A method for recording drug use planning

Through drug use planning and recording methods, registering and classifying the effects of drugs, establishing drug correlation analysis and recommending alternative drugs, the problem of imperfect drug use planning has been solved, and the recommendation of personalized treatment plans and drug selection has been achieved, which has improved the treatment effect and safety sex.

CN118692625BActive Publication Date: 2024-11-22HANGZHOU SHENGFUNUO MEDICAL TECH CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202411188178.3
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-08-28
Publication Date
2024-11-22
Estimated Expiration
2044-08-28

AI Technical Summary

Technical Problem

The drug use plan and records of existing medical institutions are not perfect, and it is difficult to personalize the use of the patient according to the patient's drug use, which makes it difficult to judge drug selection, may miss details, and it is difficult to accurately judge under various drug situations, and lacks systematic Drug tracking and analysis functions.

Method used

Through a drug use planning and recording method, register and classify the positive and negative effects of drugs, establish drug correlation analysis, trigger alarms based on the accumulated negative target effects, and automatically select alternatives with similar positive effects but different negative effects. Drugs, clustering and convolution operations are used to calculate negative effects, combined with pathological indicators to regress the cumulative negative effects, and alternative drugs are recommended.

Benefits of technology

It has achieved tracking and analysis of drug use, assisted medical personnel in recommending personalized treatment plans, improved treatment effect, reduced adverse reactions, timely adjusted treatment plans, prevented drug interactions, enhanced the trust of patients and medical personnel, and Provide data support for medical research.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118692625B_ABST
    Figure CN118692625B_ABST
Patent Text Reader

Abstract

The present application discloses a method for recording drug use planning, including: S1, registering the positive effects, negative effects, positive target effects, and negative target effects of drugs, and classifying drugs according to components and effects; S2, after the doctor diagnoses and prescribes drugs, the drugs are associated according to the user's timeline; S3, based on the analysis of drug associations in S2, an alarm is issued when the cumulative negative target effect reaches a threshold; S4, based on the alarm result in S3, drugs with different negative target effects but similar positive effects are automatically selected. This method for recording drug use planning tracks the patient's drug use, combines other drugs with the same positive effects but different negative effects stored in the system for comparative analysis, recommends appropriate replacement drugs, and assists medical personnel in selecting and replacing drugs, which has a good effect on the treatment of patients and medical research on drugs.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present application relates to the field of drug use planning and recording, and specifically to a drug use planning and recording method. Background Art

[0002] At present, the planning and record of drug use in medical institutions are not perfect. The system only has the function of storing drug use records, and cannot analyze the patient's drug use to assist medical personnel in making judgments on drug selection. Since there are many kinds of drugs to treat a disease, and the positive and negative effects of different drugs also vary in degree and type, it is necessary to adjust the drugs in time according to the patient's use. However, at present, medical personnel can only judge whether the drug needs to be replaced and what drug to replace based on the patient's description of the use and the examination report. Details are easily missed when making judgments, and when there are many types of drugs, it is difficult for humans to make accurate judgments. If the management system can track and record the drugs used by patients, record the positive and negative effects after the patients use them, and recommend suitable replacement drugs based on the patient's condition and the characteristics of the drugs stored in the system, it can assist medical personnel in providing personalized treatment to patients. By recording and analyzing individual responses to specific drugs, it can be more accurate. Providing patients with personalized treatment plans, including drug selection, dosage adjustment, etc., can also improve treatment outcomes. Continuous monitoring of the beneficial effects of drugs can help ensure that the treatment patients are receiving is the most effective. If the drug does not achieve the desired effect, the treatment plan can be adjusted in a timely manner while reducing adverse reactions. Recording and evaluating adverse drug reactions can help to promptly detect and deal with possible side effects and avoid serious consequences. It can also prevent drug interactions. When patients use multiple drugs at the same time, tracking records can help identify potential drug interactions and prevent adverse events. Correspondingly, it can promote patient compliance. Patients can intuitively see the actual improvement that the drug has brought to their condition, which enhances the trust between patients and medical staff. For patients with chronic diseases, long-term drug tracking records can help monitor disease progression and the potential risks that may be caused by long-term use of drugs. The collected drug use data can be used in medical research to help scientists and doctors better understand the mechanism of action of drugs and optimize treatment strategies. Summary of the invention

[0003] The present application provides a drug use planning and recording method to solve the problems in the above-mentioned background technology.

[0004] The technical solution adopted by the present application to solve the technical problem is: a method for recording drug use planning, including:

[0005] S1. Register the positive effects, negative effects, positive target effects, and negative target effects of drugs, and classify drugs according to their components and effects. The data sources for registration and classification include the instructions for the drugs, and use unified terms to describe the effects and components of the drugs;

[0006] S2. After the doctor diagnoses and prescribes the medicine, the medicine is associated according to the user's timeline;

[0007] S3, based on the analysis of drug associations in S2, an alarm is issued when the cumulative negative target effect reaches a threshold;

[0008] S4. Based on the alarm results in S3, drugs with different negative target effects but similar positive effects are automatically selected.

[0009] Preferably, the classification of drugs includes classification according to defined nodes, and the nodes include: drug nodes, representing each drug; ingredient nodes, representing active ingredients or excipients in the drug; action nodes, which are divided into positive effects and negative effects; target effect nodes, which are also divided into positive target effects and negative target effects; and user nodes, which represent users of the drug.

[0010] Preferably, the drug association includes defining edge relationships, associating nodes so that each user node can be associated with edges at a series of time points to form a timeline, wherein the edge relationships include: an inclusion relationship, connecting the drug node and the ingredient node, indicating the ingredients contained in the drug; a production relationship, connecting the drug node and the action node, indicating the effect produced by the drug; an achievement relationship, connecting the drug node and the target effect node, indicating the effect that the drug can achieve; an experience relationship, connecting the user node and the action node, indicating the effect experienced by the user; an achievement relationship, connecting the user node and the target effect node, indicating the target effect achieved by the user; and a taking relationship, connecting the user node and the drug node, indicating that the user has taken a specific drug.

[0011] Preferably, based on the analysis of drug associations in S2, the number and weight of negative target effect nodes connected to the user node are queried to implement an alarm when the cumulative negative target effect reaches a threshold.

[0012] Preferably, automatically selecting drugs with different negative target effects but similar positive effects based on the alarm results in S3 includes: calculating the similarity between drug nodes, and making recommendations based on the positive effects they produce and the negative target effects they avoid.

[0013] Preferably, step S4 is completed by clustering, including:

[0014] 4.1. Feature extraction: First, useful features need to be extracted from the drugs;

[0015] 4.2. Data preparation: standardize or normalize features;

[0016] 4.3. Clustering implementation and clustering result analysis: For each cluster, analyze the common characteristics of its member drugs, with special attention paid to drug groups with similar positive effects but different negative target effects;

[0017] 4.4. Recommend alternative drugs: Based on the user's positive target effect requirements, find the cluster closest to the user's needs. Within this cluster, screen out drugs with different negative target effects but similar positive effects.

[0018] Preferably, when extracting features, the features include: ingredients, the main ingredients and auxiliary ingredients of the drug; positive effects, the therapeutic effects of the drug; negative effects, the adverse reactions that the drug may cause; other related features, including pharmacological properties and bioavailability, and the extracted features are quantified.

[0019] Preferably, K-means clustering is selected as the clustering method, and a suitable K value is selected, where the K value is the number of clusters, and the K value is determined by the elbow rule or the silhouette coefficient method.

[0020] Preferably, when the accumulated negative effect is in the form of a convolution, the accumulated negative effect is calculated by using a convolution operation method, including:

[0021] 3.1. Define the negative effect function: set a time window and record the negative effect intensity at each time point in this window;

[0022] 3.2. Define the influence function: The influence function describes the contribution of the drugs taken in the past to the negative effects at the current moment;

[0023] 3.3. Calculate the convolution and apply it: For any time point t, the cumulative negative effect C(t) can be obtained by convolution of the negative effect function f(t) and the influence function h(t). The calculation formula of the cumulative negative effect C(t) is as follows:

[0024] ,

[0025] Among them, τ represents a time point in the time series, represents the intensity of the negative effect at the time point τ, It represents the degree of negative impact at the time point t-τ.

[0026] Preferably, the cumulative negative target effect is calculated by reversely calculating the accumulation of negative effects according to the patient's pathological indicators, including:

[0027] A1: Collect pathological index data: Collect the patient's pathological index data, including blood tests, liver and kidney function tests, and electrocardiograms, and record the changes of these indicators over time, especially during medication;

[0028] A2: Establish a negative effect model: Based on the drug instructions and clinical studies, establish a negative effect model to describe the possible negative effects of each drug and its impact on pathological indicators;

[0029] A3: Establish indicator-negative effect mapping: Based on clinical experience and literature, establish the mapping relationship between pathological indicators and negative effects;

[0030] A4: Establish an inverse model: Based on the negative effect model and indicator-negative effect mapping, build an inverse model to estimate the cumulative negative effects;

[0031] A5: Calculate the cumulative negative effects: Calculate the cumulative negative effects based on the changes in the patient's pathological indicators based on the inverse model.

[0032] The substantial effect of the present application is that the present drug use planning and recording method tracks the use of drugs by patients, compares and analyzes other drugs with the same positive effects but different negative effects stored in the system, recommends appropriate replacement drugs, and assists medical personnel in selecting and replacing drugs, which has a good effect on the treatment of patients and medical research on drugs. BRIEF DESCRIPTION OF THE DRAWINGS

[0033] Figure 1 It is a flowchart of the steps of an embodiment of the present application. DETAILED DESCRIPTION

[0034] The technical solution of the present application is further described below through specific embodiments.

[0035] A method for recording drug use planning, comprising:

[0036] S1. Register the positive effects, negative effects, positive target effects, and negative target effects of drugs, and classify drugs according to their components and effects. The data sources for registration and classification include the instructions for the drugs, and use unified terms to describe the effects and components of the drugs;

[0037] Among them, the classification of drugs includes classification according to definition nodes, which include: drug nodes, representing each drug; ingredient nodes, representing the active ingredients or excipients in the drug; action nodes, which are divided into positive and negative effects; target effect nodes, which are also divided into positive target effects and negative target effects; user nodes, representing the user of the drug. Drug instructions can be used as a data source for node and edge attributes, such as ingredient information, action description, etc. Unified wording can be achieved by establishing a standard vocabulary to ensure the consistency and clarity of each term.

[0038] S2. After the doctor diagnoses and prescribes the medicine, the medicine is associated according to the user's timeline. Specifically, the medicine association includes defining edge relationships and associating nodes so that each user node can be associated with edges at a series of time points to form a timeline. The use of the medicine can be tracked by timestamps, and the effects experienced by the user and the target effects achieved can be updated accordingly. Among them, the edge relationships include: inclusion relationship, connecting the drug node and the ingredient node, indicating the ingredients contained in the medicine; production relationship, connecting the drug node and the action node, indicating the effect produced by the medicine; achievement relationship, connecting the drug node and the target effect node, indicating the effect that the medicine can achieve; experience relationship, connecting the user node and the action node, indicating the effect experienced by the user; achievement relationship, connecting the user node and the target effect node, indicating the target effect achieved by the user; taking relationship, connecting the user node and the drug node, indicating that the user has taken a specific medicine.

[0039] S3. Based on the analysis of drug associations in S2, an alarm is issued when the cumulative negative target effect reaches a threshold. Specifically, the number and weight of negative target effect nodes connected to the user node can be queried to implement an alarm when the cumulative negative target effect reaches a threshold.

[0040] When the accumulated negative effects are expressed in a convolution form, the accumulated negative effects can be calculated using the convolution operation method, including:

[0041] 3.1. Define the negative effect function: set a time window and record the negative effect intensity at each time point in this window. The negative effect intensity can be a scalar value, indicating the degree of negative effect at a certain time point;

[0042] 3.2. Define the influence function: The influence function describes the contribution of the drugs taken in the past to the negative effects at the current moment. This function can be empirical or based on clinical research;

[0043] 3.3. Calculate the convolution and apply it: For any time point t, the cumulative negative effect C(t) can be obtained by convolution of the negative effect function f(t) and the influence function h(t). The calculation formula of the cumulative negative effect C(t) is as follows:

[0044] ,

[0045] Among them, τ represents a time point in the time series, represents the intensity of the negative effect at the time point τ, It represents the degree of negative effects at the time point t-τ. In practical applications, since negative effects are usually only considered within a limited time window, the integral range can be simplified to a finite interval starting from the time of the most recent medication. When applying convolution, if we consider the discrete case, the cumulative negative effects can be expressed as:

[0046] , where f[τ] is the negative effect intensity at time point τ, and h[t-τ] is the value of the influence function at time t-τ.

[0047] For example, a patient takes a drug every other day, and the side effects after each dose last for two days. We can set a simple side effect function and impact function to simulate this situation:

[0048] Negative effect function: f[0]=1, indicating the negative effect intensity of the first medication; f[2]=1, indicating the negative effect intensity of the second medication, and f[t]=0 at other time points;

[0049] Impact function: h[t]=0.5, when t is in [0, 1], it means that the negative effect lasts for two days and the intensity decreases;

[0050] Cumulative negative effect calculation:

[0051] C[0]=f[0]h[0]=1×0.5=0.5;

[0052] C[1]=f[0]h[1]+f[-1]h[2]=1×0.5+0×0.5=0.5;

[0053] C[2]=f[2]h[0]+f[0]h[2]=1×0.5+1×0.5=1.0;

[0054] C[3]=f[2]h[1]+f[1]h[2]=1×0.5+0×0.5=0.5.

[0055] Over time, the negative effect function is continuously updated and the cumulative negative effects are recalculated. In this way, we can effectively calculate the cumulative negative effects and monitor when the cumulative negative effects exceed a certain threshold, thereby triggering an alarm. This helps to adjust the treatment plan in time to avoid potential risks.

[0056] S4. Based on the alarm results in S3, drugs with different negative target effects but similar positive effects are automatically selected. Specifically, the similarity between drug nodes can be calculated, and recommendations can be made based on the positive effects they produce and the negative target effects they avoid.

[0057] Step S4 may also be completed by clustering. The clustering method is K-means clustering, and a suitable K value is selected. The K value is the number of clusters. The K value is determined by the elbow rule or the silhouette coefficient method, which specifically includes:

[0058] 4.1. Feature extraction: First, useful features need to be extracted from the drug. The features include: ingredients, the main ingredients and auxiliary ingredients of the drug; positive effects, the therapeutic effects of the drug; negative effects, the adverse reactions that the drug may cause; other related features, such as pharmacological properties and bioavailability, and the extracted features are quantified;

[0059] 4.2. Data preparation: Standardize or normalize the features to facilitate subsequent clustering algorithm processing. If necessary, the data can also be reduced in dimensionality, such as using PCA (principal component analysis) to reduce the number of features;

[0060] 4.3. Clustering implementation and clustering result analysis: For each cluster, analyze the common characteristics of its member drugs, with special attention paid to drug groups with similar positive effects but different negative target effects;

[0061] 4.4. Recommend alternative drugs: Based on the user's positive target effect requirements, find the cluster closest to the user's needs. Within this cluster, screen out drugs with different negative target effects but similar positive effects.

[0062] In this way, we can use cluster analysis to recommend alternative drugs with similar positive effects but lower negative target effects, thereby helping doctors and patients make better treatment choices.

[0063] It should be noted that the cumulative negative target effect can also be calculated by inferring the accumulation of negative effects based on the patient's pathological indicators, including:

[0064] A1: Collect pathological index data: Collect the patient's pathological index data, including blood tests, liver and kidney function tests, electrocardiograms, and record the changes of these indicators over time, especially the changes during medication. Collect the patient's pathological index data before and after medication, and record the time and dosage of medication;

[0065] A2: Establish a negative effect model: Based on the drug instructions and clinical studies, establish a negative effect model to describe the possible negative effects of each drug and its impact on pathological indicators;

[0066] A3: Establish indicator-negative effect mapping: Based on clinical experience and literature, establish the mapping relationship between pathological indicators and negative effects;

[0067] A4: Establish an inverse model: Based on the negative effect model and indicator-negative effect mapping, build an inverse model to estimate the cumulative negative effects, use cross-validation and other methods to verify the accuracy and reliability of the model, adjust the model parameters, and improve the prediction accuracy;

[0068] A5: Calculate the cumulative negative effects: Calculate the cumulative negative effects based on the changes in the patient's pathological indicators based on the inverse model.

[0069] For example, when collecting data, the patient's liver enzyme level during medication is recorded, and the time and dosage of medication are recorded; when building a model, a negative effect model that may cause liver toxicity for each drug is established based on the drug instructions and literature, and a mapping relationship between liver enzyme levels and liver toxicity is constructed; the inverse model uses historical data to train the inverse model, inputs changes in liver enzyme levels, and outputs the cumulative negative effect intensity of liver toxicity. Methods such as linear regression, support vector machine or neural network can be used; model verification can use methods such as cross-validation to verify the accuracy and stability of the model, and adjust the model parameters according to the verification results; then the inverse model is applied to calculate the patient's cumulative negative effects of liver toxicity, monitor the changing trend of the cumulative negative effects, and adjust the treatment plan as needed. In this way, we can reverse the cumulative negative effects based on the changes in the patient's pathological indicators, which is very important for optimizing treatment plans and preventing adverse reactions.

[0070] The above-described embodiment is only a preferred solution of the present application and does not limit the present application in any form. There are other variations and modifications without exceeding the technical solution described in the claims.

Claims

1. A method for recording drug use planning, characterized in that: include: S1. Register the positive effects, negative effects, positive target effects, and negative target effects of drugs, and classify drugs according to their components and effects. The data sources for registration and classification include the instructions for the drugs, and unified terms are used to describe the effects and components of the drugs. Classification of drugs includes classification according to definition nodes; S2. After the doctor diagnoses and prescribes the medicine, the medicine is associated according to the user's timeline. The drug association includes defining edge relationships and associating nodes so that each user node can be associated with edges at a series of time points to form a timeline; S3, based on the analysis of drug associations in S2, query the number and weight of negative target effect nodes connected to the user node, and issue an alarm when the cumulative negative target effect reaches a threshold; S4. Based on the alarm results in S3, drugs with different negative target effects but similar positive effects are automatically selected in a clustering manner. This step includes: 4.

1. Feature extraction: First, useful features need to be extracted from the drugs; 4.

2. Data preparation: standardize or normalize features; 4.

3. Clustering implementation and clustering result analysis: For each cluster, analyze the common characteristics of its member drugs, with special attention paid to drug groups with similar positive effects but different negative target effects; 4.

4. Recommend alternative drugs: Based on the user's positive target effect requirements, find the cluster closest to the user's needs. Within this cluster, screen out drugs with different negative target effects but similar positive effects.

2. The method for recording drug use planning according to claim 1, characterized in that: The classification of drugs includes classification according to definition nodes, and the nodes include: drug nodes, which represent each drug; ingredient nodes, which represent active ingredients or excipients in the drug; action nodes, which are divided into positive effects and negative effects; target effect nodes, which are also divided into positive target effects and negative target effects; and user nodes, which represent the users of the drug.

3. The drug use planning recording method according to claim 2, characterized in that: Drug association includes defining edge relationships and associating nodes so that each user node can be associated with edges at a series of time points to form a timeline, where the edge relationships include: inclusion relationship, connecting the drug node and the ingredient node, indicating the ingredients contained in the drug; production relationship, connecting the drug node and the action node, indicating the effect produced by the drug; achievement relationship, connecting the drug node and the target effect node, indicating the effect that the drug can achieve; experience relationship, connecting the user node and the action node, indicating the effect experienced by the user; achievement relationship, connecting the user node and the target effect node, indicating the target effect achieved by the user; and taking relationship, connecting the user node and the drug node, indicating that the user has taken a specific drug.

4. The method for recording drug use planning according to claim 3, characterized in that: Automatically selecting drugs with different negative target effects but similar positive effects based on the alarm results in S3 includes: calculating the similarity between drug nodes, and making recommendations based on the positive effects they produce and the negative target effects they avoid.

5. The method for recording drug use planning according to claim 1, characterized in that: When extracting features, the features include: ingredients, the main ingredients and auxiliary ingredients of the drug; positive effects, the therapeutic effects of the drug; negative effects, the adverse reactions that the drug may cause; other related features, including pharmacological properties and bioavailability, and the extracted features are quantified.

6. The method for recording drug use planning according to claim 5, characterized in that: Select K-means clustering as the clustering method and select an appropriate K value, which is the number of clusters. The K value is determined by the elbow rule or silhouette coefficient method.

7. The method for recording drug use planning according to claim 1, characterized in that: When the accumulated negative effects are expressed in a convolutional form, the accumulated negative effects are calculated using a convolution operation method, including: 3.

1. Define the negative effect function: set a time window and record the negative effect intensity at each time point in this window; 3.

2. Define the influence function: The influence function describes the contribution of the drugs taken in the past to the negative effects at the current moment; 3.

3. Calculate the convolution and apply it: For any time point t, the cumulative negative effect C(t) can be obtained by convolution of the negative effect function f(t) and the influence function h(t). The calculation formula of the cumulative negative effect C(t) is as follows: , Among them, τ represents a time point in the time series, represents the intensity of the negative effect at the time point τ, It represents the degree of negative impact at the time point t-τ.

8. The method for recording drug use planning according to claim 1, characterized in that: The cumulative negative target effects are calculated by inferring the accumulation of negative effects based on the patient's pathological indicators, including: A1: Collect pathological index data: Collect the patient's pathological index data, including blood tests, liver and kidney function tests, and electrocardiograms, and record the changes of these indicators over time, especially during medication; A2: Establish a negative effect model: Based on the drug instructions and clinical studies, establish a negative effect model to describe the possible negative effects of each drug and its impact on pathological indicators; A3: Establish indicator-negative effect mapping: Based on clinical experience and literature, establish the mapping relationship between pathological indicators and negative effects; A4: Establish an inverse model: Based on the negative effect model and indicator-negative effect mapping, build an inverse model to estimate the cumulative negative effects; A5: Calculate the cumulative negative effects: Calculate the cumulative negative effects based on the changes in the patient's pathological indicators based on the inverse model.