Medication evaluation system for specific diseases and computer readable storage medium

Abstract

The invention provides a medication evaluation system for specific diseases and a computer readable storage medium. The medication evaluation system is based on a pre-trained machine learning model or a neural network model, and combines input basic data of a doctor-seeing patient and a planned medication scheme to evaluate a risk condition that transaminase is abnormal after the doctor-seeing patient adopts the planned medication scheme. Particularly, the medication evaluation system can also evaluate the risk probability of transaminase abnormity after a plurality of planned medication schemes of the patient in parallel, and recommend a plurality of medication schemes with the lowest risk. The medication evaluation system provided by the invention is adopted to assist doctors in evaluating and screening medication risks of rheumatoid arthritis patients or systemic lupus erythematosus patients, so that the time of the doctors for adapting medication schemes for the patients can be saved, and the medication safety can be improved.

Description

Technical Field

[0001] The present application relates to the field of computer intelligent assisted decision-making in the medical field, and specifically to an evaluation system that can conveniently and accurately assist medical personnel in conducting medication safety evaluation (based on the risk of abnormal transaminase) for medication regimens prescribed for patients with rheumatoid arthritis and systemic lupus erythematosus. Background Art

[0002] Rheumatoid arthritis (RA) is a chronic, systemic autoimmune disease characterized by synovial inflammation, which often causes joint pain, swelling, stiffness, deformity and dysfunction. The treatment of rheumatoid arthritis is usually based on drugs. Doctors need to prescribe the corresponding medication plan after comprehensive evaluation based on the severity of the patient's disease, physical condition and other factors. There are many types of prescriptions for the treatment of rheumatoid arthritis, including non-steroidal anti-inflammatory drugs, disease-modifying antirheumatic drugs (DMARDs, including traditional synthetic DMARDs and biological DMARDs), glucocorticoids, etc. DAS28 is a scoring standard for the severity of rheumatoid arthritis recognized by the international academic community. Clinicians usually use this score to judge the degree of disease activity in patients.

[0003] Systemic lupus erythematosus (SLE) is an autoimmune disease that forms pathogenic autoantibodies and immune complexes and mediates organ and tissue damage. Clinically, multiple system involvement is often present, and there are multiple autoantibodies represented by antinuclear antibodies in the serum. SLE cannot be cured at present, and treatment should be individualized, but long-term remission can be achieved after reasonable treatment. The main treatment options are glucocorticoids combined with immunosuppressants. The principle of treatment is to actively use drugs to induce remission in the acute phase and control the activity of the disease as soon as possible; after the disease is relieved, adjust the medication, and maintain the remission treatment to keep it in remission, protect the function of important organs and reduce drug side effects. There are many standards that can be used to evaluate disease activity. The current standards include SLEDAI.

[0004] After diagnosing the severity of a patient's rheumatoid arthritis or systemic lupus erythematosus, the doctor needs to select an appropriate medication regimen and evaluate the pros and cons of each medication regimen and the patient's acceptance. This requires not only considering factors such as the therapeutic effect of the medication regimen on the specific patient and the cost of medication, but the most important consideration is the safety of the medication. Although some medication regimens have good therapeutic effects in theory, the corresponding side effects vary greatly for individual patients in different situations; severe cases may cause patients to experience abnormal transaminase or even liver failure. Therefore, the selection of medication regimens for patients with rheumatoid arthritis and systemic lupus erythematosus often requires doctors to repeatedly consider based on their own experience.

[0005] Currently, one of the reliable indicators of early liver side effects after medication for rheumatoid arthritis patients is abnormal transaminase. Abnormal transaminase is also a reliable indicator of early liver side effects of medication for systemic lupus erythematosus. For this reason, it is urgent to provide a computer system that can accurately and scientifically evaluate the safety risks of various rheumatoid arthritis or systemic lupus erythematosus medication regimens, so as to help medical personnel accurately and efficiently prescribe safe and effective medication regimens for different individual patients. Summary of the invention

[0006] In view of this, the present application provides a medication evaluation system for a specific disease. The medication evaluation system assists medical personnel in efficiently and accurately evaluating the liver safety of various medication regimens for different individual patients based on their specific conditions based on a machine learning model or a neural network model and existing patient data.

[0007] The technical solution provided by the present invention is implemented as: a medication evaluation system for a specific disease, characterized in that the medication evaluation system includes: a patient basic data input module, a medication regimen input module, an evaluation feature extraction module, a transaminase abnormality prediction model and a medication evaluation result output module.

[0008] Among them, the evaluation feature module is used to associate, clean and encode the basic data of the specific patient input through the patient basic data input module and the medication regimen of the specific patient input through the patient medication regimen input module to obtain the transaminase abnormality risk prediction characteristics of the specific patient.

[0009] The transaminase abnormality prediction model is trained by a machine learning model or a neural network, and is used to process the transaminase abnormality risk prediction feature of the specific patient to obtain the transaminase abnormality prediction result of the specific patient using the medication regimen, and output it through the medication evaluation result output module. The output of the transaminase abnormality prediction model is the prediction result of whether the patient's transaminase is abnormal (determined according to a preset probability threshold), or it can be the probability of transaminase abnormality.

[0010] Furthermore, the abnormal transaminase prediction model is trained by using a training sample consisting of historical medication data of patients with the specific disease in the past, basic data of patients corresponding to the historical medication data, and identification data of whether the transaminase corresponding to the historical medication data is abnormal. Obviously, for training models for prediction, neural network models and machine learning models are both commonly used choices. Among them, the specific type of machine learning model used in this application is not limited here, and a random forest model is preferably used.

[0011] Furthermore, the medication regimen and the medication data include: medication type, medication dosage, and medication duration. The specific diseases include: systemic lupus erythematosus and rheumatoid arthritis. For rheumatoid arthritis, the basic data include: age, disease course, gender, and DAS28 score. For systemic lupus erythematosus, the patient basic data include: age, disease course, gender, and SLEDAI score.

[0012] Preferably, the patient basic data input module provides a user input interface, in which the user selects the disease type, rheumatoid arthritis patients input the DAS28 score, and systemic lupus erythematosus patients input the SLEDAI score.

[0013] Furthermore, the medication regimen input module provides a user with an input interface for inputting multiple parallel medication regimens. The evaluation feature extraction module associates, cleans and encodes the multiple parallel medication regimens with the basic data of the corresponding patient input via the patient basic data input module to obtain multiple groups of abnormal transaminase risk prediction features for the corresponding patient. Then, each group of abnormal transaminase risk prediction features is input into the abnormal transaminase prediction model for processing (abnormal transaminase risk prediction) to obtain multiple risk probabilities of abnormal transaminase when the corresponding patient adopts the multiple parallel medication regimens, and outputs them simultaneously via the medication evaluation result output module.

[0014] Preferably, the medication evaluation result output module also recommends to the user several medication regimens with the lowest risk probability of abnormal transaminase.

[0015] Corresponding to the above-mentioned medication evaluation system for specific diseases, another aspect of the present invention further provides a computer-readable storage medium. The computer-readable storage medium stores program codes, and when the program codes are executed by a computer or a processor, the functions of the above-mentioned medication evaluation system for specific diseases are realized.

[0016] The medication evaluation system provided by the present invention makes full use of machine learning models to mine the correlation between historical patient medical data. The use of this medication safety evaluation system can assist medical personnel in scientifically and accurately evaluating the safety of medication regimens for specific rheumatoid arthritis or systemic lupus erythematosus patients, helping medical personnel to efficiently select low-risk medication regimens, improving the scientificity and safety of medication for patients with rheumatoid arthritis or systemic lupus erythematosus, and saving medical personnel time in selecting medication regimens. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings required for use in the embodiments will be briefly introduced below. It should be understood that the following drawings only show certain embodiments of the present application and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without paying creative work.

[0018] Figure 1 A schematic diagram of an embodiment of a medication evaluation system for a specific disease provided by the present invention. DETAILED DESCRIPTION

[0019] The technical solutions in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all of the embodiments. The components of the embodiments of the present application described and shown in the drawings here can be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of the present application provided in the drawings is not intended to limit the scope of the application claimed for protection, but merely represents the selected embodiments of the present application. Based on the embodiments of the present application, all other embodiments obtained by those skilled in the art without making creative work belong to the scope of protection of the present application.

[0020] It should be noted that similar reference numerals and letters represent similar items in the following drawings, so once an item is defined in one drawing, it does not need to be further defined and explained in subsequent drawings. At the same time, in the description of this application, the terms "first", "second", etc. are only used to distinguish the description and cannot be understood as indicating or implying relative importance.

[0021] like Figure 1 As shown, in one embodiment, the medication evaluation system 100 for a specific disease provided by the present invention comprises: a patient basic data input module 101, a medication regimen input module 102, an evaluation feature extraction module 103, a transaminase abnormality prediction model 104 and a medication evaluation result output module 105. The specific diseases include: systemic lupus erythematosus and rheumatoid arthritis.

[0022] In this embodiment, the patient basic data input module 101 provides an input interface for inputting the basic data of a specific patient. For patients with rheumatoid arthritis, the basic data include: age, course of disease, gender and DAS28 score. For patients with systemic lupus erythematosus, the patient basic data include: age, course of disease, gender and SLEDAI score. Among them, the DAS28 evaluation standard is the core standard for evaluating the disease activity of rheumatoid arthritis at present, and its evaluation of the disease activity of rheumatoid arthritis is quite scientific, comprehensive and authoritative. The SLEDAI evaluation standard is the current mainstream disease activity evaluation standard for systemic lupus erythematosus. The use of DAS28 scores or SLEDAI scores can comprehensively and scientifically measure the condition of patients with rheumatoid arthritis or systemic lupus erythematosus.

[0023] Obviously, the basic data of the patient may not be limited to the above-described parameters, and some additional parameters may be added as needed; the basic data of the patient will not be listed in detail here.

[0024] Preferably, the user input interface provided by the patient basic data input module 101 allows, after the user selects the disease type, rheumatoid arthritis patients to input the DAS28 score, and systemic lupus erythematosus patients to input the SLEDAI score.

[0025] The medication regimen input module 102 provides an input interface for inputting the medication regimen prepared by the doctor for a specific patient. The description parameters of the medication regimen at least include the type of medication, the dosage of medication and the duration of medication.

[0026] The evaluation feature module 103 is used to associate, clean and encode the basic data of a specific patient input via the patient basic data input module 101 and the medication regimen of the specific patient input via the patient medication regimen input module 102 to obtain the risk prediction characteristics of abnormal transaminase of the specific patient. Among them, association mainly refers to associating and corresponding the basic data of the same patient with the medication regimen of the patient to form predictive data for predicting the risk of abnormal transaminase of the patient after adopting the corresponding medication regimen. Cleaning mainly involves eliminating missing data or data with obviously abnormal values ​​according to set conditions. Encoding, this step mainly encodes various text information into a format suitable for processing by the abnormal transaminase prediction model 104 according to certain rules. There are many types of corresponding encodings, which will not be listed here.

[0027] In this embodiment, the transaminase abnormality prediction model 104 is trained using a random forest model, and is used to process the transaminase abnormality risk prediction feature of the specific patient to obtain a transaminase abnormality prediction result of the specific patient using the medication regimen, and output it via the medication evaluation result output module 105. The output of the transaminase abnormality prediction model 104 is whether the transaminase is abnormal (determined according to a preset probability threshold), or it can be the probability of the transaminase being abnormal.

[0028] Obviously, the transaminase abnormality prediction model in this case can not only be trained using the random forest model, but also other machine learning models or neural network models.

[0029] Specifically, the transaminase abnormality prediction model 104 is trained by using training samples consisting of historical medication data of patients who have previously suffered from the specific disease (rheumatoid arthritis and systemic lupus erythematosus), basic data of patients corresponding to the historical medication data, and identification data of whether the transaminase corresponding to the historical medication data is abnormal. The medication data includes the type of medication, the dosage of medication, and the duration of medication, which are the descriptive parameters of the aforementioned medication regimen.

[0030] exist Figure 1 In the illustrated embodiment, the medication regimen input module 102 provides a user with an input interface for inputting multiple parallel medication regimens. The evaluation feature extraction module 103 is implemented by associating, cleaning and encoding the multiple parallel medication regimens with the basic data of the corresponding patient input via the patient basic data input module to obtain multiple groups of abnormal transaminase risk prediction features for the corresponding patient. The evaluation feature extraction module 103 inputs each group of abnormal transaminase risk prediction features into the abnormal transaminase prediction model for processing (abnormal transaminase risk prediction) to obtain multiple risk probabilities of abnormal transaminase when the corresponding patient adopts the multiple parallel medication regimens, and outputs them simultaneously via the medication evaluation result output module.

[0031] Preferably, the confidence level of the occurrence of transaminase abnormality in the prediction result of the transaminase abnormality prediction model 104 is used as the risk probability of transaminase abnormality.

[0032] Preferably, the medication evaluation result output module also recommends to the user several medication regimens with the lowest risk probability of abnormal transaminase.

[0033] Through the above settings, the medication evaluation system in the above embodiment can quickly inform medical staff of the risk of abnormal transaminase of each candidate medication regimen relative to the patient when medical staff are considering multiple candidate medication regimens, thereby assisting medical staff to quickly select the optimal medication regimen.

[0034] Corresponding to the above-mentioned medication evaluation system for specific diseases, another aspect of the present invention further provides a computer-readable storage medium. The computer-readable storage medium stores program codes, and when the program codes are executed by a computer or a processor, the functions of the medication evaluation system for specific diseases are realized. The computer-readable storage medium includes but is not limited to network storage such as RAM, ROM, sdram, flash, norflash, optical disks, and network disks.

[0035] The above description is only the preferred embodiment of the present application and is not intended to limit the present application. For those skilled in the art, the present application may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application shall be included in the protection scope of the present application. Therefore, the protection scope of the present application shall be subject to the protection scope of the claims.

Claims

1. A medication evaluation system for a specific disease, characterized in that: The medication evaluation system includes: a patient basic data input module, a medication regimen input module, an evaluation feature extraction module, a transaminase abnormality prediction model and a medication evaluation result output module; The assessment feature module is used to associate, clean and encode the basic data of the specific patient input through the patient basic data input module and the medication regimen of the specific patient input through the patient medication regimen input module to obtain the risk prediction feature of abnormal transaminase of the specific patient; The transaminase abnormality prediction model is trained using a machine learning model or a neural network model, and is used to process the transaminase abnormality risk prediction characteristics of the specific patient to obtain the transaminase abnormality prediction result of the specific patient using the medication regimen, and output it through the medication evaluation result output module.

2. The medication evaluation system according to claim 1, characterized in that: The transaminase abnormality prediction model is trained by using a training sample consisting of historical medication data of patients who have previously suffered from the specific disease, basic data of patients corresponding to the historical medication data, and identification data of whether the transaminase corresponding to the historical medication data is abnormal to train a machine learning model or a neural network model.

3. The medication evaluation system according to claim 2, characterized in that: The specific diseases include: systemic lupus erythematosus and rheumatoid arthritis.

4. The medication evaluation system according to claim 3, characterized in that: The patient basic data include: age, disease course, gender, historical medication records, and DAS28 scores and SLEDAI scores.

5. The medication evaluation system according to claim 4, characterized in that: The medication data and the medication regimen include: medication type, medication dosage, and medication duration.

6. The medication evaluation system according to any one of claims 1 to 5, characterized in that: The prediction model is a random forest model.

7. The medication evaluation system according to claim 4 or 5, characterized in that: The patient basic data input module provides a user input interface, where the user selects the disease type, rheumatoid arthritis patients input the DAS28 score, and systemic lupus erythematosus patients input the SLEDAI score.

8. The medication evaluation system according to claim 7, characterized in that: The medication scheme input module provides a user with an input interface for inputting multiple concurrent medication schemes; the evaluation feature extraction module associates, cleans and encodes the multiple concurrent medication schemes with the basic data of the corresponding patients input via the patient basic data input module to obtain multiple groups of transaminase abnormality risk prediction features of the corresponding patients; Each set of transaminase abnormality risk prediction features is input into the transaminase abnormality prediction model for processing to obtain multiple risk probabilities of transaminase abnormality when the corresponding patients adopt the multiple parallel medication regimens, and are output simultaneously through the medication evaluation result output module.

9. The medication evaluation system according to claim 8, characterized in that: The medication evaluation result output module also recommends to the user several medication regimens with the lowest risk probability of abnormal transaminase.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores program codes, and when the program codes are executed by a computer or a processor, the functions of the medication evaluation system for a specific disease according to any one of claims 1 to 9 are implemented.

Images