Method and apparatus for recommending medical plan on basis of feature analysis

Abstract

The present invention belongs to the technical field of medical intelligent data analysis, and relates to a method and apparatus for recommending a medical plan on the basis of feature analysis. The method comprises: S1, establishing a decision tree model, and a medical resource information base used for storing historical medical information; S2, screening target historical medical information falling within the coverage of medical insurance, performing DRG grouping processing, and using the target historical medical information after grouping processing to train the decision tree model; S3, acquiring a user portrait of a target patient, and extracting target features from the user portrait, wherein the target features comprise diagnostic and therapeutic features and economic features; and S4, inputting the target features into an optimal decision tree model obtained by means of training, so as to predict a diagnosis and treatment decision plan, wherein a prediction process specifically comprises: predicting a pre-selected plan set on the basis of the diagnostic and therapeutic features, on the basis of the economic features, screening a decision plan from the pre-selected plan set, and recommending the decision plan. In this way, the accuracy of plan recommendation is high.

Description

A method and device for recommending medical solutions based on feature analysis Technical Field

[0001] The present invention belongs to the technical field of medical intelligent data analysis, and in particular relates to a method and device for recommending medical solutions based on feature analysis. Background Art

[0002] Currently, most existing medical recommendation systems use fixed search methods or simply use historical interaction information between doctors and patients as input to recommend relevant medical information. This recommendation method does not comprehensively consider the patient's personal information, such as the patient's financial information, and therefore may easily lead to inaccurate recommended medical plans and even potential medical risks.

[0003] In addition, some medical systems use rule-tree-based recommendation methods based on clinical guidelines. However, the rule-tree models constructed in existing technologies mainly decompose clinical guidelines and then simulate the doctor's decision-making process through the decomposed information. Therefore, they still cannot accurately cover the actual symptoms of different patients.

[0004] Therefore, how to accurately generate medical recommendation plans in the medical recommendation system has become a technical problem that needs to be solved urgently. From the perspective of clinical medical knowledge, the common methods for grouping and summarizing medical record data include DRG / DIP grouping:

[0005] DRG refers to grouping according to disease diagnosis-related information, dividing hospitalized patients into a certain number of disease groups according to clinical similarities and resource consumption similarities (i.e., according to the severity of the patient's disease, the complexity of the treatment method, and the degree of resource consumption);

[0006] DIP refers to the principle of the work-point system, which converts the relative price relationship between the medical expenses of different diseases and their weights into a score for each disease, and then groups them according to the score;

[0007] In summary, this application proposes a medical plan recommendation method and device based on DRG / DIP grouping technology that can effectively consider patient personal information. Summary of the Invention

[0008] In view of the above deficiencies in the existing technology, the purpose of the present invention is to provide a medical plan recommendation method and device based on feature analysis, which accurately determines the medical plan that meets the user's needs by conducting a comprehensive matching analysis of the user profile of the target patient and historical medical information.

[0009] In the first aspect, to achieve the above-mentioned purpose, the present invention provides the following technical solutions:

[0010] A method for recommending medical solutions based on feature analysis, comprising:

[0011] S1. Establish a decision tree model and a medical resource information database for storing historical medical information;

[0012] S2. Filter the target historical medical information that falls within the scope of medical insurance, perform DRG grouping, and use the target historical medical information after grouping to train the decision tree model;

[0013] S3. Obtain a user profile of the target patient and extract target features from the user profile, wherein the target features include diagnostic and therapeutic features and economic features;

[0014] S4. Input the target features into the trained optimal decision tree model to predict the diagnosis and treatment decision plan, and the prediction process specifically includes:

[0015] predicting a set of preselected options based on the diagnostic and therapeutic characteristics;

[0016] A decision solution is screened from the set of pre-selected solutions according to the economic characteristics, and the decision solution is recommended.

[0017] Furthermore, the historical medical information includes medical department classification information, medical equipment classification information and medical case information.

[0018] Furthermore, the medical department classification information includes at least one or more of department classification information based on treatment methods, department classification information based on disease type, department classification information based on disease location, and department classification information based on the severity of symptoms.

[0019] The department classification information based on treatment methods includes internal medicine classification and surgical classification;

[0020] The department classification information based on disease type includes oncology department classification, infectious disease department classification, and health care department classification;

[0021] The department classification information based on disease location includes ophthalmology classification, stomatology classification, otolaryngology classification, dermatology classification, orthopedics classification, and neurology classification;

[0022] The department classification information based on the severity of the disease includes routine outpatient department classification, emergency outpatient department classification and severe disease department classification.

[0023] Furthermore, the medical device classification information includes at least one or more of the following: device classification information based on structural features and device classification information based on operation methods.

[0024] The device classification information based on structural features includes passive medical device classification and active medical device classification;

[0025] The instrument classification information based on the operation method includes the classification of instruments that contact the human body and the classification of instruments that do not contact the human body.

[0026] Furthermore, the medical case information includes at least identity information, symptom information, pathological diagnosis information, clinical treatment plan information and treatment result information extracted from the patient's medical case text.

[0027] Furthermore, in step S2:

[0028] The target historical medical information after grouping is used to form a training set and a test set respectively;

[0029] Extracting features from the training set based on natural semantic processing, and constructing a historical diagnosis and treatment knowledge graph through the extracted training features;

[0030] Using the historical diagnosis and treatment knowledge graph to train the decision tree model;

[0031] Extracting features from the test set based on natural semantic processing, and inputting the extracted test features into the trained decision tree model to obtain test results;

[0032] Performing a DIP confidence analysis on the test results, wherein the DIP confidence analysis includes a safety analysis and an economic analysis; calculating a safety score through the safety analysis, calculating an economic score through the economic analysis, and weighting the safety score and the economic score to obtain a DIP decision score;

[0033] Determine whether the DIP decision score of the test result exceeds the DIP decision threshold. If so, complete the training and output the current decision tree model as the optimal decision tree model. Otherwise, re-extract the training features and perform optimization and update training.

[0034] Furthermore, the diagnostic and therapeutic characteristics include disease characteristics and physiological induced characteristics. In step S4, the preselected solution set predicted based on the diagnostic and therapeutic characteristics includes:

[0035] Matching the symptom level and symptom cause according to the symptom characteristics;

[0036] Optimizing screening of the causes of the disease through the physiologically induced characteristics;

[0037] A set of preselected solutions is predicted based on the symptom level and the screened symptom causes, and all preselected solutions in the set of preselected solutions are sorted in descending order of DIP decision scores.

[0038] Furthermore, in step S4, screening a decision solution from the set of pre-selected solutions according to the economic characteristics includes:

[0039] Obtain cost control conditions based on the economic characteristics;

[0040] The pre-selected scheme that meets the cost control conditions in the pre-selected scheme set is used as the decision scheme.

[0041] In a second aspect, to achieve the above-mentioned purpose, the present invention also provides the following technical solutions:

[0042] A device for recommending medical solutions based on feature analysis, comprising:

[0043] A construction module for building a decision tree model and a medical resource information base for storing historical medical information;

[0044] The information processing module screens the target historical medical information that falls within the scope of medical insurance and performs DRG grouping processing;

[0045] A training module, configured to train the decision tree model using the grouped target historical medical information;

[0046] Data acquisition module, used to obtain user portraits of target patients;

[0047] A feature extraction module, configured to extract target features from the user portrait;

[0048] The prediction module is used to input the target features into the trained optimal decision tree model to predict the diagnosis and treatment decision plan, and the prediction process specifically includes:

[0049] predicting a set of preselected options based on the diagnostic and therapeutic characteristics;

[0050] A decision solution is screened from the set of pre-selected solutions according to the economic characteristics, and the decision solution is recommended.

[0051] As a general inventive concept, the present invention also provides:

[0052] A computer-readable storage medium stores a computer program, which, when executed by a processor, implements the above-mentioned method for recommending a medical plan based on feature analysis.

[0053] An electronic device comprising a processor, a communication interface, a memory, and a communication bus;

[0054] The processor, communication interface, and memory communicate with each other via a communication bus;

[0055] The memory is used to store computer programs;

[0056] The processor is configured to execute a computer program stored in the memory, and when the computer program is executed, the method for recommending a medical plan based on feature analysis as described above is implemented.

[0057] The beneficial effects of the present invention are as follows:

[0058] In this invention, a decision tree model is obtained by training the target historical medical information that falls within the scope of medical insurance, and the user profile of the target patient is fully matched and analyzed through the decision tree model, so as to quickly obtain a medical plan that can meet the user's needs and effectively improve the accuracy of the plan recommendation;

[0059] Specifically, when extracting features from user portraits, diagnostic and treatment features and economic features were extracted respectively. The decision tree model completes the distribution prediction of medical links such as pathology sorting, cause confirmation, disease refinement and medical process optimization through comprehensive matching analysis of diagnostic and treatment features and historical medical information. The predicted pre-selected options are all based on a comprehensive confidence analysis based on DRG / DIP grouping technology. Finally, taking into account the economic characteristics, a highly reliable recommended decision plan is screened, and the decision plan can effectively fall within the scope of medical insurance, reducing the economic pressure on users. BRIEF DESCRIPTION OF THE DRAWINGS

[0060] The accompanying drawings are only for the purpose of illustrating specific embodiments and are not to be considered as limiting the present invention. Throughout the drawings, the same reference numerals represent the same components. Obviously, the drawings described below are only some of the embodiments of the present invention. Those skilled in the art can also derive other drawings based on these drawings.

[0061] FIG1 is a flow chart of a method for recommending medical treatment plans based on feature analysis provided by the present invention;

[0062] FIG2 is a schematic diagram of a user portrait of a target patient obtained by the present invention;

[0063] FIG3 is a schematic diagram of a historical diagnosis and treatment knowledge graph constructed by the present invention;

[0064] FIG4 is a structural diagram of a device for recommending medical solutions based on feature analysis provided by the present invention. DETAILED DESCRIPTION

[0065] In order to enable those skilled in the art to better understand the technical solutions in the embodiments of the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the embodiments described are part of the embodiments of the present invention, rather than all of the embodiments. It should be understood that these descriptions are merely exemplary and are not intended to limit the scope of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative work should fall within the scope of protection of the present invention.

[0066] Furthermore, in the following description, descriptions of well-known structures and technologies are omitted to avoid unnecessarily obscuring the concepts disclosed in the present invention.

[0067] Exemplary embodiments will be described in detail herein, examples of which are illustrated in the accompanying drawings. In the following description, when referring to the drawings, like numbers in different figures represent the same or similar elements unless otherwise indicated. The embodiments described in the following exemplary embodiments are not intended to represent all possible embodiments consistent with the present invention. Rather, they are merely examples of methods and systems consistent with certain aspects of the present invention, as detailed in the appended claims.

[0068] In order to solve the problem of poor accuracy in existing medical plan recommendations, the present invention specifically provides a medical plan recommendation method and device based on feature analysis.

[0069] Method Example

[0070] As shown in FIG1 , the method for recommending medical solutions based on feature analysis of the present invention mainly includes the following steps:

[0071] S1. Establish a decision tree model and medical resource information database;

[0072] A decision tree model specifically refers to a model that can simulate the human decision-making process based on basic information. The decision tree model performs classification and prediction based on the shape of the tree. Each internal node represents a judgment condition on a feature attribute, each branch represents a possible attribute value, and each leaf node represents a category, thus conveniently and intuitively displaying the entire decision process.

[0073] The medical resource information database is used to store historical medical information, and its historical medical information mainly includes medical department classification information, medical device classification information and medical case information;

[0074] (11) Information on the classification of medical departments: Departments can be classified in different ways. The following are some common classification methods:

[0075] According to the treatment methods: Internal medicine and surgery are the two most common departments. Internal medicine mainly treats diseases through drug therapy, including respiratory medicine, gastroenterology, etc.; while surgery treats diseases through surgical treatment, including general surgery, neurosurgery, etc.

[0076] Classification by disease site: such as ophthalmology, dentistry, otolaryngology, dermatology, orthopedics, neurology, etc.;

[0077] Classification by disease type: such as oncology, infectious diseases, health care, etc.;

[0078] Divided by the severity of the disease: for example, routine outpatient department, emergency outpatient department and intensive care department, etc.

[0079] The above classification method is only one specific example of medical department classification information, and the present invention includes but is not limited to the above classification information. With the continuous development of medical technology, new departments and treatment items are constantly emerging. Therefore, in practical applications, medical department classification information also needs to be continuously updated and adapted.

[0080] (12) Information on medical device classification: Medical devices can be classified according to different characteristics:

[0081] According to structural characteristics: they can be divided into passive medical devices and active medical devices;

[0082] According to the use period (this classification is mainly for passive medical devices): they can be divided into disposable devices, short-term reusable devices, and long-term reusable devices;

[0083] According to the operation method: it can be divided into human contact devices and non-human contact devices;

[0084] According to the operating site (this classification is mainly for devices that come into contact with the human body): they can be divided into skin or cavity (oral) devices, trauma or tissue devices, blood circulation system devices or central nervous system devices;

[0085] According to the structural characteristics and operation methods, they can be divided into passive human-contacting devices, passive human-non-contacting devices, active human-contacting devices, and active human-non-contacting devices.

[0086] The above classification method is only a specific classification example of medical device classification information, and the present invention includes but is not limited to the above classification information. For example, medical devices can also be classified according to their use, degree of impact on the human body, etc.

[0087] (13) About medical case information

[0088] A complete patient medical case text generally includes the patient's identity information, medical history, doctor's orders, nursing documents, examination findings, examination conclusions and other documents. These documents reflect the patient's complete treatment process. With the establishment and improvement of the medical system information, these case texts are usually entered and stored in the form of electronic files. Most of the text contents are composed of natural language, and the sentence structure is relatively complex. Therefore, in an embodiment of the present invention, by dividing the text subject fields, identity information, symptom information, pathological diagnosis information, clinical treatment plan information and treatment result information are extracted from the patient's medical case text, and stored as medical case information corresponding to the patient's medical case text.

[0089] S2. Filter the target historical medical information that falls within the scope of medical insurance, perform DRG grouping, and use the target historical medical information after grouping to train the decision tree model;

[0090] (21) Regarding DRG grouping processing - DRG refers to grouping according to disease diagnosis-related information. Therefore, the above-mentioned target historical medical information that is preferably included in the medical insurance coverage is divided into a certain number of disease groups according to clinical similarity and resource consumption similarity (mainly referring to the severity of the disease, the complexity of the treatment method and the resource consumption of each medical case information).

[0091] (22) About decision tree model training

[0092] The target historical medical information after grouping is used to form a training set and a test set respectively;

[0093] The training set is subjected to feature extraction based on natural semantic processing, and the historical diagnosis and treatment knowledge graph shown in FIG3 is constructed through the extracted training features; specifically, the training feature graph includes aspects such as the medical staff team, outpatient department visit status, department size data, work saturation, inpatient bed and scheduling, operating room utilization status, operating table number statistics and scheduling data;

[0094] Using the historical diagnosis and treatment knowledge graph to train the decision tree model;

[0095] Extracting features from the test set based on natural semantic processing, and inputting the extracted test features into the trained decision tree model to obtain test results;

[0096] Performing a DIP confidence analysis on the test results, wherein the DIP confidence analysis includes a safety analysis and an economic analysis; calculating a safety score through the safety analysis, calculating an economic score through the economic analysis, and weighting the safety score and the economic score to obtain a DIP decision score;

[0097] Determine whether the DIP decision score of the test result exceeds the DIP decision threshold. If so, complete the training and output the current decision tree model as the optimal decision tree model. Otherwise, re-extract the training features and perform optimization and update training.

[0098] S3. Obtain a user portrait of the target patient as shown in FIG2 , and extract target features from the user portrait, wherein the target features include diagnostic and therapeutic features and economic features, and the diagnostic and therapeutic features include symptom features and physiologically induced features.

[0099] S4. Input the target features into the trained optimal decision tree model to predict the diagnosis and treatment decision plan, and the prediction process specifically includes:

[0100] Matching the symptom level and symptom cause according to the symptom characteristics;

[0101] Optimizing screening of the causes of the disease through the physiologically induced characteristics;

[0102] Predicting a set of preselected solutions based on the symptom level and the screened symptom causes, and sorting all preselected solutions in the set of preselected solutions by descending DIP decision scores;

[0103] Obtain cost control conditions based on the economic characteristics;

[0104] The pre-selected scheme that meets the cost control conditions in the pre-selected scheme set is used as the decision scheme, and the decision scheme is recommended.

[0105] In summary, taking a patient who needs to undergo percutaneous coronary balloon angioplasty as an example, the following detailed description of the solution recommendations of the embodiment of the present invention is provided:

[0106] a) Obtain user portraits of target patients through the patient consultation process

[0107] If there is a pre-treatment online consultation, preliminary feature extraction is performed on the consultation information. Based on the extracted features, the target patient's address and regional medical resources can be combined to allocate a hospital for treatment. The severity of the symptoms can also be used to determine whether to register for a regular outpatient or emergency department. If emergency treatment is not required, the specific department to register for is determined based on the location of the disease.

[0108] After registering online or offline, a user profile of the target patient is constructed through the target patient's electronic medical record information and online consultation information.

[0109] b) Extract features and conduct in-depth analysis of user portraits to obtain decision-making solutions

[0110] Predicted pre-selection options:

[0111] b1) Extracting economic characteristics, disease characteristics, and physiological induced characteristics from user portraits;

[0112] b2) Screen and match similar historical cases by symptom characteristics, and deeply analyze common causes, differential causes, symptom levels, precautions, etc.;

[0113] b3) Refined screening of common and differential etiologies through physiological induction characteristics to determine the specific etiology and pathology;

[0114] b4) Perform manual intervention based on differential etiologies and precautions, conduct detailed analysis and obtain clear etiologies and pathologies, and use the decision tree model to output pre-selected plans. Generally speaking, there are multiple matching pre-selected plans, which are combined to form a pre-selected plan set, and each pre-selected plan includes the corresponding surgical plan steps and the full amount of available surgical equipment, medical auxiliary equipment, etc.

[0115] For percutaneous coronary balloon angioplasty, one of the preselected options is as follows:

[0116] (01)Surgical steps:

[0117] The patient lies supine under DSA;

[0118] The Alln test on the right upper limb was negative;

[0119] Perform routine disinfection and lay sterile towels;

[0120] Local anesthesia was performed with 2% lidocaine infiltration, and the right radial artery was punctured, a sheath was placed, and systemic heparinization was performed.

[0121] (02) Coronary angiography showed:

[0122] The left main coronary artery showed no stenosis, and the anterior descending branch had a wide stent width and no stenosis;

[0123] The original stent of the circumflex artery was patent and no stenosis was found;

[0124] There is irregularity in the proximal and mid-segment walls of the coronary artery and 99% stenosis in the distal stent;

[0125] JR3.5 guide catheter to the right coronary artery ostium;

[0126] The BMW guidewire was passed to the distal end of the anterior descending branch and the stenosis was pre-dilated by balloon;

[0127] A cutting balloon was used to remove the thrombus in the original stent, and a rapamycin-eluting stent was placed in the original stent in the distal right coronary artery.

[0128] After dilation, angiography showed that the stenosis at the stent disappeared and the distal T1-MI blood flow was grade 3;

[0129] Remove the tube and sheath, and apply pressure bandage to the puncture site;

[0130] (02) Code search: Angioplasty - Coronary - Percutaneous Transluminal (Balloon) (Single Vessel) 00.66 (Note: Alternative code).

[0131] (03) Operation code: Percutaneous coronary balloon angioplasty 00.66.

[0132] Decision-making pre-selection plan:

[0133] b5) For all the preselected options predicted above, a DIP confidence analysis is performed on each preselected option in the decision tree model, specifically including a safety analysis and a computational analysis of the personnel, equipment, drugs, medical devices, etc. involved in each preselected option. A safety score is calculated through the safety analysis. The safety score is used to reflect diagnostic and treatment factors such as surgical risk, success rate, and postoperative recovery. An economic score is calculated through the economic analysis. The economic score is mainly used to reflect economic level. The safety score and economic score are weighted to obtain a DIP decision score;

[0134] The following table shows the results of the DIP confidence analysis for the preselected options for percutaneous coronary balloon angioplasty:

[0135] b6) Obtaining cost control conditions for target patients based on economic characteristics extracted from user profiles;

[0136] b7) Taking the pre-selected scheme that meets the cost control conditions in the pre-selected scheme set as the decision scheme, and recommending the decision scheme.

[0137] In the actual medical diagnosis and treatment process, facing the recommended decision-making plan, further realistic communication is required between the doctor and the patient. If secondary screening is required in terms of economic level, surgical risk level, etc., a new selection is made through the secondary screening and the calculated DIP decision score to obtain the actual decision-making plan.

[0138] During the overall process, the patient's condition data is accurately obtained through detailed analysis of the user portrait of the target patient, and wrong and erroneous registrations are reduced through rapid matching of departments and pathologies, so that decision-making plans that meet the needs of the target patients can be quickly obtained, the economic situation of the decision-making plans can be clarified, and financial disputes between doctors and patients can be reduced.

[0139] Device embodiment

[0140] As shown in FIG4 , the device for recommending medical solutions based on feature analysis of the present invention mainly includes the following structures:

[0141] A construction module for building a decision tree model and a medical resource information base for storing historical medical information;

[0142] The information processing module screens the target historical medical information that falls within the scope of medical insurance and performs DRG grouping processing;

[0143] A training module, configured to train the decision tree model using the grouped target historical medical information;

[0144] Data acquisition module, used to obtain user portraits of target patients;

[0145] A feature extraction module, configured to extract target features from the user portrait;

[0146] The prediction module is used to input the target features into the trained optimal decision tree model to predict the diagnosis and treatment decision plan, and the prediction process specifically includes:

[0147] predicting a set of preselected options based on the diagnostic and therapeutic characteristics;

[0148] A decision solution is screened from the set of pre-selected solutions according to the economic characteristics, and the decision solution is recommended.

[0149] Specifically, the device provided in this embodiment performs feature analysis and decision-making recommendations for DRG / DIP medical solutions according to the method provided in the first embodiment above during execution.

[0150] In addition, in the present invention, the following embodiments are also provided based on the same inventive concept: an electronic device includes a processor, a communication interface, a memory, and a communication bus;

[0151] The processor, communication interface, and memory communicate with each other via a communication bus;

[0152] The memory is used to store computer programs;

[0153] The processor is used to execute the computer program stored in the memory, and when the computer program is executed, the medical treatment plan recommendation method based on feature analysis of the present invention is implemented.

[0154] The communication bus mentioned in the above terminal can be a Peripheral Component Interconnect (PCI) bus or an Extended Industry Standard Architecture (EISA) bus, etc. The communication bus can be divided into an address bus, a data bus, a control bus, etc. For ease of representation, only one thick line is used in the figure, but it does not mean that there is only one bus or one type of bus. The communication interface is used for communication between the above terminal and other devices. The memory can include a random access memory (RAM) or a non-volatile memory, such as at least one disk storage. Optionally, the memory can also be at least one storage system located away from the aforementioned processor.

[0155] The above-mentioned processor can be a general-purpose processor, including a central processing unit (CPU), a network processor (NP), etc.; it can also be a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, and discrete hardware components.

[0156] In addition, to achieve the above-mentioned purpose, an embodiment of the present invention further proposes a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the feature analysis-based medical plan recommendation method of the present invention.

[0157] Those skilled in the art will appreciate that embodiments of the present invention may be provided as methods, systems, or computer program products. Thus, embodiments of the present invention may take the form of entirely hardware embodiments, entirely software embodiments, or embodiments combining software and hardware aspects. Furthermore, embodiments of the present invention may take the form of a computer program product implemented on one or more computer-usable devices (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0158] The embodiments of the present invention are described with reference to the flowcharts and / or block diagrams of the methods, terminal devices (systems), and computer program products according to the embodiments of the present invention. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing terminal device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing terminal device produce a system for implementing the functions specified in one or more processes in the flowchart and / or one or more boxes in the block diagram.

[0159] These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing terminal device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce a manufactured product including an instruction system that implements the functions specified in one or more processes in the flowchart and / or one or more boxes in the block diagram.

[0160] These computer program instructions can also be loaded onto a computer or other programmable data processing terminal device so that a series of operating steps are executed on the computer or other programmable terminal device to produce computer-implemented processing, so that the instructions executed on the computer or other programmable terminal device provide steps for implementing the functions specified in one or more processes in the flowchart and / or one or more boxes in the block diagram.

[0161] In this document, relational terms such as first and second are merely used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply that there is any such actual relationship or order between these entities or operations. "And / or" means that either one of the two can be selected, or both can be selected. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or terminal device comprising a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or terminal device. In the absence of further restrictions, the elements defined by the sentence "comprising a ..." do not exclude the presence of other identical elements in the process, method, article or terminal device comprising the elements.

[0162] Finally, it should be noted that the above embodiments are merely illustrative of the technical solutions of the embodiments of the present invention, and are not intended to limit them. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they may still modify the technical solutions described in the aforementioned embodiments, or replace some of the technical features therein with equivalents; and such modifications or replacements do not deviate from the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention. Any changes or replacements that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be covered by the scope of protection of the present invention.

Claims

1. A medical solution recommendation method based on feature analysis, characterized in that It includes the following steps: S1. Establish a decision tree model and a medical resource information database for storing historical medical information; S2. Screen the target historical medical information that falls within the scope of medical insurance, perform DRG grouping processing, and use the grouped target historical medical information to train the decision tree model; S3. Obtain the user portrait of the target patient, and extract target features from the user portrait, and the target features include diagnostic features and economic features; S4. Input the target features into the trained optimal decision tree model to predict the diagnosis and treatment decision-making plan, and the prediction process specifically includes: Predict the preselected plan set according to the diagnostic features; Screen the decision-making plan from the preselected plan set according to the economic features, and recommend the decision-making plan.

2. The medical solution recommendation method based on feature analysis according to claim 1, wherein: The historical medical information includes medical department classification information, medical device classification information, and medical case information; and The medical department classification information includes at least one or more of department classification information based on treatment means, department classification information based on disease types, department classification information based on disease locations, and department classification information based on the severity of the disease; The medical device classification information includes at least one or more of device classification information based on structural features and device classification information based on operation methods; The medical case information includes at least identity information, symptom information, pathological diagnosis information, clinical treatment plan information, and treatment result information extracted from the patient's medical case text.

3. The method for recommending a medical plan based on feature analysis according to claim 2, wherein: The department classification information based on treatment means includes internal medicine classification and surgery classification; The department classification information based on disease types includes oncology classification, infectious disease classification, and health care classification; The department classification information based on disease locations includes ophthalmology classification, stomatology classification, otolaryngology classification, dermatology classification, orthopedics classification, and brain department classification; The department classification information based on the severity of the disease includes general outpatient department classification, emergency outpatient department classification, and intensive care unit classification.

4. The method for recommending a medical plan based on feature analysis according to claim 2, wherein: The device classification information based on structural features includes passive medical device classification and active medical device classification; The device classification information based on operation methods includes contact human body device classification and non-contact human body device classification.

5. The method for recommending medical solutions based on feature analysis according to claim 2, wherein In step S2: Use the grouped target historical medical information to form a training set and a test set respectively; Extract features from the training set based on natural semantic processing, and construct a historical diagnosis and treatment knowledge graph through the extracted training features; Use the historical diagnosis and treatment knowledge graph to train the decision tree model; Extract features from the test set based on natural semantic processing, and input the extracted test features into the trained decision tree model to obtain a test result; Performing a DIP confidence analysis on the test results, wherein the DIP confidence analysis includes a safety analysis and an economic analysis; obtaining a safety score through the safety analysis, obtaining an economic score through the economic analysis, and weighting the safety score and the economic score to obtain a DIP decision score; It is determined whether the DIP decision score of the test result exceeds the DIP decision threshold. If yes, the training is completed and the current decision tree model is output as the optimal decision tree model. Otherwise, the training features are re-extracted and the optimization and update training is performed.

6. The method for recommending medical solutions based on feature analysis according to claim 5, characterized in that: The diagnostic and therapeutic characteristics include symptom characteristics and physiological induced characteristics. In step S4, the preselected solution set predicted according to the diagnostic and therapeutic characteristics includes: matching the symptom level and the symptom cause according to the symptom characteristics; Optimizing the screening of the cause of the disease through the physiologically induced characteristics; A set of pre-selected solutions is predicted according to the symptom level and the screened symptom causes, and all pre-selected solutions in the set of pre-selected solutions are sorted in descending order of DIP decision scores.

7. The method for recommending medical solutions based on feature analysis according to claim 6, wherein: In step S4, selecting a decision solution from the set of pre-selected solutions according to the economic characteristics includes: Obtain cost control conditions based on the economic characteristics; The pre-selected scheme in the pre-selected scheme set that meets the cost control condition is taken as the decision scheme.

8. A medical solution recommendation device based on feature analysis, characterized in that, Includes the following structure: A construction module for building a decision tree model and a medical resource information base for storing historical medical information; The information processing module screens the target historical medical information that falls within the scope of medical insurance and performs DRG grouping processing; A training module, used to train the decision tree model using the target historical medical information after grouping processing; A data acquisition module is used to obtain the user portrait of the target patient; A feature extraction module, used to extract target features from the user portrait; The prediction module is used to input the target features into the trained optimal decision tree model to predict the diagnosis and treatment decision plan, and the prediction process specifically includes: Predicting a set of preselected options based on the diagnostic and therapeutic characteristics; A decision plan is screened from the set of pre-selected plans according to the economic characteristics, and the decision plan is recommended.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the method for recommending a medical plan based on feature analysis as described in any one of claims 1 to 7 is implemented.

10. An electronic device, characterized in that: including a processor, a communication interface, a memory and a communication bus; The processor, communication interface, and memory communicate with each other via a communication bus; The memory is used to store computer programs; The processor is used to execute a computer program stored in the memory, and when the computer program is executed, the medical treatment plan recommendation method based on feature analysis as described in any one of claims 1 to 7 is implemented.

Images