Clinical treatment scheme recommendation method based on brain diseases

Abstract

The invention relates to the technical field of clinical treatment scheme recommendation. The invention relates to a clinical treatment scheme recommendation method based on brain diseases. The method comprises the following steps: S1, collecting brain examination data of a patient and other collectable body data, and then extracting disease features and other features of the brain; s2, collecting a historical clinical treatment scheme set and corresponding historical patient data, and combining the historical clinical treatment scheme set with the historical patient data to score the effect of the treatment scheme; according to the scheme adjusting model, the system can carry out personalized adjustment on the treatment scheme according to individual differences of patients, the unique characteristics of each patient are considered, the most suitable treatment scheme is customized for the patient, and meanwhile, the treatment effect of the treatment scheme is predicted by applying a simulation analysis technology; a doctor and a patient can roughly know effects possibly brought by different schemes before treatment, so that the doctor can select a scheme with a better curative effect expectation, and blind attempts are avoided.

Description

Technical Field

[0001] The present invention relates to the technical field of clinical treatment plan recommendation, and in particular to a method for recommending clinical treatment plans based on brain diseases. Background Art

[0002] In the clinical treatment of brain diseases, doctors can currently make a preliminary judgment on the type and severity of a patient's brain disease through conventional diagnostic procedures, such as brain imaging (MRI, CT, etc.), clinical symptom assessment, and medical history inquiries. Subsequently, doctors formulate corresponding treatment plans based on their own professional knowledge and clinical experience, combined with reference to existing medical literature and treatment guidelines;

[0003] However, doctors' decisions are highly dependent on personal experience and knowledge reserves. There are large differences in the levels of different doctors, which makes it difficult to ensure the accuracy and consistency of treatment plans. This prevents some patients from receiving the most appropriate treatment, affecting the treatment effect and recovery process. Although the existing medical literature and treatment guidelines have certain guiding significance, it is difficult to fully cover the individual differences of all patients. Factors such as the patient's genetic characteristics, lifestyle habits, and other comorbidities are difficult to be fully considered in the traditional treatment decision-making process. Therefore, a method for recommending clinical treatment plans based on brain diseases is proposed. Summary of the invention

[0004] The purpose of the present invention is to provide a method for recommending clinical treatment plans based on brain diseases to solve the problems raised in the above background technology.

[0005] To achieve the above purpose, a method for recommending a clinical treatment plan based on brain diseases is provided, comprising the following steps:

[0006] S1. Collect the patient's brain examination data and other body data that can be collected, and then extract disease characteristics and other characteristics from the brain;

[0007] S2. Collect historical clinical treatment plan sets and corresponding historical patient data, and combine the historical clinical treatment plan sets with historical patient data to score the effectiveness of the treatment plans;

[0008] S3. Perform treatment adaptability analysis based on other characteristics of the patient, then extract treatment plans based on the treatment adaptability analysis combined with disease characteristics and historical clinical treatment plan sets, and screen the extracted treatment plans by effect scoring;

[0009] S4. Establish a regimen adjustment model, through which the treatment regimen after screening is adjusted individually according to the patient's disease characteristics and other characteristics, and then predict the efficacy of the adjusted treatment regimen;

[0010] S5. Recommend and compare treatment plans based on the efficacy prediction results, recommend treatment plans to doctors for review based on the comparison results, and make overall adaptive modifications to the treatment plans based on the doctors' review opinions.

[0011] As a further improvement of the technical solution, the steps of S1 are as follows:

[0012] S1.1. By establishing an information collection module in the hospital system, providing an information input channel to doctors through the information collection module, and using the data received through the information input channel as brain examination data and other body data;

[0013] S1.2. Disease characteristics and other characteristics are extracted based on brain examination data and other physical data combined with the correlation with the incidence of brain diseases.

[0014] As a further improvement of the technical solution, the brain examination data in S1 includes brain imaging data and electrophysiological data, and other body data includes basic information of the patient and description of clinical symptoms;

[0015] Disease signatures represent key features extracted from imaging data;

[0016] Other relevant features represent physical characteristics extracted from the patient's basic information.

[0017] As a further improvement of the technical solution, the steps of S2 are as follows:

[0018] S2.1. Collect historical clinical treatment plan sets and extract historical patient data corresponding to each clinical treatment plan, so as to build a knowledge base containing various brain disease treatment plans in the information collection module;

[0019] S2.2. Extract the brain characteristic data before treatment and the brain characteristic data after treatment from the historical patient data, and combine the historical clinical treatment plan set with the brain characteristic data before treatment and the brain characteristic data after treatment to score the effect of the treatment plan, so as to obtain the effect value of each clinical treatment plan.

[0020] As a further improvement of the present technical solution, the effect score in S2.2 includes the improvement of disease-related indicators, treatment safety and side effects, and the risk of disease recurrence and progression.

[0021] As a further improvement of this technical solution, the steps of S3 are as follows:

[0022] S3.1. Conduct treatment adaptability analysis based on other characteristics of the patient to obtain treatment methods that the patient's physical condition can adapt to;

[0023] S3.2. Combine the analyzed treatment methods with the disease characteristics and the historical clinical treatment plan set to extract the treatment plan, and obtain the treatment plan set that is suitable for the patient's disease characteristics;

[0024] S3.3. The treatment plans obtained in S3.2 are concentrated and each treatment plan is screened by effect score, so that only the treatment plans with higher effect scores are retained among similar treatment plans.

[0025] As a further improvement of the technical solution, the step of S4 is as follows:

[0026] S4.1. Establish a regimen adjustment model, and then input the selected treatment regimen into the regimen adjustment model in combination with the patient's disease characteristics and other characteristics. The regimen adjustment model will make personalized adjustments to the input treatment regimen based on the individual differences of the patient;

[0027] S4.2. Use simulation analysis technology to predict the efficacy of the treatment plan after personalized adjustment in S4.1 in combination with the patient, and obtain the efficacy prediction results of each treatment plan.

[0028] As a further improvement of this technical solution, the formula of S4.1 is as follows:

[0029] T=kT 0

[0030] Among them, T is the adjusted operation time, k is the adjustment coefficient calculated by the model according to the patient characteristics, and T 0 is the original operation time;

[0031] D=fD 0

[0032] Where D is the adjusted drug dose, f is the dose adjustment factor calculated by the model, and D 0 The original drug dose.

[0033] As a further improvement of the technical solution, the step of S5 is as follows:

[0034] S5.1. Integrate the efficacy prediction results of each treatment plan for recommendation comparison, select similar treatment plans for better preservation of efficacy prediction results, and then send the compared and saved treatment plans to doctors for review;

[0035] S5.2. Make overall adaptation modifications to the saved treatment plan based on the doctor’s review opinion.

[0036] Compared with the prior art, the present invention has the following beneficial effects:

[0037] In this method of recommending clinical treatment plans for brain diseases, the system can make personalized adjustments to the treatment plan based on the individual differences of the patient through the plan adjustment model. Taking into account the unique characteristics of each patient, the most suitable treatment plan can be tailored for the patient. At the same time, simulation analysis technology is used to predict the efficacy of the treatment plan, so that doctors and patients can have a general understanding of the possible effects of different plans before treatment. This helps doctors choose plans with better expected efficacy and avoid blind attempts. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] Figure 1 It is the overall flow chart of the present invention;

[0039] Figure 2 A flowchart of the present invention for establishing an information collection module in a hospital system;

[0040] Figure 3 A flowchart of collecting a historical clinical treatment plan set for the present invention;

[0041] Figure 4 A flowchart of the present invention for analyzing treatment adaptation according to other characteristics of the patient;

[0042] Figure 5 A flowchart of a scheme adjustment model for the present invention;

[0043] Figure 6 This is a flowchart of the present invention for overall adaptive modification of a stored treatment plan based on a doctor's review opinion. DETAILED DESCRIPTION

[0044] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0045] See also Figure 1 - Figure 6 As shown, the purpose of this embodiment is to provide a method for recommending a clinical treatment plan based on brain diseases, comprising the following steps:

[0046] S1. Collect the patient's brain examination data and other body data that can be collected, and then extract disease characteristics and other characteristics from the brain;

[0047] The steps of S1 are as follows:

[0048] S1.1. By establishing an information collection module in the hospital system, providing an information input channel to doctors through the information collection module, and using the data received through the information input channel as brain examination data and other body data;

[0049] A specially designed information collection module is embedded in the hospital's existing information management system. This module has good compatibility and can be seamlessly connected with the hospital's imaging system, testing system, and electronic medical record system.

[0050] In the electronic medical record system operation interface of the doctor's workstation, a special data collection portal is set up and presented in the form of a pop-up window or sidebar. For brain examination data, doctors can directly select the corresponding brain MRI, CT, PET and other imaging examination results from the PACS system and associate them with the current patient's medical record;

[0051] S1.2. Disease characteristics and other characteristics are extracted based on brain examination data and other physical data combined with the correlation with the incidence of brain diseases.

[0052] Brain examination data in S1 include brain imaging data and electrophysiological data, and other physical data include basic patient information and clinical symptom description;

[0053] Disease signatures represent key features extracted from imaging data;

[0054] Other relevant features represent physical characteristics extracted from the patient's basic information.

[0055] Brain disease feature extraction: For MRI and CT images, the U-Net network structure is used. Its input is the brain MRI image. It extracts features through a series of convolutional layers and pooling layers, and then upsamples through a deconvolution layer. Finally, the segmented brain lesion area and the extracted lesion features are output. At the same time, features such as disease attack frequency and duration are extracted from the patient's medical history.

[0056] Extraction of other physical features: Extract age, gender and other features from the patient's basic information as categorical variables, and extract key indicators from laboratory test data, such as blood sugar, blood lipids, liver and kidney function indicators, etc. These indicators are directly used as features for subsequent analysis.

[0057] S2. Collect historical clinical treatment plan sets and corresponding historical patient data, and combine the historical clinical treatment plan sets with historical patient data to score the effectiveness of the treatment plans;

[0058] The steps of S2 are as follows:

[0059] S2.1. Collect historical clinical treatment plans and extract historical patient data corresponding to each clinical treatment plan, so as to build a knowledge base containing various brain disease treatment plans in the information collection module. The specific steps are as follows:

[0060] Collect historical clinical treatment plan sets: Screen all records involving brain disease treatment from the hospital's electronic medical record system. In addition to the electronic medical record system, integrate brain imaging examination reports in the hospital's picture archiving and communication system (PACS) and related test reports in the laboratory information management system (LIS);

[0061] Extract historical patient data corresponding to each clinical treatment plan: Extract the patient's basic demographic information from the electronic medical record, including age, gender, height, weight, etc., and extract and collect disease-related data;

[0062] A knowledge base is built into the information collection module: a relational database structure is used to construct the knowledge base, and multiple related tables are designed. By establishing associations between tables (such as foreign key constraints), effective data integration and query are achieved, and then the collected and extracted historical clinical treatment plan sets and corresponding historical patient data are entered according to the designed knowledge base structure.

[0063] S2.2. Extract the brain characteristic data before treatment and the brain characteristic data after treatment from the historical patient data, and combine the historical clinical treatment plan set with the brain characteristic data before treatment and the brain characteristic data after treatment to score the effect of the treatment plan, so as to obtain the effect value of each clinical treatment plan.

[0064] The effect score in S2.2 includes the improvement of disease-related indicators, treatment safety and side effects, and the risk of disease recurrence and progression. The specific steps are as follows:

[0065] Extraction of brain feature data before treatment: For MRI images, image segmentation algorithms, such as the U-Net network based on deep learning, are used to segment brain tissue and lesion areas. For CT images, tissue characteristics are analyzed based on CT values. For electroencephalogram (EEG) data, Fourier transform is used to convert time domain signals into frequency domain signals.

[0066] Extraction of brain feature data after treatment: Similar to the data extraction method before treatment, but focusing on the changes in imaging and electrophysiological data after treatment, observing the reduction or disappearance of lesions on MRI images, recalculating the size, shape and other characteristics of lesions, and comparing the numerical changes before and after treatment. For EEG data, compare the changes in brain wave power spectrum density in different frequency bands before and after treatment to evaluate the recovery of brain function;

[0067] Improvement score of disease-related indicators: For brain lesions, such as tumors, the change rate of lesion volume before and after treatment is calculated, and the scoring standard is set according to the volume change rate. The greater the volume change rate, the higher the score. At the same time, taking epilepsy patients as an example, the frequency change of epileptiform discharges in EEG before and after treatment is compared. The scoring standard is also set according to the frequency change rate. The higher the frequency change rate, the higher the score. The comprehensive lesion improvement score and electrophysiological function improvement score are used as the improvement score of disease-related indicators;

[0068] Treatment safety and side effect score: The types and severity of drug side effects experienced by patients during treatment are counted, and the side effects are divided into three levels: mild, moderate, and severe. Scores are assigned to each level, and then the total score is calculated based on the level and number of drug side effects. For surgical treatment plans, the occurrence of surgery-related complications, such as infection, bleeding, and other complications, is counted. Each complication is assigned a corresponding score based on its severity. The drug side effect score and surgical complication score are combined to obtain the treatment safety and side effect score;

[0069] Disease recurrence and progression risk score: Count the number of patients who have a disease relapse among those who receive a certain treatment regimen within a certain period of time, and set a score based on the recurrence rate. The lower the recurrence rate, the higher the score. For brain diseases, such as neurodegenerative diseases, the progression risk is measured by evaluating the rate of deterioration of specific symptoms after treatment. It is assumed that the smaller the decrease in MMSE score each year after treatment, the higher the score. Then the recurrence risk score and progression risk score are combined as the disease recurrence and progression risk score;

[0070] Obtain the effect value of each clinical treatment plan: Based on the scores of the above three aspects, the weighted average method is used to calculate the final effect value of each clinical treatment plan. The formula is as follows:

[0071] G=J×W 1 +Z×W 2 +B×W 3

[0072] Among them, G is the effect value of the clinical treatment plan, J is the improvement score of disease-related indicators, Z is the treatment safety and side effect score, B is the risk score of disease recurrence and progression, and W is the risk score of disease recurrence and progression. 1 The weight for scoring the improvement of disease-related indicators, W 2 is the weight of the treatment safety and side effect score, W 3 The weights of the disease recurrence and progression risk scores are used to derive a quantitative effect value for each clinical treatment option, which facilitates comparison and ranking of the effects of different treatment options.

[0073] S3. Perform treatment adaptability analysis based on other characteristics of the patient, then extract treatment plans based on the treatment adaptability analysis combined with disease characteristics and historical clinical treatment plan sets, and screen the extracted treatment plans by effect scoring;

[0074] The steps for S3 are as follows:

[0075] S3.1. Conduct treatment adaptability analysis based on other characteristics of the patient to obtain treatment methods that the patient's physical condition can adapt to;

[0076] S3.2. Combine the analyzed treatment methods with the disease characteristics and the historical clinical treatment plan set to extract the treatment plan, and obtain the treatment plan set that is suitable for the patient's disease characteristics. The specific steps are as follows:

[0077] Establish a fitness evaluation model: A decision tree model can be used to analyze the adaptability of patient characteristics and treatment methods. Taking age as an example, if the patient is older than 65 years old, the decision tree can judge that the patient has low fitness for some more invasive surgical treatments. If the patient also suffers from severe cardiopulmonary diseases, the fitness score for surgical treatment will be further reduced. For drug treatment, the adaptability to the metabolism of different drugs can be judged based on the patient's liver and kidney function indicators. If the patient's liver function indicators (such as alanine aminotransferase and aspartate aminotransferase) are beyond the normal range, the fitness of some drugs metabolized by the liver will be reduced;

[0078] Quantitative calculation of fitness: For each treatment method, assign weights corresponding to different features, and then calculate the comprehensive fitness score. Assuming that the treatment method is surgical treatment, the formula is as follows:

[0079] S=W A ×f(A)+W H ×H+W L ×g(L)

[0080] Where S is the surgical treatment fitness score, A is age, H is the history of heart disease, L is the liver function index, f(A) and g(L) are score functions pre-set according to the age and liver function index range. For example, the older the age, the lower the score, and the worse the liver function index, the lower the score. W A , W H , W L The weights of age, heart disease history, and liver function index are respectively used. In a similar way, the adaptability score of each treatment method (such as drug therapy, physical therapy, rehabilitation therapy, etc.) for the patient is calculated;

[0081] Then a fitness threshold is set, and treatments with scores greater than the fitness threshold are considered to be adaptable to the patient's physical condition. Treatments with fitness scores exceeding the threshold are screened out to form a preliminary set of adaptive treatments.

[0082] Extract treatment plans based on disease characteristics: Analyze the patient's brain disease characteristics, such as disease type (tumor, cerebrovascular disease, neurodegenerative disease, etc.), lesion site, severity, etc., and then screen out treatment plans that match the patient's disease characteristics and initial adapted treatment methods from the historical clinical treatment plan collection. Summarize all eligible treatment plans that have been screened out to form a treatment plan set that can be applied to the patient's disease characteristics.

[0083] S3.3, screening the effect scores of each treatment plan in the treatment plans obtained in S3.2, so that only the treatment plans with higher effect scores are retained among similar treatment plans;

[0084] Similarity is defined based on key factors such as the main treatment method, type of drug, surgical method, etc. of the treatment plan. For example, if two treatment plans both use a specific drug as the main treatment, and the drug dosage and treatment course are similar, they can be considered similar treatment plans; or if two surgical treatment plans use similar key operations such as surgical approach and resection range, they are also classified as similar treatment plans. All plans in the treatment plan set are traversed, and for each plan, they are compared with other plans one by one to see if they are similar. If similar plans are found, their comprehensive effect scores are compared, and the plan with the higher score is retained, and the plan with the lower score is deleted.

[0085] S4. Establish a regimen adjustment model, through which the treatment regimen after screening is adjusted individually according to the patient's disease characteristics and other characteristics, and then predict the efficacy of the adjusted treatment regimen;

[0086] The steps of S4 are as follows:

[0087] S4.1. Establish a regimen adjustment model, and then input the selected treatment regimen into the regimen adjustment model in combination with the patient's disease characteristics and other characteristics. The regimen adjustment model makes personalized adjustments to the input treatment regimen based on the individual differences of the patient. The specific steps are as follows:

[0088] Model selection: You can choose from a variety of machine learning models, such as neural networks (such as multi-layer perceptron MLP, recurrent neural network RNN ​​and its variant LSTM, etc.), decision trees, random forests, etc. Taking the multi-layer perceptron as an example, it consists of an input layer, a hidden layer, and an output layer, and each layer is connected by weights. The input layer receives the patient's disease characteristics and other feature data, the hidden layer extracts and transforms the data through nonlinear activation functions (such as ReLU functions), and the output layer outputs the adjusted treatment plan parameters;

[0089] Input data: After preprocessing, the selected treatment plans and the patient's disease characteristics and other characteristic data are input into the trained plan adjustment model. For example, for a patient with a brain tumor, the disease characteristics include the size, location, pathological type, etc. of the tumor, and other characteristics include age, physical condition, underlying diseases, etc. At the same time, the initially screened surgical treatment plans (such as surgical method, resection range, etc.) are used as input;

[0090] Model prediction and adjustment: The model outputs personalized treatment plan adjustment suggestions for the patient based on the relationship between the learned patient characteristics and treatment plan adjustments. If the model predicts that the patient needs to shorten the operation time and reduce the dose of anesthetic drugs during brain surgery due to his / her old age and weak body, the formula is as follows:

[0091] T=kT 0

[0092] Among them, T is the adjusted operation time, k is the adjustment coefficient calculated by the model according to the patient characteristics, and T 0 is the original operation time;

[0093] For drug treatment plans, the model may adjust the drug dosage based on the patient's liver and kidney function indicators. The formula is as follows:

[0094] D=fD 0

[0095] Where D is the adjusted drug dose, f is the dose adjustment factor calculated by the model, and D 0 The original drug dose.

[0096] S4.2. Use simulation analysis technology to predict the efficacy of the treatment plan after personalized adjustment in S4.1 combined with the patient, and obtain the efficacy prediction results of each treatment plan. The specific steps are as follows:

[0097] Pharmacokinetic model: used to predict the absorption, distribution, metabolism and excretion of drugs in the body, so as to evaluate the efficacy of drugs. The formula is as follows:

[0098] C(t)=C 0 e -kt

[0099] Where C(t) is the concentration of the drug in the body, t is the time, and C 0 is the initial drug concentration, k is the elimination rate constant;

[0100]

[0101] Where E is the drug effect, E maxFor maximum effect, EC 50 The drug concentration that produces 50% of the maximum effect can predict the therapeutic effect of the drug. The drug efficacy is related to the concentration of the drug at a specific target site.

[0102] Disease progression model: constructed based on the natural progression of the disease and the impact of therapeutic interventions. The formula for neurodegenerative diseases is as follows:

[0103] S(t)=S 0 +αt

[0104] Among them, S 0 is the initial disease severity, S is the severity of the disease, t is the time, and α is the disease progression rate;

[0105] When there is a therapeutic intervention, the rate of disease progression may change as follows:

[0106] α 1 =α(1-β)

[0107] Among them, α 1 is the rate of disease progression, β is the inhibitory coefficient of treatment on disease progression, and by adjusting α 1 To simulate the effect of treatment on disease progression and thus predict disease severity after treatment;

[0108] Data preparation: Organize and preprocess the detailed information of the personalized treatment plan and the individual characteristic data of the patient, input the prepared data into the established simulation model, and calculate the values ​​of various efficacy prediction indicators according to the formula;

[0109] Obtain the efficacy prediction results for each treatment plan: For each personalized treatment plan, the model will output the corresponding efficacy prediction results, that is, the predicted values ​​of each efficacy prediction indicator. These results can help doctors and patients understand the expected effects of different treatment plans and provide a reference for treatment decisions. Doctors and patients can choose a more appropriate treatment plan based on these prediction results and other factors (such as treatment risks, patient willingness, etc.).

[0110] S5. Recommend and compare treatment plans based on the efficacy prediction results, recommend treatment plans to doctors for review based on the comparison results, and make overall adaptive modifications to the treatment plans based on the doctors' review opinions.

[0111] The steps of S5 are as follows:

[0112] S5.1. Integrate the efficacy prediction results of each treatment plan for recommendation comparison, select similar treatment plans for better preservation of efficacy prediction results, and then send the compared and saved treatment plans to doctors for review;

[0113] S5.2. Modify the saved treatment plan based on the doctor's review. The specific steps are as follows:

[0114] Integrate the efficacy prediction results: ensure that the efficacy prediction results of each treatment plan have a unified data structure, and then organize the efficacy prediction results of all treatment plans into a two-dimensional matrix;

[0115] Recommended comparison and screening of similar regimens: Based on factors such as the similarity of treatment methods, drugs used or surgical procedures, for drug treatment regimens, if the main drugs used are the same and the drug dosage and frequency of use are similar within a certain range, the two regimens can be considered similar;

[0116] For each group of similar treatment plans, their efficacy prediction results are compared, and weights are assigned to each indicator according to its importance. Then, in each group of similar treatment plans, the plan with the highest comprehensive efficacy score is selected and saved to form a preliminary recommended treatment plan set.

[0117] Send to doctor for review: The preliminary recommended treatment plans are compiled into a detailed report, which includes the specific content of each treatment plan, the predicted results of the efficacy, the comparison with other similar plans, etc. The report is sent to the doctor's work terminal through the hospital's internal information system;

[0118] Opinion classification and analysis: Doctors may raise various opinions during the review process, such as concerns about the safety of the treatment plan, doubts about the accuracy of the efficacy prediction results, and considerations about the feasibility of the treatment plan. The doctor's review opinions are classified and analyzed to determine the specific content and direction that need to be modified. If the doctor believes that the drug dosage of a certain treatment plan may cause a greater burden on the patient's liver and kidney function, this is a safety opinion and the drug dosage needs to be adjusted;

[0119] Plan adjustment: According to the doctor's review opinion, the treatment plan is modified accordingly. If it is an adjustment to the drug dosage, the dosage value is directly modified according to the doctor's advice; if it is an adjustment to the treatment method, it is necessary to re-screen the appropriate plan from the historical treatment plan set, re-predict the efficacy and similarity comparison of the modified treatment plan to ensure that the modified plan still meets the requirements in terms of efficacy and similarity. Feedback the re-evaluation results to the doctor for review again until the doctor is satisfied.

[0120] The above shows and describes the basic principles, main features and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited by the above embodiments. The above embodiments and descriptions are only preferred examples of the present invention and are not intended to limit the present invention. Without departing from the spirit and scope of the present invention, the present invention may have various changes and improvements, which fall within the scope of the present invention. The scope of protection of the present invention is defined by the attached claims and their equivalents.

Claims

1. A method for recommending clinical treatment plans for brain diseases, characterized in that: The following steps are involved: S1. Collect the patient's brain examination data and other body data that can be collected, and then extract disease characteristics and other characteristics from the brain; S2. Collect historical clinical treatment plan sets and corresponding historical patient data, and combine the historical clinical treatment plan sets with historical patient data to score the effectiveness of the treatment plans; S3. Perform treatment adaptability analysis based on other characteristics of the patient, then extract treatment plans based on the treatment adaptability analysis combined with disease characteristics and historical clinical treatment plan sets, and screen the extracted treatment plans by effect scoring; S4. Establish a regimen adjustment model, through which the treatment regimen after screening is adjusted individually according to the patient's disease characteristics and other characteristics, and then predict the efficacy of the adjusted treatment regimen; S5. Recommend and compare treatment plans based on the efficacy prediction results, recommend treatment plans to doctors for review based on the comparison results, and make overall adaptive modifications to the treatment plans based on the doctors' review opinions.

2. A method for recommending clinical treatment plans for brain diseases according to claim 1, characterized in that: The steps of S1 are as follows: S1.

1. By establishing an information collection module in the hospital system, providing an information input channel to doctors through the information collection module, and using the data received through the information input channel as brain examination data and other body data; S1.

2. Disease characteristics and other characteristics are extracted based on brain examination data and other physical data combined with the correlation with the incidence of brain diseases.

3. The method for recommending clinical treatment plans for brain diseases according to claim 1, characterized in that: The brain examination data in S1 include brain imaging data and electrophysiological data, and other physical data include basic information of the patient and description of clinical symptoms; Disease signatures represent key features extracted from imaging data; Other relevant features represent physical characteristics extracted from the patient's basic information.

4. The method for recommending clinical treatment plans for brain diseases according to claim 1, characterized in that: The steps of S2 are as follows: S2.

1. Collect historical clinical treatment plan sets and extract historical patient data corresponding to each clinical treatment plan, so as to build a knowledge base containing various brain disease treatment plans in the information collection module; S2.

2. Extract the brain characteristic data before treatment and the brain characteristic data after treatment from the historical patient data, and combine the historical clinical treatment plan set with the brain characteristic data before treatment and the brain characteristic data after treatment to score the effect of the treatment plan, so as to obtain the effect value of each clinical treatment plan.

5. The method for recommending clinical treatment plans for brain diseases according to claim 1, characterized in that: The efficacy score in S2.2 includes the improvement of disease-related indicators, treatment safety and side effects, and the risk of disease recurrence and progression.

6. The method for recommending clinical treatment plans for brain diseases according to claim 1, characterized in that: The steps of S3 are as follows: S3.

1. Conduct treatment adaptability analysis based on other characteristics of the patient to obtain treatment methods that the patient's physical condition can adapt to; S3.

2. Combine the analyzed treatment methods with the disease characteristics and the historical clinical treatment plan set to extract the treatment plan, and obtain the treatment plan set that is suitable for the patient's disease characteristics; S3.

3. The treatment plans obtained in S3.2 are concentrated and each treatment plan is screened by effect score, so that only the treatment plans with higher effect scores are retained among similar treatment plans.

7. The method for recommending clinical treatment plans for brain diseases according to claim 1, characterized in that: The steps of S4 are as follows: S4.

1. Establish a regimen adjustment model, and then input the selected treatment regimen into the regimen adjustment model in combination with the patient's disease characteristics and other characteristics. The regimen adjustment model will make personalized adjustments to the input treatment regimen based on the individual differences of the patient; S4.

2. Use simulation analysis technology to predict the efficacy of the treatment plan after personalized adjustment in S4.1 in combination with the patient, and obtain the efficacy prediction results of each treatment plan.

8. The method for recommending clinical treatment plans for brain diseases according to claim 7, characterized in that: The formula of S4.1 is as follows: T=kT0 Among them, T is the adjusted operation time, k is the adjustment coefficient calculated by the model according to the patient characteristics, and T0 is the original operation time; D=fD0 Among them, D is the adjusted drug dose, f is the dose adjustment factor calculated by the model, and D0 is the original drug dose.

9. The method for recommending clinical treatment plans for brain diseases according to claim 1, characterized in that: The steps of S5 are as follows: S5.

1. Integrate the efficacy prediction results of each treatment plan for recommendation comparison, select similar treatment plans for better preservation of efficacy prediction results, and then send the compared and saved treatment plans to doctors for review; S5.

2. Make overall adaptation modifications to the saved treatment plan based on the doctor’s review opinion.

Images