 8 A schematic diagram showing the distribution of doctors provided by an embodiment of the present application;

[0078] Figure 9 A schematic diagram showing the distribution of DRG departments provided by an embodiment of the present application. Detailed implementation manners

[0079] For better explaining the present invention for easy understanding, the present invention will be described in detail below with reference to the accompanying drawings through specific implementation manners.

[0080] Currently, a main problem is that there are significant differences in the medical paths of the same diagnosis group among different departments and / or treatment groups. Therefore, the present invention provides a patient-centered medical path optimization system, which includes: a data storage module, a diagnosis group determination module, a medical path optimization module, and a display module; wherein, the data storage module is used to store medical data and the standard data of each diagnosis group; the medical data is the data involved in the hospital diagnosis and treatment process, and the standard data is regularly released by relevant personnel; the diagnosis group determination module is used to determine the target diagnosis group to be optimized for the medical path based on the standard data and medical data of each diagnosis group; the medical path optimization module is used to optimize the medical path of the target diagnosis group based on the patients of the target diagnosis group; the display module is used to display the medical data, standard data, and the medical paths of each diagnosis group. The medical path optimization module of the system of the present invention can optimize the medical path of the target diagnosis group based on the patients of the target diagnosis group, and then display it through the display module, reducing the differences in the medical paths of the same diagnosis group among different departments and / or treatment groups.

[0081] This embodiment provides a patient-centered medical path optimization system, which can be deployed in electronic devices such as computers and servers. The electronic device deploying the patient-centered medical path optimization system includes a memory that stores computer programs related to the patient-centered medical path optimization system. In addition, the electronic device deploying the patient-centered medical path optimization system further includes a processor that is configured to execute the computer programs stored in the memory related to the patient-centered medical path optimization system to implement the patient-centered medical path optimization system. And / or,

[0082] The patient-centered medical path optimization system provided in this embodiment is deployed in a device connected to a computer-readable storage medium, and the computer-readable storage medium stores computer programs related to the patient-centered medical path optimization system. The computer programs related to the patient-centered medical path optimization system are executed by a processor to implement the patient-centered medical path optimization system.

[0083] As Figure 1 shown, the patient-centered medical path optimization system provided in this embodiment includes: a data storage module 101, a diagnosis group determination module 102, a medical path optimization module 103, and a display module 104.

[0084] 1. Data storage module 101

[0085] Among them, the data storage module is used to store all data related to the hospital, such as storing medical data and standard data of each diagnosis group.

[0086] The medical data is the data involved in the hospital diagnosis and treatment process, and the standard data is regularly released by relevant personnel.

[0087] For example, the data involved in the hospital diagnosis and treatment process is the data generated during the hospital operation, such as hospital patient data (such as name, gender, age, home address, contact information, etc.), the relevant data of each outpatient diagnosis and treatment of patients recorded by the outpatient-related systems in the hospital (such as Figure 2 shown admission records, test data, examination data, etc.), the relevant data of each inpatient diagnosis and treatment of patients recorded by the inpatient-related systems (such as inpatient cases, daily test data, daily examination data, daily medication data, etc.), and the respective relevant data recorded by each business system (such as the drug procurement-related data and drug distribution-related data recorded by the pharmacy-related business system).

[0088] The standard data can be generated by a hospital or non - hospital. The standard data is relevant data used to provide guidance, standards, benchmarks, norms, policies, opinions, notices, etc., such as DRG (Diagnosis Related Groups) / DIP (Disease Severity and Intervention Complexity Points) related data released by the medical management department, the standard data of each diagnosis group (such as the cost standard of each diagnosis group, the diagnosis and treatment duration standard of each diagnosis group, the relevant standards of the diagnosis and treatment path of each diagnosis group, etc.), the relevant data of centralized drug procurement, the implementation guidelines and standard norms announced by the hospital, etc.

[0089] In specific implementation, the data storage module 101 can be the memory of the electronic device that deploys the patient - centered medical path optimization system provided in this embodiment, or a hard disk, etc., or it can also be a cloud memory connected to the electronic device that deploys the patient - centered medical path optimization system provided in this embodiment.

[0090] 2. Diagnosis group determination module 102

[0091] The diagnosis group determination module is used to determine the target diagnosis group for which the medical path needs to be optimized based on the standard data and medical data of each diagnosis group.

[0092] The diagnosis group determination module can determine the target diagnosis group for which the medical path needs to be optimized based on the standard data and medical data of each diagnosis group through the following process.

[0093] 201. The diagnosis group determination module classifies the medical data by diagnosis group to obtain the medical data of each diagnosis group.

[0094] Each time a patient visits a doctor, information related to the patient's condition is recorded. In step 201, the diagnosis group determination module extracts the information related to the patient's condition from the medical data of each visit of each patient, determines the diagnosis group, and then classifies the medical data of each hospitalization of each patient by diagnosis group, so as to obtain the medical data of each diagnosis group.

[0095] For example, extract Figure 2 the present illness history, past history, etc. in, and through the DRG grouping scheme, obtain the diagnosis group of the medical data. Then classify the medical data by diagnosis group to obtain the medical data of each diagnosis group.

[0096] It should be noted that, on the one hand, due to the large amount of medical data, and on the other hand, due to the timeliness of medical data (for example, data from twenty years ago may no longer be of reference value to the current medical path due to the progress of drugs and medical means), the medical data mentioned in this embodiment can be medical data for a recent period of time, such as medical data for the most recent ten months, one year, three years, five years, etc., so as to ensure that the system provided in this embodiment can obtain the most accurate and optimized medical path. Similarly, the standard data also has timeliness, and the standard data mentioned in this embodiment is the latest standard data.

[0097] 202. The diagnosis group determination module determines the medical data belonging to each diagnosis group in each department according to the diagnosis group to which the medical data of each department belongs.

[0098] Since patients with the same condition can be treated by different departments. For example, patients with pulmonary mycosis are normally admitted to the Department of Respiratory Medicine, but patients with sudden severe symptoms will also be admitted to the Emergency Department. In this way, the Department of Respiratory Medicine will generate medical data belonging to the diagnosis group "ES2 Pulmonary Mycosis", and the Emergency Department will also generate medical data belonging to the diagnosis group "ES2 Pulmonary Mycosis". This makes it possible that the medical data belonging to the same diagnosis group can originate from different departments, as shown in Figure 3.

[0099] The existing problem is that there are significant differences in the treatment paths for the same diagnosis group of diseases in different departments. The system provided in this embodiment is to solve this problem. To solve this problem, in step 202, the diagnosis group determination module 102 determines the medical data belonging to each diagnosis group in each department according to the diagnosis group to which the medical data of each department belongs. In this way, the medical data of the same diagnosis group will be summarized at the department dimension, and then the medical data of different departments of the same diagnosis group will be compared to obtain the overall situation of the medical paths of each department for this diagnosis group. Then, according to the overall situation, the target diagnosis group is determined. In this way, the determination of the target diagnosis group is not only based on the medical path situation of the main department diagnosing the diseases related to this diagnosis group, but also comprehensively considers the overall medical path situation of all possible departments in the hospital for the diseases related to this diagnosis group, that is, it takes into account the differences between departments. Subsequently, when optimizing the medical path, an optimal medical path applicable to each department is given based on the overall situation. In this way, each department conducts diagnosis and treatment based on the optimal medical path, avoiding the differences in the treatment effects and the consumption of medical resources caused by the significant differences in the treatment paths for the same diagnosis group of diseases in different departments.

[0100] 203. The diagnosis group determination module determines the target patients for each diagnosis group in each department and the target doctors for each diagnosis group in each department.

[0101] Among them, the target patients for any diagnosis group i in any department j are the patients corresponding to the medical data belonging to diagnosis group i in department j, and the target doctors for diagnosis group i in department j are the doctors corresponding to the medical data belonging to diagnosis group i in department j.

[0102] For example, if any department j is the emergency department and any diagnosis group i is "ES2 pulmonary mycosis", then the target patients are the patients to whom the medical treatment data originating from the emergency department and with the disease condition belonging to "ES2 pulmonary mycosis" belong (such as the patients who have visited the emergency department for diseases related to "ES2 pulmonary mycosis"). The target doctors are the doctors to whom the medical treatment data originating from the emergency department and with the disease condition belonging to "ES2 pulmonary mycosis" belong (such as the doctors among all the doctors affiliated with the emergency department who have treated diseases related to "ES2 pulmonary mycosis").

[0103] In this way, the target patients are all the patients who have visited the hospital for diseases related to the diagnosis group "ES2 pulmonary mycosis", regardless of which department they visited. The target doctors are all the doctors who have treated diseases related to the diagnosis group "ES2 pulmonary mycosis", regardless of which department the doctor is affiliated with. All the patients and doctors of the diagnosis group "ES2 pulmonary mycosis" can be obtained.

[0104] 204. The diagnosis group determination module determines the probabilities of each diagnosis group for each department according to the medical data belonging to each diagnosis group in each department, the target patients and target doctors for each diagnosis group in each department.

[0105] For any diagnosis group i and any department j, the probability of diagnosis group i for department j is determined through the following steps:

[0106] 301. Determine the number of target patients NPAij of diagnosis group i in department j and the number of target doctors NDOij of diagnosis group i in department j.

[0107] For example, if any department j is the emergency department and any diagnosis group i is "ES2 pulmonary mycosis", then NPAij is the number of target patients of "ES2 pulmonary mycosis" in the emergency department, and NDOij is the number of target doctors of "ES2 pulmonary mycosis" in the emergency department.

[0108] 302. Based on the medical data, determine the total number of medical treatments of each target patient of diagnosis group i in department j, the number of medical treatments of each target patient of diagnosis group i in department j in department j, and the number of medical treatments belonging to diagnosis group i in the medical treatments of each target patient of diagnosis group i in department j in department j.

[0109] For example, if any department j is the emergency department, and any diagnosis group i is "ES2 pulmonary mycosis", patient A has visited three departments in the hospital, namely the emergency department, the orthopedics department, and the cardiovascular medicine department. Among them, the patient has visited the emergency department twice, once due to pulmonary mycosis and once due to a fracture. The patient has visited the orthopedics department once due to a fracture. The patient has visited the cardiovascular medicine department 12 times, 10 times due to premature beats and 2 times due to hyperlipidemia. Since the patient has visited the emergency department due to diseases related to "ES2 pulmonary mycosis", patient A is a target patient of "ES2 pulmonary mycosis" in the emergency department. Then:

[0110] 1. The total number of medical treatments of each target patient of diagnosis group i in department j

[0111] The total number of medical treatments of each target patient of diagnosis group i in department j is the total number of times the target patient of "ES2 pulmonary mycosis" in the emergency department has visited the hospital, regardless of whether they visited due to diseases related to "ES2 pulmonary mycosis" and regardless of whether they visited the emergency department.

[0112] For example, the total number of medical treatments of patient A is 2 + 1 + 12 = 15 times.

[0113] 2. The number of medical treatments of each target patient of diagnosis group i in department j in department j

[0114] The number of medical treatments of each target patient of diagnosis group i in department j in department j is the total number of times the target patient of "ES2 pulmonary mycosis" in the emergency department has visited the emergency department, regardless of whether they visited due to diseases related to "ES2 pulmonary mycosis".

[0115] For example, the number of medical treatments of patient A in the emergency department is 2 times.

[0116] 3. The number of medical treatments belonging to diagnosis group i of each target patient of diagnosis group i in department j during the medical treatment in department j

[0117] The number of medical treatments belonging to diagnosis group i of each target patient of diagnosis group i in department j during the medical treatment in department j is the number of times the target patient of "ES2 pulmonary mycosis" in the emergency department has visited the emergency department due to the cause of "ES2 pulmonary mycosis".

[0118] For example, the number of medical treatments of patient A belonging to diseases related to "ES2 pulmonary mycosis" in the emergency department is 1 time.

[0119] 303. Determine the weight of each target patient of diagnosis group i in department j for department j.

[0120] Among them, the weight Wuj of any target patient u of diagnosis group i in department j for department j = the number of medical treatments of target patient u in department j / the total number of medical treatments of target patient u.

[0121] Taking any department j as the emergency department, any diagnosis group i as "ES2 pulmonary mycosis", and any target patient u in the emergency department with the diagnosis group "ES2 pulmonary mycosis" as patient A as an example, then the weight Wuj of patient A for the emergency department = the number of times patient A was treated in the emergency department / the total number of times patient A was treated = 2 / 15.

[0122] Wuj represents the probability that the target patient u visits department j. The larger this value, the more concentrated the main symptoms of the target patient u are on the symptoms involved in department j, that is, the medical data of the target patient u is concentrated on department j. Then, the medical data of the target patient u represents the medical conditions and medical level of department j better.

[0123] 304. Determine the weights of each target patient in diagnosis group i of department j for diagnosis group i in department j.

[0124] Among them, the weight Wuij of any target patient u in diagnosis group i of department j for diagnosis group i in department j = the number of times the target patient u belongs to diagnosis group i in the treatment in department j / the total number of times the target patient u was treated.

[0125] Taking any department j as the emergency department, any diagnosis group i as "ES2 pulmonary mycosis", and any target patient u in the emergency department with the diagnosis group "ES2 pulmonary mycosis" as patient A as an example, then the weight Wuij of patient A for "ES2 pulmonary mycosis" in the emergency department = the number of times patient A belongs to the symptoms related to "ES2 pulmonary mycosis" in the emergency department / the total number of times patient A was treated = 1 / 15.

[0126] Wuij represents the probability that the target patient u visits department j due to the diseases related to diagnosis group i. The larger this value, the more concentrated the main symptoms of the target patient u are on the symptoms involved in diagnosis group i of department j, that is, the medical data of the target patient u is concentrated on diagnosis group i of department j. Then, the medical data of the target patient u represents the medical conditions and medical level of diagnosis group i of department j better.

[0127] 305. Determine the patient weight WU = ∑u(Wuij) / ∑u(Wuj).

[0128] The patient weight WU represents the probability that the patient visits due to the diseases related to diagnosis group i on the premise of visiting department j. Therefore, the patient weight WU = the average probability that the patients in diagnosis group i of department j visit due to the diseases related to diagnosis group i in department j / the average probability that the patients in diagnosis group i of department j visit department j.

[0129] The average probability that the patients in diagnosis group i of department j visit due to the diseases related to diagnosis group i in department j = ∑u(Wuij) / U, where U is the total number of target patients in diagnosis group i of department j.

[0130] The average probability of patients in diagnosis group i of department j seeking medical treatment in department j = ∑u(Wuj) / U.

[0131] Therefore, WU = [∑u(Wuij) / U] / [∑u(Wuj) / U] = ∑u(Wuij) / ∑u(Wuj).

[0132] 306. Determine the patient ratio Rij(1) = WU × [NPAij / the total number of patients treated in department j].

[0133] NPAij / the total number of patients treated in department j represents the proportion of patients with diseases related to diagnosis group i treated in department j. The larger this value, the more department j is a department for diagnosing diseases related to diagnosis group i. WU represents the probability that a patient seeks medical treatment due to diseases related to diagnosis group i on the premise of seeking medical treatment in department j, and represents the degree of manifestation of the patient for diseases related to diagnosis group i. Therefore, when determining the patient ratio, WU is used as the weight of the target patients in diagnosis group i in department j to correct NPAij / the total number of patients treated in department j. That is to say, if there are many patients with diseases related to diagnosis group i treated in department j, but these patients are not the main patients with diseases related to diagnosis group i (for example, the diseases related to diagnosis group i are only a complication caused by another main disease, and this time the patient happens to seek medical treatment due to the diseases related to diagnosis group i, then the medical treatment path reflected by the medical data of this patient will be affected by the medical treatment path of the main cause, and the degree of manifestation of the medical data of this patient for the medical treatment path of diseases related to diagnosis group i will be affected), then when determining the patient ratio Rij(1), the large number of patients also needs to be adjusted downward, and the adjustment basis is WU.

[0134] The patient ratio Rij(1) represents the probability that a patient in department j has diseases related to diagnosis group i. The larger this value, the more department j is the main department for diagnosing diseases related to diagnosis group i (that is, the department that should diagnose diseases related to diagnosis group i). For example, Figure 3 As shown, the departments that can diagnose diseases related to "ES2 pulmonary mycosis" are the Department of Rheumatology and Immunology, the Department of Infection and Clinical Microbiology, the Department of Respiratory and Critical Care Medicine, the Emergency Department, and the Department of Occupational Diseases and Poisoning Medicine. And the department that should diagnose diseases related to "ES2 pulmonary mycosis" is the Department of Respiratory and Critical Care Medicine. The patient ratio Rij(1) reflects the departments that diagnose diseases related to "ES2 pulmonary mycosis". The larger this value, the more likely it is to be a department for diagnosing diseases related to "ES2 pulmonary mycosis".

[0135] 307. Determine the professional titles of each target doctor in diagnosis group i of department j.

[0136] The professional titles of doctors can be determined according to the existing professional title evaluation standards.

[0137] For example, professional titles include: resident physician, attending physician, deputy chief physician, chief physician, etc. The evaluation criteria are as follows:

[0138] Resident physician: Have completed the standardized training for resident physicians and initially possess the ability to independently diagnose and treat common diseases.

[0139] Attending physician: Can handle relatively complex diseases and provide guidance to junior physicians.

[0140] Deputy chief physician: Have a relatively deep attainment in the professional field and can solve difficult and complicated diseases.

[0141] Chief physician: Be the academic leader in the professional field and have high authority in both academic and clinical aspects.

[0142] In specific implementation, the representation values of doctors' professional titles can be set in advance, and the value range of the representation values can be set in advance, such as 0 - 1, or 1 - 10, or 1 - 100, etc. As long as the higher the professional title, the larger the representation value.

[0143] For example, the value range of the representation value is 1 - 10, and the corresponding relationship between the professional title and the representation value is shown in Table 1.

[0144] Table 1

[0145] Professional title Characteristic value Resident doctor 1 Attending physician 4 Associate chief physician 7 Chief physician 10

[0146] 308. Based on the medical data, determine the total number of diagnoses of each target doctor in diagnosis group i of department j, the number of diagnoses of each target doctor in diagnosis group i of department j in department j, and the number of diagnoses of each target doctor in diagnosis group i of department j in department j that belong to diagnosis group i.

[0147] For example, any department j is the emergency department, and any diagnosis group i is "ES2 pulmonary mycosis". The emergency department includes 3 doctors, namely Doctor A, Doctor B, and Doctor C. Among them, Doctor A used to belong to the Department of Respiratory and Critical Care Medicine and was later transferred to the emergency department. Doctor A had 100 diagnoses in the Department of Respiratory and Critical Care Medicine and 3 diagnoses in the emergency department, and all 3 diagnoses in the emergency department were related to "ES2 pulmonary mycosis". Doctor B has always belonged to the emergency department. He had 2 diagnoses in the emergency department, one of which was related to "ES2 pulmonary mycosis" and the other was asthma. Doctor C has always belonged to the emergency department. He had 5 diagnoses in the emergency department, all of which were asthma. Since both Doctor A and Doctor B have diagnosed diseases related to "ES2 pulmonary mycosis", Doctor A and Doctor B are the target doctors of "ES2 pulmonary mycosis" in the emergency department. Then:

[0148] 1. The total number of diagnoses of each target doctor in diagnosis group i of department j

[0149] The total number of medical treatments for each target doctor in diagnostic group i of department j is the total number of medical treatments of doctor A and doctor B, regardless of whether they diagnose diseases related to "ES2 pulmonary mycosis" and regardless of whether the diagnosis is made in the emergency department.

[0150] For example, the total number of medical treatments of doctor A is 100 + 3 = 103 times, and the total number of medical treatments of doctor B is 2 times.

[0151] 2. The number of medical treatments in department j for each target doctor in diagnostic group i of department j

[0152] The number of medical treatments in department j for each target doctor in diagnostic group i of department j is the number of medical treatments of doctor A and doctor B in the emergency department, regardless of whether they diagnose diseases related to "ES2 pulmonary mycosis".

[0153] For example, the number of medical treatments of doctor A in the emergency department is 3 times, and the number of medical treatments of doctor B in the emergency department is 2 times.

[0154] 3. The number of medical treatments in department j for each target doctor in diagnostic group i of department j that belong to diagnostic group i

[0155] The number of medical treatments in department j for each target doctor in diagnostic group i of department j that belong to diagnostic group i is the number of times doctor A and doctor B diagnose diseases related to "ES2 pulmonary mycosis" in the emergency department.

[0156] For example, the number of medical treatments of doctor A in the emergency department that belong to "ES2 pulmonary mycosis" is 3 times, and the number of medical treatments of doctor B in the emergency department that belong to "ES2 pulmonary mycosis" is 1 time.

[0157] 309. Determine the weight of each target doctor in diagnostic group i of department j for department j.

[0158] Among them, the weight Wvj of any target doctor v in diagnostic group i of department j for department j = the number of medical treatments of target doctor v in department j / the total number of medical treatments of target doctor v.

[0159] Still taking any department j as the emergency department, any diagnostic group i as "ES2 pulmonary mycosis", and the target doctors in diagnostic group "ES2 pulmonary mycosis" in the emergency department as doctor A and doctor B as an example, then the weight Wvj of doctor A for the emergency department = the number of medical treatments of doctor A in the emergency department / the total number of medical treatments of doctor A = 3 / 103. The weight Wvj of doctor B for the emergency department = the number of medical treatments of doctor B in the emergency department / the total number of medical treatments of doctor B = 2 / 2.

[0160] Wvj represents the probability of target doctor v diagnosing and treating in department j. The larger the value is, the more representative the target doctor v's main work is in department j. In other words, the medical data of target doctor v is concentrated in department j. Therefore, the medical data of target doctor v is more representative of the medical condition and medical level of department j.

[0161] 310, determine the weight of each target doctor of diagnosis group i in department j for diagnosis group i in department j.

[0162] Among them, the weight of any target doctor v in diagnosis group i in department j to diagnosis group i in department j is Wvij = the number of diagnosis and treatment times of target doctor v in diagnosis group i in department j / the total number of diagnosis and treatment times of target doctor v.

[0163] Still taking any department j as the emergency department, any diagnosis group i as "ES2 pulmonary fungal disease", and the target doctors of the diagnosis group "ES2 pulmonary fungal disease" in the emergency department as doctors A and B as an example, then the weight Wvij of doctor A for "ES2 pulmonary fungal disease" in the emergency department = the number of times doctor A diagnosed and treated "ES2 pulmonary fungal disease" in the emergency department / the total number of times doctor A diagnosed and treated = 3 / 103. The weight Wvij of doctor B for "ES2 pulmonary fungal disease" in the emergency department = the number of times doctor B diagnosed and treated "ES2 pulmonary fungal disease" in the emergency department / the total number of times doctor B diagnosed and treated = 1 / 2.

[0164] Wvij represents the probability that the target doctor v diagnoses and treats diseases related to diagnosis group i in department j. The larger the value is, the more representative the target doctor v is in treating diseases related to diagnosis group i in department j. Then the medical data of target doctor v is in representing the medical condition and medical level of diagnosis group i in department j.

[0165] 311, determine the doctor weight WD = [∑v(Wvij) / ∑v(Wvj)].

[0166] The doctor weight WD represents the probability of the doctor diagnosing the relevant diseases of diagnosis group i under the premise of seeing the doctor in department j. Therefore, the doctor weight WD = the average probability of the doctor diagnosing and treating diseases related to "ES2 pulmonary fungal disease" in department j / the average probability of the doctor diagnosing and treating diseases in department j.

[0167] The average probability of a doctor treating "ES2 pulmonary fungal disease" related diseases in department j = ∑v(Wvij) / V, where V is the total number of all target doctors in diagnosis group i in department j.

[0168] The average probability of a doctor treating patients in department j = ∑v(Wvj) / V.

[0169] Therefore, WD=[∑v(Wvij) / V] / [∑v(Wvj) / V]=[∑v(Wvij) / ∑v(Wvj)].

[0170] 312. Determine the professional title weight WT = the highest professional title among all target doctors of diagnosis group i in department j / the highest professional title of all doctors in department j.

[0171] The professional title weight WT can be calculated based on the guaranteed value. For example, determine the maximum value V1 among the representation values corresponding to the professional titles of all target doctors of diagnosis group i in department j, and the maximum value V2 among the representation values corresponding to the professional titles of all doctors of diagnosis group i in department j. Determine the professional title weight WT = V1 / V2.

[0172] The professional title weight WT characterizes the situation of doctors related to the diseases of diagnosis group i in department j. The higher this value, the more it indicates that the diagnosis and treatment ability in department j tends to diagnosis group i, diagnosis group i is the key diagnosis group of department j, and the ability of the treating doctors in diagnosis group i is stronger.

[0173] 313. Determine the doctor ratio Rij(2) = [NDOij / the total number of doctors in department j] × (WD + WT) / 2.

[0174] NDOij / the total number of doctors in department j is the proportion of doctors in department j who diagnose diseases related to diagnosis group i. The larger this proportion, the more it indicates that department j is a department for diagnosing diseases related to diagnosis group i, and department j is more suitable for diagnosing diagnosis group i.

[0175] (WD + WT) / 2 is the average of the doctor weight WD and the professional title weight WT. The doctor weight WD describes the diagnostic ability of doctors related to the diseases of diagnosis group i in department j from the perspective of the number of doctor diagnoses. The higher WD, the higher the proportion of doctors diagnosing diseases related to diagnosis group i, and the more familiar and capable the doctors are in diagnosing diseases related to diagnosis group i. The professional title weight WT describes the diagnostic ability of doctors related to the diseases of diagnosis group i in department j from the perspective of doctor support. The higher WT, the stronger the ability of doctors diagnosing diagnosis group i. (WD + WT) / 2 comprehensively considers both the quantity and the professional title to obtain the diagnosis and treatment ability of doctors related to the diseases of diagnosis group i in department j for diseases related to diagnosis group i. The greater this diagnosis and treatment ability, the more research the doctors in department j have done on diseases related to diagnosis group i, and the more suitable department j is for diagnosing diseases related to diagnosis group i.

[0176] The doctor ratio Rij(2) reflects the department for diseases related to diagnosis group i from two aspects: the suitability of department j for diagnosis group i and the diagnostic ability of the treating doctors. The larger this value, the more likely it is to be a department for diseases related to diagnosis group i.

[0177] 314. Determine the probability Pij of diagnosis group i for department j = [Rij(1) + Rij(2)] / 2.

[0178] $P_{ij}$ represents the importance of diagnosis group $i$ for department $j$ from both the perspectives of doctors and patients. The larger this value, the more important diagnosis group $i$ is for department $j$, and the more department $j$ is a department for treating diseases related to diagnosis group $i$.

[0179] 205. The diagnosis group determination module determines the diagnosis and treatment optimization degree of each diagnosis group for each department according to the standard data of each diagnosis group, the medical data belonging to each diagnosis group in each department, the target patients and target doctors for each diagnosis group in each department.

[0180] For any diagnosis group $i$ and any department $j$, the diagnosis and treatment optimization degree of diagnosis group $i$ for department $j$ is determined through the following steps:

[0181] 401. Determine the matching degree between the medical data of each target patient of diagnosis group $i$ for department $j$ and the standard data of diagnosis group $i$.

[0182] The matching degree can be achieved through existing feature comparison schemes. For example, construct the feature indicators of patients in advance, such as the age, gender, condition, test results, examination results, diagnosis, length of hospital stay, daily diagnosis and treatment conditions, costs, medications, etc. of patients (the features here can adopt existing evaluation features). When the diagnosis group determination module executes step 401, first, based on the pre-constructed feature indicators of patients, obtain the feature vectors of the medical data of each target patient of diagnosis group $i$ for department $j$. Then, based on the pre-constructed feature indicators of patients, obtain the feature vectors related to patients in the standard data of diagnosis group $i$. Subsequently, calculate the distance (such as Euclidean distance) between the two feature vectors, and this distance value is the matching degree between the medical data of each target patient of diagnosis group $i$ for department $j$ and the standard data of diagnosis group $i$.

[0183] Taking diagnosis group $i$ as "ES2 pulmonary mycosis" as an example, if the departments that have diagnosed "ES2 pulmonary mycosis" are: Rheumatology and Immunology Department, Infectious Diseases and Clinical Microbiology Department, Respiratory and Critical Care Medicine Department, Emergency Department, Occupational Diseases and Poisoning Medicine Department, then in 401, the diagnosis group determination module will determine the feature vectors of the medical data of each target patient of "ES2 pulmonary mycosis" for the Rheumatology and Immunology Department (i.e., patients who have visited the Rheumatology and Immunology Department due to diseases related to "ES2 pulmonary mycosis"), and then calculate the distance (such as Euclidean distance) between this feature vector and the feature vector related to patients in the standard data of "ES2 pulmonary mycosis". This distance value is the matching degree $SIM1$ between the medical data of each target patient of "ES2 pulmonary mycosis" for the Rheumatology and Immunology Department and the standard data of "ES2 pulmonary mycosis".

[0184] The diagnosis group determination module also determines the feature vector of the medical data of each target patient of "ES2 pulmonary mycosis" for the Department of Infectious Diseases and Clinical Microbiology (i.e., patients who seek medical treatment in the Department of Infectious Diseases and Clinical Microbiology due to diseases related to "ES2 pulmonary mycosis"), and then calculates the distance (such as the Euclidean distance) between this feature vector and the feature vector related to the patient in the standard data of "ES2 pulmonary mycosis". This distance value is the matching degree SIM2 between the medical data of each target patient of "ES2 pulmonary mycosis" for the Department of Infectious Diseases and Clinical Microbiology and the standard data of "ES2 pulmonary mycosis".

[0185] The diagnosis group determination module also determines the feature vector of the medical data of each target patient of "ES2 pulmonary mycosis" for the Department of Respiratory and Critical Care Medicine (i.e., patients who seek medical treatment in the Department of Respiratory and Critical Care Medicine due to diseases related to "ES2 pulmonary mycosis"), and then calculates the distance (such as the Euclidean distance) between this feature vector and the feature vector related to the patient in the standard data of "ES2 pulmonary mycosis". This distance value is the matching degree SIM3 between the medical data of each target patient of "ES2 pulmonary mycosis" for the Department of Respiratory and Critical Care Medicine and the standard data of "ES2 pulmonary mycosis".

[0186] The diagnosis group determination module also determines the feature vector of the medical data of each target patient of "ES2 pulmonary mycosis" for the Emergency Department (i.e., patients who seek medical treatment in the Emergency Department due to diseases related to "ES2 pulmonary mycosis"), and then calculates the distance (such as the Euclidean distance) between this feature vector and the feature vector related to the patient in the standard data of "ES2 pulmonary mycosis". This distance value is the matching degree SIM4 between the medical data of each target patient of "ES2 pulmonary mycosis" for the Emergency Department and the standard data of "ES2 pulmonary mycosis".

[0187] The diagnosis group determination module also determines the feature vector of the medical data of each target patient of "ES2 pulmonary mycosis" for the Department of Occupational Diseases and Poisoning Medicine (i.e., patients who seek medical treatment in the Department of Occupational Diseases and Poisoning Medicine due to diseases related to "ES2 pulmonary mycosis"), and then calculates the distance (such as the Euclidean distance) between this feature vector and the feature vector related to the patient in the standard data of "ES2 pulmonary mycosis". This distance value is the matching degree SIM5 between the medical data of each target patient of "ES2 pulmonary mycosis" for the Department of Occupational Diseases and Poisoning Medicine and the standard data of "ES2 pulmonary mycosis".

[0188] 402. Determine the minimum value Simu(min) and the standard deviation Simu(sd) of the matching degree between the medical data of each target patient of diagnosis group i for department j and the standard data of diagnosis group i.

[0189] Still taking the diagnosis group i as "ES2 pulmonary mycosis" as an example, Simu(min) = min{SIM1, SIM2, SIM3, SIM4, SIM5}. Simu(sd) = sqrt{[(SIM1 - avgu)×(SIM1 - avgu) + (SIM2 - avgu)×(SIM2 - avgu) + (SIM3 - avgu)×(SIM3 - avgu) + (SIM4 - avgu)×(SIM4 - avgu) + (SIM5 - avgu)×(SIM5 - avgu)] / 5}.

[0190] Among them, min{} is the minimum value function, sqrt{} is the square root formula, and avgu is the average of the matching degrees between the medical data of each target patient of the diagnosis group i for the department j and the standard data of the diagnosis group i, that is, avgu = (SIM1 + SIM2 + SIM3 + SIM4 + SIM5) / 5.

[0191] 403. Determine the matching degree between the medical data of each target doctor of the diagnosis group i for the department j and the standard data of the diagnosis group i.

[0192] The determination scheme of the matching degree can be the same as the determination scheme in step 401, that is, it can be realized by the existing feature comparison scheme. For example, pre-construct the feature indicators of doctors, such as the affiliated department, the main diseases treated, education background, years of medical practice, professional title, number of operating tables, participation in scientific research projects, published academic papers, obtained scientific research achievements, etc. (the features here can adopt the existing evaluation features). When the diagnosis group determination module executes step 403, first, based on the pre-constructed feature indicators of doctors, obtain the feature vectors of the medical data of each target doctor of the diagnosis group i for the department j. Then, based on the pre-constructed feature indicators of doctors, obtain the doctor-related feature vectors in the standard data of the diagnosis group i. Subsequently, calculate the distance (such as the Euclidean distance) between the two feature vectors, and this distance value is the matching degree between the medical data of each target doctor of the diagnosis group i for the department j and the standard data of the diagnosis group i.

[0193] Taking the diagnostic group i as "ES2 pulmonary mycosis" as an example, if the departments that have diagnosed "ES2 pulmonary mycosis" are: Rheumatology and Immunology Department, Infectious and Clinical Microbiology Department, Respiratory and Critical Care Medicine Department, Emergency Department, Occupational Disease and Poisoning Medicine Department, then in 403, the diagnostic group determination module will determine the feature vectors of the medical data of each target doctor of "ES2 pulmonary mycosis" for the Rheumatology and Immunology Department (i.e., the doctors who diagnose "ES2 pulmonary mycosis" related diseases in the Rheumatology and Immunology Department), and then calculate the distance (such as Euclidean distance) between this feature vector and the doctor-related feature vector in the standard data of "ES2 pulmonary mycosis". This distance value is the matching degree SIM6 between the medical data of each target doctor of "ES2 pulmonary mycosis" for the Rheumatology and Immunology Department and the standard data of "ES2 pulmonary mycosis".

[0194] The diagnostic group determination module will also determine the feature vectors of the medical data of each target doctor of "ES2 pulmonary mycosis" for the Infectious and Clinical Microbiology Department (i.e., the doctors who diagnose "ES2 pulmonary mycosis" related diseases in the Infectious and Clinical Microbiology Department), and then calculate the distance (such as Euclidean distance) between this feature vector and the doctor-related feature vector in the standard data of "ES2 pulmonary mycosis". This distance value is the matching degree SIM7 between the medical data of each target doctor of "ES2 pulmonary mycosis" for the Infectious and Clinical Microbiology Department and the standard data of "ES2 pulmonary mycosis".

[0195] The diagnostic group determination module will also determine the feature vectors of the medical data of each target doctor of "ES2 pulmonary mycosis" for the Respiratory and Critical Care Medicine Department (i.e., the doctors who diagnose "ES2 pulmonary mycosis" related diseases in the Respiratory and Critical Care Medicine Department), and then calculate the distance (such as Euclidean distance) between this feature vector and the doctor-related feature vector in the standard data of "ES2 pulmonary mycosis". This distance value is the matching degree SIM8 between the medical data of each target doctor of "ES2 pulmonary mycosis" for the Respiratory and Critical Care Medicine Department and the standard data of "ES2 pulmonary mycosis".

[0196] The diagnostic group determination module will also determine the feature vectors of the medical data of each target doctor of "ES2 pulmonary mycosis" for the Emergency Department (i.e., the doctors who diagnose "ES2 pulmonary mycosis" related diseases in the Emergency Department), and then calculate the distance (such as Euclidean distance) between this feature vector and the doctor-related feature vector in the standard data of "ES2 pulmonary mycosis". This distance value is the matching degree SIM9 between the medical data of each target doctor of "ES2 pulmonary mycosis" for the Emergency Department and the standard data of "ES2 pulmonary mycosis".

[0197] The diagnosis group determination module also determines the feature vectors of the medical data of "ES2 pulmonary mycosis" for each target doctor in the Department of Occupational Diseases and Poisoning Medicine (i.e., the doctor who diagnoses and treats diseases related to "ES2 pulmonary mycosis" in the Department of Occupational Diseases and Poisoning Medicine), and then calculates the distance (such as the Euclidean distance) between the feature vector and the doctor-related feature vector in the standard data of "ES2 pulmonary mycosis". This distance value is the matching degree SIM10 between the medical data of "ES2 pulmonary mycosis" for each target doctor in the Department of Occupational Diseases and Poisoning Medicine and the standard data of "ES2 pulmonary mycosis".

[0198] 404. Determine the minimum value Simv(min) and the standard deviation Simv(sd) of the matching degree between the medical data of each target doctor in diagnosis group i for department j and the standard data of diagnosis group i.

[0199] Still taking diagnosis group i as "ES2 pulmonary mycosis" as an example, Simv(min) = min{SIM6, SIM7, SIM8, SIM9, SIM10}. Simv(sd) = sqrt{[(SIM6 - avgv)×(SIM6 - avgv) + (SIM7 - avgv)×(SIM7 - avgv) + (SIM8 - avgv)×(SIM8 - avgv) + (SIM9 - avgv)×(SIM9 - avgv) + (SIM10 - avgv)×(SIM10 - avgv)] / 5}.

[0200] Among them, min{} is the minimum value function, sqrt{} is the square root formula, avgv is the average value of the matching degree between the medical data of each target doctor in diagnosis group i for department j and the standard data of diagnosis group i, that is, avgv = (SIM6 + SIM7 + SIM8 + SIM9 + SIM10) / 5.

[0201] 405. Determine the diagnosis and treatment optimization degree ODij of diagnosis group i for department j = min{Simu(min) / patient matching degree threshold, Simv(min) / doctor matching degree threshold, 1 - [Simu(sd) / standard deviation threshold], 1 - [Simv(sd) / standard deviation threshold]}.

[0202] Among them, min{} is the minimum value function.

[0203] The patient matching degree threshold represents the minimum matching degree of normal patients when department j diagnoses diseases related to diagnosis group i (that is, the minimum matching degree of patients suffering from diseases related to diagnosis group i and whose diseases related to diagnosis group i need to be diagnosed in department j). Patients with a matching degree not less than this patient matching degree threshold indicate that the patient truly needs to be diagnosed in department j for diseases related to diagnosis group i; patients with a matching degree lower than this patient matching degree threshold indicate that the patient does not need to be diagnosed in department j for diseases related to diagnosis group i. One possible reason for this is that the patient does not mainly suffer from diseases related to diagnosis group i (such as incorrect input of case or other diseases, or the diseases related to diagnosis group i are complications caused by other etiologies), and another reason is that the diseases related to diagnosis group i are not the main diseases diagnosed by department j (for example, the emergency department is not the main department for "ES2 pulmonary mycosis", but the patient is diagnosed in this department only due to special reasons (such as suddenly showing obvious symptoms and urgently needing medical treatment)). For example, the larger the Simu(min) / patient matching degree threshold, the more normal the target patients of diagnosis group i for department j are, indicating that from the patient dimension, the current diagnosis of diseases related to diagnosis group i in department j is normal and no treatment optimization is required.

[0204] The doctor matching degree threshold represents the minimum matching degree of normal doctors in department j diagnosing diseases related to diagnosis group i (that is, department j is the main department for diagnosing diseases related to diagnosis group i, and the minimum matching degree of doctors who mainly diagnose diseases related to diagnosis group i). Doctors with a matching degree not less than this doctor matching degree threshold indicate that they are doctors good at diagnosing diseases related to diagnosis group i in department j; doctors with a matching degree lower than this doctor matching degree threshold indicate that they are not doctors diagnosing diseases related to diagnosis group i in department j. One possible reason for this is that the doctor does not mainly diagnose diseases related to diagnosis group i (for example, a doctor in the Department of Respiratory and Critical Care Medicine mainly studies asthma and can perform simple diagnoses for diseases related to "ES2 pulmonary mycosis", but for in-depth and accurate diagnoses, it still needs to be transferred to a doctor mainly studying diseases related to "ES2 pulmonary mycosis"), and another reason is that the diseases related to diagnosis group i are not the main diseases diagnosed by department j (for example, the emergency department is not the main department for "ES2 pulmonary mycosis", but the patient is diagnosed in this department only due to special reasons (such as suddenly showing obvious symptoms and seeking medical treatment)). For example, the larger the Simv(min) / doctor matching degree threshold, the more normal the target doctors of diagnosis group i for department j are, indicating that from the doctor dimension, the current diagnosis of diseases related to diagnosis group i in department j is normal and no treatment optimization is required.

[0205] The standard deviation threshold characterizes the fluctuations in the matching degrees between normal patients / doctors. If the fluctuations in the matching degrees among all target patients / target doctors of diagnosis group i for department j are lower than the standard deviation threshold, it indicates that the fluctuations of the target patients / target doctors are normal; otherwise, they are abnormal. The smaller the Simu(sd) / standard deviation threshold, the more normal the fluctuations in the matching degrees among the target patients of diagnosis group i for department j. At this time, the larger the 1 - [Simu(sd) / standard deviation threshold]. That is to say, the larger the 1 - [Simu(sd) / standard deviation threshold], the more normal it is for the current department j to diagnose the diseases related to diagnosis group i from the patient dimension, and no diagnosis and treatment optimization is required. The smaller the Simv(sd) / standard deviation threshold, the more normal the fluctuations in the matching degrees among the target doctors of diagnosis group i for department j. At this time, the larger the 1 - [Simv(sd) / standard deviation threshold]. That is to say, the larger the 1 - [Simv(sd) / standard deviation threshold], the more normal it is for the current department j to diagnose the diseases related to diagnosis group i from the doctor dimension, and no diagnosis and treatment optimization is required.

[0206] Based on the above analysis, comprehensively evaluate the situation of diagnosing the diseases related to diagnosis group i in the previous department j from four aspects: the matching degree of target patients, the matching degree of target doctors, the fluctuations in the matching degree of target patients, and the fluctuations in the matching degree of target doctors. As long as the value in one aspect is lower, it indicates that this aspect reflects abnormal characteristics, and it is determined that diagnosis and treatment optimization is required. Therefore, take the minimum value of the matching degree of target patients, the matching degree of target doctors, the fluctuations in the matching degree of target patients, and the fluctuations in the matching degree of target doctors in the four aspects as the diagnosis and treatment optimization degree of diagnosis group i for department j.

[0207] 206. The diagnosis group determination module determines the diagnosis and treatment optimization degrees of each diagnosis group according to the probabilities of each diagnosis group for each department and the diagnosis and treatment optimization degrees of each diagnosis group for each department.

[0208] Among them, the diagnosis and treatment optimization degree ODi of any diagnosis group i = ∑j(ODij × Pij) / J. Where J is the total number of departments, ODij is the diagnosis and treatment optimization degree of diagnosis group i for department j, and Pij is the probability of diagnosis group i for department j.

[0209] In step 205, the diagnosis optimization degrees of any diagnosis group i in the diagnosis of each department are obtained (that is, the abnormal degrees of any diagnosis group i in the diagnosis of each department). In step 204, the probabilities of any diagnosis group i for each department are obtained (that is, the importance degrees of any diagnosis group i for department j). ODij × Pij comprehensively considers the importance degree of diagnosis group i for department j and the abnormal degree of diagnosis group i in the diagnosis of each department, and obtains the diagnosis and treatment optimization degree of diagnosis group i for department j. ODi obtains the average diagnosis and treatment optimization degree of diagnosis group i for all departments. The lower this value is, the more optimization is required.

[0210] 207. The diagnosis group determination module determines the diagnosis group with a medical treatment optimization degree less than the preset threshold as the target diagnosis group.

[0211] If there are multiple diagnosis groups with a medical treatment optimization degree less than the preset threshold, the diagnosis group determination module determines all of them as the target diagnosis groups.

[0212] 3. Medical path optimization module 103

[0213] The medical path optimization module is used to optimize the medical path of the target diagnosis group based on the patients in the target diagnosis group.

[0214] If there are multiple target diagnosis groups, the medical path optimization module will optimize the medical path for each target diagnosis group respectively based on its patients.

[0215] For example, for any target diagnosis group, the process of the medical path optimization module optimizing the medical path of the target diagnosis group based on the patients in the target diagnosis group is as follows:

[0216] 501. Determine the medical data of the target diagnosis group as the target medical data.

[0217] The medical data here includes all medical data such as patients and doctors. As long as the medical data related to the target diagnosis group is the target medical data.

[0218] 502. According to the target medical data, determine the medical paths of each optimizing doctor for each optimizing patient, and the standard degree of each medical path.

[0219] Among them, the optimizing doctor is the doctor involved in the target medical data, and the optimizing patient is the patient involved in the medical data of the target diagnosis group.

[0220] In addition, the medical path is composed of medical nodes arranged in the order of medical treatment.

[0221] For example, if the target medical data is the data of a certain hospitalization of patient 1, the attending doctor during hospitalization is doctor 1, and the patient was hospitalized for 5 days, then the optimizing patient is patient 1, the optimizing doctor is doctor 1, and each day corresponds to a medical node. Then the medical path is composed of 5 nodes corresponding to 5 days, arranged in the order of medical treatment (i.e., the admission date), as Figure 4 shown. Each medical node will correspond to the relevant medical data of that day.

[0222] The standard degree of each medical path can be determined by the matching degree between the medical path and its corresponding standard path.

[0223] The matching degree can also be achieved through existing feature comparison schemes. For example, the characteristic indicators of the medical path are pre-constructed, such as the patient's age, condition, test results, examination results, diagnosis, length of hospital stay, daily diagnosis and treatment conditions, expenses, medications, the professional title of the attending doctor, etc. (the features here can adopt existing evaluation features). When the medical path optimization module executes step 502, first, based on the pre-constructed characteristic indicators of the medical path, the characteristic vector of a certain medical path is obtained. Then, based on the pre-constructed characteristic indicators of the medical path, the standard characteristic indicators of a certain medical path are obtained (the characteristic indicators can be extracted from relevant standard data based on the diagnosis group, doctor, patient, etc. involved in a certain medical path). Subsequently, the distance (such as the Euclidean distance) between the two calculated characteristic vectors is calculated, and this distance value is the standard degree of this medical path.

[0224] Among them, the standard characteristic indicators represent the standard path for doctors to diagnose diseases related to the target diagnosis group (such as the standard length of hospital stay for doctors to diagnose diseases related to the target diagnosis group, the examination content and medication conditions every day, etc.). The standard characteristic indicators only reflect the situation of the recommended medical path. The recommended medical path is only guiding and not necessarily to be executed. Therefore, the recommended medical path can be the optimal path or the common path for general patients, etc.

[0225] The medical path of any optimizing doctor for any optimizing patient is the real diagnosis and treatment data of the doctor's treatment of the patient. The standard degree of this medical path is the matching degree between the real data of this diagnosis and treatment and the standard. The lower the standard degree of the medical path, the lower the matching degree. This low may be caused by problems in this diagnosis and treatment or by the differences of patients.

[0226] 503. Obtain the medical path of the optimized target diagnosis group according to the medical path whose standard degree is not less than the standard degree threshold.

[0227] The medical path optimization module can obtain the medical path of the optimized target diagnosis group through the following steps.

[0228] 601. Construct a path graph of the target disease according to all target medical paths.

[0229] Among them, the target medical path is the medical path whose standard degree is not less than the standard degree threshold.

[0230] The points in the path graph correspond one by one to the medical nodes in all medical paths, and the edges between the nodes in the path graph correspond to the diagnosis and treatment order between the medical nodes in the medical path.

[0231] Any point in the path graph has a point attribute, and the point attribute value of any node is the number of medical paths containing that node.

[0232] Each edge in the path graph has an edge attribute, and the edge attribute value of each edge is the number of medical paths having the execution order between two points connected by each edge.

[0233] The target diagnosis group is the diagnosis group whose diagnosis and treatment optimization degree is less than the preset threshold, that is, the diagnosis group that needs to be optimized for the medical pathway. The medical pathway is the medical pathway for each diagnosis and treatment in the target diagnosis group. The target medical pathway (i.e., the medical pathway with a standard degree not less than the standard degree threshold) is the medical pathway in the target diagnosis group that meets the standard conditions. The pathway diagram brings together all the medical pathways in the target diagnosis group that meet the standard conditions, so the diagnosis and treatment conditions that meet the standards for the target diagnosis group can be obtained through the pathway diagram.

[0234] Since there are differences between different patients, if the distance between the features of the medical data of a certain node in multiple target medical pathways is less than a certain threshold, it means that the diagnosis and treatment conditions of the corresponding patients at the node are the same, and the corresponding nodes can be merged. For example, one target medical pathway is node 1->node 2->node 3, one target medical pathway is node 4->node 5->node 6->node 7, and the other target medical pathway is node 8->node 9->node 10. Wherein, both node 1 and node 8 include: electrocardiogram and routine blood test; node 4 includes: electrocardiogram, routine blood test and fasting blood glucose; according to the degree of influence of fasting blood glucose on the treatment of related diseases of the target diagnosis group and the degree of relevance of the examination items of node 1, node 4 and node 8, it is determined that the distance between the features of the medical data of node 1, node 4 and node 8 is not less than the threshold, then they can be merged (such as merged into node 11, node 11 includes the medical data of node 1, node 4 and node 8). Similarly, if the distance between the features of the medical data of node 5 and node 9 is not less than the threshold, then node 5 and node 9 can also be merged into node 12 (node ​​12 includes the medical data of node 5 and node 9), and the result is Figure 5 The path diagram shown.

[0235] In this way, all nodes in the medical pathways that meet the standard conditions can be merged to obtain a path graph. Therefore, the points in the path graph correspond one-to-one to the medical nodes in all medical pathways (except that a certain point in the path graph may correspond to medical nodes in multiple medical pathways, that is, a certain node in the path graph may be merged from multiple medical nodes), and the edges between the nodes in the path graph correspond to the diagnosis and treatment sequence between the medical nodes in the medical pathway. Also because a certain node in the path graph may be merged from multiple medical nodes, therefore, any point in the path graph has point attributes, and the point attribute value of any node is the number of medical paths containing any node (that is, the number of merged diagnosis and treatment nodes, such as Figure 5Among them, the point attribute of node 11 is 3 because it is merged from nodes 1, 4, and 8, and the point attribute of node 12 is 2 because it is merged from nodes 5 and 9). Any edge in the path graph has an edge attribute, and the edge attribute value of any edge is the number of medical paths of the execution order between the two points connected by any edge (that is, the number of merged edges, such as Figure 5 Among them, the edge attribute of the edge between node 11 and node 12 is 2 because it is merged from the edges of node 4 and node 5 and the edges of node 8 and node 9).

[0236] The path graph reflects the diagnostic paths that meet the standards for the target diagnosis group in the hospital. Subsequently, the medical paths of the optimized target diagnosis group will be obtained in steps 602 to 606.

[0237] 602. Among the points corresponding to the first medical node of all target medical paths in the path graph, the point with the largest point attribute value is used as the starting node.

[0238] The starting node is the most frequently used starting node among all diagnostic paths that meet the standards for the target diagnosis group, that is, the first medical node used by most diagnostic paths among all diagnostic paths that meet the standards.

[0239] 603. Among the points corresponding to the last medical node of all target medical paths in the path graph, the point with the largest point attribute value is used as the ending node.

[0240] The ending node is the most frequently used ending node among all diagnostic paths that meet the standards for the target diagnosis group, that is, the last medical node used by most diagnostic paths among all diagnostic paths that meet the standards.

[0241] 604. Determine the target link in the path graph.

[0242] Among them, the starting point of any target link is the starting node, and the ending point is the ending node.

[0243] The target link is all the links that start with the starting node and end with the ending node among all diagnostic paths that meet the standards for the target diagnosis group.

[0244] 605. Determine the link value of each target link according to the point attribute value and edge attribute value of the target link.

[0245] For example, for any target link l, its link value is determined through the following steps:

[0246] 701. For any edge of the target link l, determine its edge value as the edge attribute value of any edge × (the average value of the point attribute values of the two points connected by any edge).

[0247] For example, any target link l is node 11 -> node 12 -> node 10. For any edge between node 11 and node 12, its edge value = the edge attribute value of the edge between node 11 and node 12 × (the average of the node attribute values of node 11 and node 12).

[0248] If the node attribute value of node 11 is 3, the node attribute value of node 12 is 2, and the edge attribute value of the edge between node 11 and node 12 is 2, then the edge value of the edge between node 11 and node 12 is 2 × [(3 + 2) / 2] = 5.

[0249] The edge value characterizes the degree to which the edge and the two involved nodes are adopted by the medical paths that meet the standards. The higher this value, the more likely the edge and the two involved nodes are to appear in the optimized medical path.

[0250] 702. Determine the average value Eavg(l) and the standard deviation Esd(l) of the edge values of all edges of the target link l.

[0251] Taking any target link l as an example of node 11 -> node 12 -> node 10, Eavg(l) = (the edge value of the edge between node 11 and node 12 + the edge value of the edge between node 12 and node 10) / 2, and the standard deviation Esd(l) = sqrt{[(the edge value of the edge between node 11 and node 12 - Eavg(l)) × (the edge value of the edge between node 11 and node 12 - Eavg(l)) + (the edge value of the edge between node 12 and node 10 - Eavg(l)) × (the edge value of the edge between node 12 and node 10 - Eavg(l))] / 2}

[0252] 703. Determine that the link value of the target link l is Esd(l) / Eavg(l) × the typicality of the target link l.

[0253] Among them, the typicality of the target link l = [(the professional title of the doctor corresponding to the target link l / the highest professional title of all doctors corresponding to the target link) + the matching degree between the medical data corresponding to the target link l and the standard data of the diagnostic group i] / 2.

[0254] Esd(l) / Eavg(l) characterizes the degree of difference between the edge values of each edge in the target link l. The larger this value is, the greater the degree of difference between the edge values of each edge. For example, some edges may have a greater possibility of appearing in the optimized medical path, while some edges may have a smaller possibility of appearing in the optimized medical path. In such a link, some edges are adopted by most of the medical paths that meet the standards, while some edges are not adopted by most of the medical paths that meet the standards. Such a path will affect the possibility of being adopted due to some atypical edges. The smaller this value is, the more balanced the possibilities of each edge appearing in the optimized medical path are. The edges in such a link are more likely to be adopted by the medical paths that meet the standards on average. Such a path is more likely to be adopted because of the overall balance of the medical path.

[0255] The typicality of the target link l evaluates the matching situation of the target link with the standard from two aspects: doctors and diagnosis and treatment situations. The larger this value is, the better the matching.

[0256] Therefore, the link value of the target link l characterizes whether the target link l is suitable as an optimized medical path from the degree of matching with the standard and the possibility of the overall adoption of the target link l.

[0257] 606. According to the link value, obtain the medical path of the optimized target diagnosis group.

[0258] The implementation process of this step is as follows:

[0259] 701. Determine the patients corresponding to each target link.

[0260] 702. Cluster the corresponding patients according to the patient attributes.

[0261] For example, for any target link l, this step can obtain the feature vectors of each patient corresponding to the target link l based on the pre-constructed feature indicators of the patients (such as the feature indicators of the patients in step 401), and cluster each patient corresponding to the target link l according to the feature vectors, clustering the similar patients into one category. In this way, each category of patients is composed of patients with similar features (such as similar diseases, similar ages, etc.).

[0262] 703. Determine the standard deviation of the link values of all target links in each category.

[0263] The standard deviation of the link values of all target links in each category characterizes the degree of variation of the link values of all target links corresponding to patients with similar characteristics. The larger this value, the greater the fluctuation of each link of the corresponding characteristic patients as the optimized medical path (that is, some are more suitable as the optimized medical path, and some are less suitable as the optimized medical path). The optimized medical path is used to standardize the medical paths of the same diagnosis group in the hospital and avoid large differences in the diagnosis paths of the same diagnosis group in different departments. Therefore, the links of similar patients should be more similar, that is, the fluctuation is not large. Therefore, if the standard deviation of the link values of all target links in a category is large, it means that its fluctuation is large and it is not suitable as the optimized medical path.

[0264] 704. For the category whose standard deviation is not greater than the standard deviation threshold, the target link with the largest link value in this category is determined as the medical path of the optimized target diagnosis group.

[0265] Among them, the standard deviation threshold is set in advance and can be obtained through training with sample data or according to empirical values. For example, the standard deviation threshold is a value between 1 and 3.

[0266] If the category has a standard deviation not greater than the standard deviation threshold, it means that the fluctuation between the medical treatment paths of the patients in this category is not large. Then, the target link with the largest link value (that is, the highest possibility of being adopted) in this category is determined as the medical path of the optimized target diagnosis group.

[0267] 705. For the category whose standard deviation is greater than the standard deviation threshold, the target link with the highest typicality in this category is determined as the medical path of the optimized target diagnosis group.

[0268] If the category has a standard deviation greater than the standard deviation threshold, it means that the fluctuation between the medical treatment paths of the patients in this category is relatively large. Then, the target link with the highest typicality in this category can be selected as the medical path of the optimized target diagnosis group.

[0269] The evaluation of typicality can be determined by relevant personnel. For example, the target link selected by expert voting, or the target link of the leader.

[0270] In this embodiment, the medical path is optimized by diagnosis group, which can unify the medical paths of the same diagnosis group in different departments, avoiding the large differences in medical paths for diseases of the same diagnosis group in different departments in the prior art and the resulting differences in diagnosis and treatment. In addition, when optimizing the medical path of the same diagnosis group, different treatment plans will be obtained for different patients. Therefore, patients will be clustered, with each cluster being a type of patient, and an optimized medical path for the target diagnosis group will be obtained for each cluster, achieving the unification of medical paths for different types of patients with the same diagnosis group in different departments. Simply optimizing by diagnosis group may result in a mismatch with patient characteristics, and it is very likely that the optimized medical path cannot be promoted. Therefore, the system provided in this embodiment can obtain a more implementable and benchmark-quality optimized medical path for the target diagnosis group based on the diagnosis group and patient characteristics.

[0271] In addition, the path diagram reflects the diagnostic path that meets the standards for the target diagnosis group in the hospital. By obtaining the optimized medical path for the target diagnosis group through the path diagram, it not only meets the standards of the target diagnosis group but also conforms to the resource allocation and diagnosis and treatment capabilities of the target diagnosis group in the hospital, ensuring the standardization and practicability of the optimized medical path for the target diagnosis group.

[0272] 4. Display module 104

[0273] The display module is used to display medical data, standard data, the medical paths of each diagnosis group, etc.

[0274] For example, medical data includes all medical-related data such as doctor information, patient information, case information, department distribution, drug information, etc.

[0275] As shown in the display module Figure 6 the patient distribution situation, Figure 7 the case situation in the medical data, Figure 8 the doctor distribution situation, Figure 9 the DRG department distribution situation, etc.

[0276] The standard data can be announcements, data, etc. issued by relevant units. The medical paths of each diagnosis group include the medical paths of each doctor for each patient, the optimized medical paths, etc.

[0277] The display module can display all data. In specific implementation, different data can be displayed according to conditions such as the level and authorization of the display object. For example, only display the relevant data of a certain doctor's department for a certain doctor, etc., to achieve flexible and secure display of data.

[0278] It should be noted that in this embodiment, each example is only used to explain the implementation process of the system provided in this embodiment, and none of them are real data. In specific implementation, real data processing can be adopted.

[0279] This embodiment relates to a patient-centered medical path optimization system, which includes: a data storage module, a diagnosis group determination module, a medical path optimization module, and a display module. Among them, the data storage module is used to store medical data and standard data of each diagnosis group; the medical data is the data involved in the hospital diagnosis and treatment process, and the standard data is regularly released by relevant personnel; the diagnosis group determination module is used to determine the target diagnosis group to be optimized for the medical path based on the standard data and medical data of each diagnosis group; the medical path optimization module is used to optimize the medical path of the target diagnosis group based on the patients in the target diagnosis group; the display module is used to display medical data, standard data, and the medical paths of each diagnosis group. The system of this embodiment can optimize the medical path of the target diagnosis group based on the patients in the target diagnosis group, and then display it, reducing the difference in the medical path of the same diagnosis group among different departments and / or treatment groups.

[0280] It should be clear that the present invention is not limited to the specific configurations and processes described above and shown in the figures. For the sake of brevity, the detailed description of known methods or systems is omitted here. In the above embodiments, several specific steps are described and shown as examples. However, the system process of the present invention is not limited to the specific modules and steps described and shown. Those skilled in the art can make various changes, modifications, and additions after understanding the spirit of the present invention, or change the order of the steps.

[0281] It should also be noted that in the exemplary embodiments mentioned in the present invention, some methods or systems are described based on a series of steps or devices. However, the present invention is not limited to the order of the above steps, that is, the steps can be executed in the order mentioned in the embodiments, or different from the order in the embodiments, or several steps can be executed simultaneously.

[0282] Finally, it should be noted that the above-described embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions described in the foregoing embodiments, or perform equivalent replacements on some or all of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A patient-centered medical path optimization system, characterized in that, The system includes: a data storage module, a diagnosis group determination module, a medical path optimization module, and a display module; Among them, the data storage module is used to store medical data and the standard data of each diagnosis group; the medical data is the data involved in the hospital diagnosis and treatment process, and the standard data is regularly released by relevant personnel; The diagnosis group determination module is used to determine the target diagnosis group to be optimized for the medical path based on the standard data and medical data of each diagnosis group; The medical path optimization module is used to optimize the medical path of the target diagnosis group based on the patients of the target diagnosis group; The display module is used to display medical data, standard data, and the medical paths of each diagnosis group.

2. The system according to claim 1, wherein The determination of the target diagnosis group to be optimized for the medical path based on the standard data and medical data of each diagnosis group includes: Classify the medical data by diagnosis group to obtain the medical data of each diagnosis group; Determine the medical data belonging to each diagnosis group in each department according to the diagnosis group to which the medical data of each department belongs; Determine the target patients for each diagnosis group in each department and the target doctors for each diagnosis group in each department; among them, for any diagnosis group i in any department j, the target patients are the patients corresponding to each medical data belonging to diagnosis group i in department j, and the target doctor for diagnosis group i in department j is the doctor corresponding to each medical data belonging to diagnosis group i in department j; Determine the probability of each diagnosis group for each department according to the medical data belonging to each diagnosis group in each department, the target patients for each diagnosis group in each department, and the target doctors; Determine the diagnosis and treatment optimization degree of each diagnosis group for each department according to the standard data of each diagnosis group, the medical data belonging to each diagnosis group in each department, the target patients for each diagnosis group in each department, and the target doctors; Determine the diagnosis and treatment optimization degree of each diagnosis group according to the probability of each diagnosis group for each department and the diagnosis and treatment optimization degree of each diagnosis group for each department; among them, the diagnosis and treatment optimization degree ODi of any diagnosis group i = ∑j(ODij × Pij) / J; where J is the total number of departments, ODij is the diagnosis and treatment optimization degree of diagnosis group i for department j, and Pij is the probability of diagnosis group i for department j; Determine the diagnosis group with a diagnosis and treatment optimization degree less than the preset threshold as the target diagnosis group.

3. The system according to claim 2, wherein The determination of the probability of each diagnosis group for each department according to the medical data belonging to each diagnosis group in each department, the target patients for each diagnosis group in each department, and the target doctors includes: For any diagnosis group i and any department j, determine the probability of diagnosis group i for department j through the following steps: Determine the number NPAij of target patients of diagnosis group i in department j and the number NDOij of target doctors of diagnosis group i in department j; Based on the medical data, determine the total number of diagnoses and treatments of each target patient of diagnosis group i in department j, the number of diagnoses and treatments of each target patient of diagnosis group i in department j in department j, and the number of diagnoses and treatments belonging to diagnosis group i in the diagnoses and treatments of each target patient of diagnosis group i in department j in department j; Determine the weight of each target patient in diagnosis group i of department j for department j; among them, the weight Wuj of any target patient u in diagnosis group i of department j for department j = the number of times target patient u is treated in department j / the total number of times target patient u is treated; Determine the weight of each target patient in diagnosis group i of department j for diagnosis group i in department j; among them, the weight Wuij of any target patient u in diagnosis group i of department j for diagnosis group i in department j = the number of times target patient u's treatment in department j belongs to diagnosis group i / the total number of times target patient u is treated; Determine the patient weight WU = ∑u(Wuij) / ∑u(Wuj); Determine the patient ratio Rij(1) = WU × [NPAij / the total number of all patients treated in department j]; Determine the professional titles of each target doctor in diagnosis group i of department j; Based on the medical data, determine the total number of times each target doctor in diagnosis group i of department j is treated, the number of times each target doctor in diagnosis group i of department j is treated in department j, and the number of times each target doctor in diagnosis group i of department j is treated in department j and belongs to diagnosis group i; Determine the weight of each target doctor in diagnosis group i of department j for department j; among them, the weight Wvj of any target doctor v in diagnosis group i of department j for department j = the number of times target doctor v is treated in department j / the total number of times target doctor v is treated; Determine the weight of each target doctor in diagnosis group i of department j for diagnosis group i in department j; among them, the weight Wvij of any target doctor v in diagnosis group i of department j for diagnosis group i in department j = the number of times target doctor v's treatment in department j belongs to diagnosis group i / the total number of times target doctor v is treated; Determine the doctor weight WD = [∑v(Wvij) / ∑v(Wvj)]; Determine the title weight WT = the highest professional title among all target doctors in diagnosis group i of department j / the highest professional title among all doctors in department j; Determine the doctor ratio Rij(2) = [NDOij / the total number of doctors in department j] × (WD + WT) / 2; Determine the probability Pij of diagnosis group i for department j = [Rij(1) + Rij(2)] / 2.

4. The system according to claim 2, characterized in that, Said determining the treatment optimization degree of each diagnosis group for each department according to the standard data of each diagnosis group, the medical data belonging to each diagnosis group in each department, the target patients and target doctors for each diagnosis group in each department includes: For any diagnosis group i and any department j, determine the treatment optimization degree of diagnosis group i for department j through the following steps: Determine the matching degree between the medical data of each target patient in diagnosis group i for department j and the standard data of diagnosis group i; Determine the minimum value Simu(min) and standard deviation Simu(sd) of the matching degrees between the medical data of each target patient in diagnosis group i for department j and the standard data of diagnosis group i; Determine the matching degree between the medical data of each target doctor in diagnosis group i for department j and the standard data of diagnosis group i; Determine the minimum value Simv(min) and the standard deviation Simv(sd) of the matching degree between the medical data of each target doctor in diagnosis group i for department j and the standard data of diagnosis group i. Determine the diagnosis and treatment optimization degree ODij of diagnosis group i for department j = min{Simu(min) / patient matching threshold, Simv(min) / doctor matching threshold, 1 - [Simu(sd) / standard deviation threshold], 1 - [Simv(sd) / standard deviation threshold]}; where min{} is the function of taking the minimum value.

5. The system according to claim 1, characterized in that, For the patients in the target diagnosis group, optimizing the medical path of the target diagnosis group includes: Determine the medical data of the target diagnosis group as the target medical data. According to the target medical data, determine the medical paths of each optimizing doctor for each optimizing patient and the standard degree of each medical path; where the optimizing doctor is the doctor involved in the target medical data, and the optimizing patient is the patient involved in the medical data of the target diagnosis group. Obtain the optimized medical path of the target diagnosis group according to the medical paths with a standard degree not less than the standard degree threshold.

6. The system according to claim 5, wherein The medical path is composed of medical nodes arranged in the diagnosis and treatment order. The obtaining of the optimized medical path of the target diagnosis group according to the medical paths with a standard degree not less than the standard degree threshold includes: Construct a path graph of the target disease according to all target medical paths; where the target medical path is the medical path with a standard degree not less than the standard degree threshold, the points in the path graph correspond one by one to the medical nodes in all medical paths, and the edges between the nodes in the path graph correspond to the diagnosis and treatment order between the medical nodes in the medical path; any point in the path graph has a point attribute, and the point attribute value of any node is the number of medical paths containing the any node; any edge in the path graph has an edge attribute, and the edge attribute value of any edge is the number of medical paths with the execution order between the two points connected by the any edge. Among the points in the path graph corresponding to the first medical node of all target medical paths, take the point with the largest point attribute value as the starting node. Among the points in the path graph corresponding to the last medical node of all target medical paths, take the point with the largest point attribute value as the ending node. Determine the target link in the path graph; where the starting point of any target link is the starting node and the ending point is the ending node. Determine the link value of each target link according to the point attribute value and the edge attribute value of the target link. Obtain the optimized medical path of the target diagnosis group according to the link value.

7. The system according to claim 6, wherein The determining of the link value of each target link according to the point attribute value and the edge attribute value of the target link includes: For any target link l, determine its link value through the following steps: For any edge of target link l, determine its edge value as the edge attribute value of any edge × (the average value of the point attribute values of the two points connected by any edge). Determine the average value Eavg(l) and the standard deviation Esd(l) of the edge values of all edges of target link l. Determine that the link value of the target link l is Esd(l) / Eavg(l) × the typicality of the target link l; where the typicality of the target link l = [(the professional title of the doctor corresponding to the target link l / the highest professional title of all doctors corresponding to the target link) + the matching degree between the medical data corresponding to the target link l and the standard data of the diagnosis group i] / 2.

8. The system according to claim 6, characterized in that, The obtaining of the optimized medical path of the target diagnosis group according to the link value includes: Determine the patients corresponding to each target link; Cluster the corresponding patients according to patient attributes; Determine the standard deviation of the link values of all target links in each category; For the category whose standard deviation is not greater than the standard deviation threshold, determine the target link with the largest link value in this category as the optimized medical path of the target diagnosis group; For the category whose standard deviation is greater than the standard deviation threshold, determine the target link with the largest typicality in this category as the optimized medical path of the target diagnosis group.

9. The system according to claim 1, wherein The computer program involved in the patient-centered medical path optimization system is stored in the memory; The memory is located inside the electronic device, and the electronic device further includes: a processor; The processor is configured to execute the computer program involved in the patient-centered medical path optimization system to implement the patient-centered medical path optimization system.

10. The system according to claim 1, wherein The computer program involved in the patient-centered medical path optimization system is stored in a computer-readable storage medium; The computer program involved in the patient-centered medical path optimization system is executed by a processor to implement the patient-centered medical path optimization system.

Images