Tumor chemotherapy patient vomit-stopping medication management method and related device
By building a comprehensive database and knowledge graph reasoning algorithm, dynamically predicting CINV risk and optimizing antiemetic drug treatment regimens, we solved the problem of poor efficacy of chemotherapy-related vomiting and improved the efficiency of antiemetic drug use and patient treatment compliance.
Patent Information
- Application Number
- CN202510674564.1
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-05-23
- Publication Date
- 2025-09-12
AI Technical Summary
The incidence of chemotherapy-related vomiting (CINV) is high. The irrational selection of existing antiemetic drugs leads to poor efficacy, inability to effectively deal with complex vomiting mechanisms, and inappropriate administration time leads to insufficient drug efficacy.
A comprehensive database was constructed, combining patient physiological parameters and chemotherapy drug information, and personalized antiemetic drug treatment plans were recommended through knowledge graph reasoning algorithms. CINV risk was dynamically predicted, and the probability of CINV occurrence was predicted through ARIMA and GBDT models, optimizing antiemetic drug treatment plans to minimize risks and adverse reactions.
It improves the efficacy of antiemetic drugs, reduces unnecessary drug use and adverse reactions, and enhances the medication compliance and treatment effect of chemotherapy patients.
Smart Images
Figure CN120636672A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of medical information technology, and in particular to a method and device for managing antiemetic medication for patients undergoing chemotherapy for tumors, and a computing device. Background Art
[0002] Chemotherapy is an important treatment for cancer, but many chemotherapy drugs can cause adverse reactions, including vomiting, known as chemotherapy-induced vomiting (CINV). The occurrence of CINV not only increases patient suffering but also severely impacts adherence to treatment regimens, thereby compromising treatment efficacy. Among cancer patients receiving chemotherapy without antiemetic measures, 70% to 80% experience CINV, which is typically categorized as acute, delayed, explosive, refractory, and anticipatory nausea and vomiting.
[0003] In order to alleviate CINV, clinical practice usually administers antiemetics to cancer patients while they are receiving anti-tumor drug treatment. However, due to reasons such as unreasonable selection of antiemetics and unreasonable administration time, the efficacy of antiemetics is often poor. For example, some antiemetics have a short half-life, and when vomiting occurs, the antiemetics may have been basically metabolized and cannot be effectively exerted. For situations where antiemetics with different mechanisms of action need to be used in combination to cope with complex vomiting mechanisms, antiemetics with a single mechanism of action may not achieve the ideal therapeutic effect. Administering the antiemetic too early or too late may result in a low blood concentration of the antiemetic when the patient vomits, making it unable to exert its effect.
[0004] Therefore, in order to solve the above problems, the present invention proposes a method for managing antiemetic medication for tumor chemotherapy patients to overcome the problem of poor efficacy of antiemetic drugs for chemotherapy patients, thereby improving the medication compliance of chemotherapy patients. Summary of the Invention
[0005] In view of the above problems, the present invention provides a method and apparatus, as well as a computing device, for managing antiemetic medication for patients undergoing chemotherapy for tumors.
[0006] According to one aspect of the present invention, a method for managing antiemetic medication for patients undergoing chemotherapy for tumors is provided, comprising:
[0007] Construct a comprehensive database containing the emetogenicity level of chemotherapy drugs, patient risk factors, and antiemetic drug information. The antiemetic drug information includes drug type, dosage form, dosage, route of administration, administration time, treatment course, half-life, liver and kidney function requirements, adverse reactions, and drug interactions. The patient risk factors include age, gender, previous history of vomiting, and anxiety level. A baseline database of patient physiological parameters including blood routine, liver and kidney function, electrolytes, and electrocardiogram indicators is established.
[0008] Dynamically predicting CINV warning information for the patient based on the comprehensive database, wherein the warning information includes CINV risk level, CINV risk probability, severity, and expected time of occurrence;
[0009] Based on the CINV warning information, the comprehensive database, the patient physiological parameter baseline database, the patient's drug allergy history, and drug interaction information, a knowledge graph reasoning algorithm is used to automatically match and recommend a personalized antiemetic drug treatment plan; wherein the antiemetic drug treatment plan includes the type of antiemetic drug, dosage form, dosage, route of administration, time of first administration, frequency of subsequent administration, and course of treatment;
[0010] The antiemetic drug treatment plan is sent to the nursing mobile terminal to remind clinical nurses to give antiemetic drugs and record the administration time and route; and the recorded patient vomiting, nausea level and adverse reaction information are fed back.
[0011] In an optional manner, the dynamically predicting the patient's CINV warning information based on the comprehensive database further includes:
[0012] Establishing a time series database based on the patient's CINV incidence data, the emetogenicity level of chemotherapy drugs, patient risk factors, and the use of antiemetic drugs;
[0013] The CINV occurrence data is decomposed into time series using an autoregressive integrated moving average model to predict the probability of CINV occurrence in the future; wherein the autoregressive integrated moving average model is ARIMA(p, d, q), where p is the order of the autoregressive term, d is the order of the difference term, and q is the order of the moving average term;
[0014] The patient's age, gender, previous vomiting history, and anxiety level were used as input features, and a gradient boosting decision tree was used to train a GBDT classification model to predict the risk level of CINV.
[0015] Comprehensive CINV warning information is constructed based on the CINV occurrence probability predicted by the ARIMA model and the CINV risk level predicted by the GBDT classification model.
[0016] In an optional embodiment, the automatic matching and recommendation of personalized antiemetic drug treatment plans by the knowledge graph reasoning algorithm further includes:
[0017] The nodes of the knowledge graph are entities including chemotherapy drugs, antiemetic drugs, and patient diseases, and the edges of the knowledge graph are relationships between entities;
[0018] According to the patient's CINV warning information, drug allergy history, and drug interaction information, the best antiemetic drug treatment plan is searched in the knowledge graph using the PathRanking algorithm.
[0019] In an optional embodiment, the method for recommending a personalized antiemetic drug treatment regimen further comprises:
[0020] Three optimization objectives are defined, wherein the optimization objectives include minimizing the CINV risk score S(t), minimizing the incidence of adverse drug reactions P, and minimizing the treatment cost C total ;
[0021] The multi-objective problem is transformed into a single-objective optimization through the ε constraint method. The optimization formula is:
[0022] min(S(t)+max(P adverse -P max ,0)+max(C total -C max ,0))
[0023] Among them, P adverse is the actual observed or predicted incidence of adverse drug reactions; P max is the maximum acceptable threshold for the incidence of adverse drug reactions; C max is the maximum acceptable threshold for total treatment cost.
[0024] In an optional manner, the minimization of the CINV risk score S(t) satisfies the Bellman equation, which is:
[0025] S(t)=min u(t) L(x(t),u(t),t)+E[S(t+1)|x(t),u(t)]
[0026] where u(t) is the antiemetic treatment regimen taken at time t; x(t) is the patient's state at time t; L(x(t),u(t),t) is the immediate CINV risk resulting from treatment u(t) at time t when the patient's state is x(t); and E[S(t+1)|x(t),u(t)] is the expected optimal CINV risk score at time t+1, conditional on the state x(t) and treatment u(t) at time t.
[0027] In an optional manner, the minimization of the incidence of adverse drug reactions P and the minimization of treatment costs C total The constraint is embedded into the single-objective optimization function through the Lagrangian relaxation method. The optimization formula of the single-objective optimization function is:
[0028] min uS(t)+λ1(P adverse (u)-P max ) + +λ2(C total (u)-C max ) +
[0029] Where u is the antiemetic drug treatment regimen; (.) + is a positive function; λ1 is the penalty factor when the adverse drug reaction violates the constraint; λ2 is the penalty factor when the treatment cost violates the constraint.
[0030] In an optional manner, the method further includes:
[0031] Assigning time-sensitive weights to the historical CINV probability based on the patient's real-time physiological data;
[0032] A two-stage parameter optimization objective function is constructed, which consists of the first and second stages. In the first stage, a genetic algorithm is used to search for the minimum parameter combination (p, d, q) that satisfies the AIC criterion. In the second stage, the Bayesian Information Criterion is used to regularize the candidate model. The constraint condition of the Bayesian Information Criterion is:
[0033] BIC=-2lnL+ln(n)·(1+ω(t))
[0034] Where L is the model likelihood function; n is the sample size; ω(t) is the weight function; ω(t) = e -Δt ·[1+α·Var(x physio (t))]; Δt is the data time interval; α is the physiological parameter fluctuation adjustment factor; Var(x physio (t)) is the variance of the patient's real-time physiological parameters at time t; x physio (t) is the patient's real-time physiological parameter value at time t;
[0035] When it is detected that the change rate of chemotherapy drug infusion rate is greater than the preset threshold, the ARIMA model is forced to switch to the low-order configuration.
[0036] In an optional manner, the constraint embedding process of the Lagrangian relaxation method further includes:
[0037] Construct a nonlinear constraint function for the incidence of adverse drug reactions and convert the continuous probability density function into discrete constraints through piecewise linearization.
[0038] The multi-objective optimization problem is transformed into a constrained stochastic differential equation, and the expected optimal solution of the nonlinear constraint function is calculated using the Ito formula.
[0039] According to another aspect of the present invention, a device for managing antiemetic medication for patients undergoing chemotherapy for tumors is provided, comprising:
[0040] A data construction module is used to construct a comprehensive database containing the emetogenicity level of chemotherapy drugs, patient risk factors, and antiemetic drug information, wherein the antiemetic drug information includes drug type, dosage form, dosage, route of administration, administration time, treatment course, half-life, liver and kidney function requirements, adverse reactions and drug interactions. The patient risk factors include age, gender, previous vomiting history, and anxiety level; and establish a patient physiological parameter baseline database including the patient's blood routine, liver and kidney function, electrolytes, and electrocardiogram indicators;
[0041] A CINV early warning module, configured to dynamically predict CINV early warning information of a patient based on the comprehensive database, wherein the early warning information includes CINV risk level, CINV risk probability, severity, and expected occurrence time;
[0042] a regimen recommendation module for automatically matching and recommending personalized antiemetic drug treatment regimens based on the CINV warning information, the comprehensive database, the patient physiological parameter baseline database, the patient's drug allergy history, and drug interaction information through a knowledge graph reasoning algorithm; wherein the antiemetic drug treatment regimen includes the type of antiemetic drug, dosage form, dosage, route of administration, time of first administration, frequency of subsequent administration, and course of treatment;
[0043] The execution and feedback module is used to send the antiemetic drug treatment plan to the nursing mobile terminal to remind clinical nurses to give antiemetic drugs and record the administration time and route; and to provide feedback on the recorded patient vomiting, nausea level and adverse reaction information.
[0044] According to another aspect of the present invention, there is provided a computing device, comprising: a processor, a memory, a communication interface, and a communication bus, wherein the processor, the memory, and the communication interface communicate with each other via the communication bus;
[0045] The memory is used to store at least one executable instruction, and the executable instruction enables the processor to execute operations corresponding to the above-mentioned method for managing antiemetic medication for patients undergoing tumor chemotherapy.
[0046] According to the solution provided by the present invention, a comprehensive database containing the emetogenicity level of chemotherapy drugs, patient risk factors and antiemetic drug information is constructed, wherein the antiemetic drug information includes drug type, dosage form, dosage, route of administration, administration time, course of treatment, half-life, liver and kidney function requirements, adverse reactions and drug interactions, and the patient risk factors include age, gender, previous vomiting history and anxiety level; a patient physiological parameter baseline database containing the patient's blood routine, liver and kidney function, electrolytes and electrocardiogram indicators is established; the patient's CINV early warning information is dynamically predicted based on the comprehensive database, wherein the early warning information includes CINV risk level, CINV risk profile, and CINV risk level. rate, severity and expected time of occurrence; based on the CINV warning information, the comprehensive database, the patient physiological parameter baseline database, the patient's drug allergy history and drug interaction information, automatically match and recommend personalized antiemetic drug treatment plans through the knowledge graph reasoning algorithm; wherein, the antiemetic drug treatment plan includes the type of antiemetic drug, dosage form, dosage, route of administration, first administration time, subsequent administration frequency and course of treatment; the antiemetic drug treatment plan is sent to the nursing mobile terminal to remind clinical nurses to give antiemetic drugs and record the administration time and route; and the recorded patient vomiting, nausea level and adverse reaction information are fed back. The present invention matches appropriate antiemetic drugs and provides antiemetic treatment plans based on therapeutic principles and drug properties, and promptly reminds clinical nurses to give the optimal interval time of antiemetic drugs to ensure the best efficacy of antiemetic drugs. Specifically, by constructing a comprehensive database and a patient physiological parameter baseline database and combining the patient's specific situation, the patient's CINV warning information is dynamically predicted, and the knowledge graph reasoning algorithm is used to automatically match and recommend personalized antiemetic drug treatment plans, reducing unnecessary drug use and adverse reactions. The multi-objective problem is converted into a single-objective optimization through the ε constraint method, ensuring that the optimal treatment plan is found while meeting multiple goals. The antiemetic drug treatment plan is sent to the nursing mobile terminal to remind clinical nurses to give antiemetic drugs and record the time and route of administration. The best antiemetic drug treatment plan is searched in the knowledge graph through the Path Ranking algorithm, further improving the accuracy of the recommended plan. The historical CINV occurrence probability is assigned a time-sensitive weight based on the patient's real-time physiological data, and a two-stage parameter optimization objective function is constructed. The model is optimized using the Bayesian Information Criterion to ensure the adaptability of the prediction model. When the change rate of the chemotherapy drug infusion rate is detected to be greater than the preset threshold, the ARIMA model is forced to switch to a low-order configuration so that the model can remain accurate under dynamic changes.
[0047] The above description is only an overview of the technical solution of the present invention. In order to more clearly understand the technical means of the present invention, it can be implemented in accordance with the contents of the specification. In order to make the above and other purposes, features and advantages of the present invention more obvious and easy to understand, the specific implementation methods of the present invention are specifically listed below. BRIEF DESCRIPTION OF THE DRAWINGS
[0048] Various other advantages and benefits will become apparent to those skilled in the art upon reading the detailed description of the preferred embodiment below. The accompanying drawings are for illustration purposes only and are not to be considered as limiting the present invention. The same reference symbols are used throughout the drawings to represent the same components. In the drawings:
[0049] Figure 1 A schematic flow chart showing a method for managing antiemetic medication for patients undergoing chemotherapy for tumors according to an embodiment of the present invention is shown;
[0050] Figure 2 A schematic diagram showing a framework of an antiemetic medication management device for cancer chemotherapy patients according to an embodiment of the present invention;
[0051] Figure 3 A schematic structural diagram of a computing device according to an embodiment of the present invention is shown. DETAILED DESCRIPTION
[0052] Exemplary embodiments of the present invention will be described in more detail below with reference to the accompanying drawings. Although exemplary embodiments of the present invention are shown in the accompanying drawings, it should be understood that the present invention can be implemented in various forms and should not be limited by the embodiments set forth herein. Rather, these embodiments are provided to enable a more thorough understanding of the present invention and to fully convey the scope of the present invention to those skilled in the art.
[0053] Figure 1 The flowchart of the antiemetic medication management method for cancer chemotherapy patients according to an embodiment of the present invention is shown. Figure 1 As shown, the following steps are included:
[0054] Step S101, constructing a comprehensive database containing the emetogenicity level of chemotherapy drugs, patient risk factors and antiemetic drug information, wherein the antiemetic drug information includes drug type, dosage form, dosage, route of administration, administration time, course of treatment, half-life, liver and kidney function requirements, adverse reactions and drug interactions, and the patient risk factors include age, gender, previous vomiting history and anxiety level; establishing a patient physiological parameter baseline database including the patient's blood routine, liver and kidney function, electrolytes and electrocardiogram indicators.
[0055] In this embodiment, the emetogenicity level of chemotherapy drugs, the patient's risk factors (such as age, gender, previous history of vomiting, anxiety level) and antiemetic drug information (such as drug type, dosage form, dosage, route of administration, administration time, course of treatment, half-life, liver and kidney function requirements, adverse reactions and drug interactions) are collected. In addition, the patient's blood routine, liver and kidney function, electrolyte and electrocardiogram indicators are collected to establish a baseline database of the patient's physiological parameters. The CINV occurrence data are decomposed into time series using an autoregressive integral sliding average model to predict the future probability of CINV occurrence. The classification model is trained using a gradient boosting decision tree to predict the risk level of CINV. A knowledge graph containing chemotherapy drugs, antiemetic drugs and patient diseases is constructed, with nodes being entities and edges being the relationships between entities. Based on the patient's CINV warning information, drug allergy history and drug interaction information, the best antiemetic drug treatment plan is searched in the knowledge graph using the Path Ranking algorithm.
[0056] For example, a patient is receiving chemotherapy, and the chemotherapy drug has a high emetogenicity level. The patient has a history of vomiting and is highly anxious. The following information is obtained through the comprehensive database and the patient's physiological parameter baseline database:
[0057] Age: 45
[0058] Gender: Female
[0059] Previous history of vomiting: Yes
[0060] Anxiety level: High
[0061] Chemotherapy drug information:
[0062] Emeticity level: High
[0063] Drug type: Ondansetron
[0064] Dosage form: Tablet
[0065] Dosage: 8 mg
[0066] Route of administration: Oral
[0067] Administration time: 30 minutes before chemotherapy
[0068] Treatment course: once a day for 5 days
[0069] Half-life: 3.5 hours
[0070] Liver and kidney function requirements: normal
[0071] Adverse reactions: headache, fatigue
[0072] Drug interactions: interacts with certain antibiotics A
[0073] Blood test: normal
[0074] Liver and kidney function: normal
[0075] Electrolytes: Normal
[0076] Electrocardiogram: normal
[0077] Through dynamic prediction and knowledge graph reasoning algorithms, the following personalized antiemetic drug treatment plan is recommended:
[0078] Antiemetic drug type: Ondansetron
[0079] Dosage form: Tablet
[0080] Dosage: 8 mg
[0081] Route of administration: Oral
[0082] First dose time: 30 minutes before chemotherapy
[0083] Subsequent dosing frequency: once a day
[0084] Treatment course: 5 days
[0085] Through the mobile nursing platform, clinical nurses administer antiemetics according to the recommended treatment plan and record the time and route of administration. At the same time, they provide feedback on the patient's vomiting, nausea level, and adverse reactions, forming a closed-loop management system.
[0086] Step S102: dynamically predicting the patient's CINV warning information based on the comprehensive database, wherein the warning information includes CINV risk level, CINV risk probability, severity, and expected occurrence time.
[0087] In this embodiment, not only is CINV predicted, but it is also further refined to the risk level, probability, severity and expected time of occurrence, thereby providing more comprehensive early warning information. For example, the chemotherapy drug for patient A is cisplatin (highly emetogenic). The risk factors of patient A are obtained from the comprehensive database: age 60 years old, male, no history of vomiting, and moderate anxiety level. The blood routine, liver and kidney function and other indicators of patient A are obtained from the patient physiological parameter baseline database, and the historical CINV data of similar patients are obtained from the time series database. It is predicted that the probability of CINV occurrence of patient A within 24 hours after chemotherapy is 70%, the CINV risk level is "medium", the severity of CINV: expected mild to moderate, and the expected time of occurrence: within 6-24 hours after chemotherapy. The above early warning information is sent to the nursing mobile terminal to remind the nurse to pay attention to the CINV situation of patient A and give antiemetic drugs as prescribed by the doctor.
[0088] In an optional manner, the dynamically predicting the patient's CINV warning information based on the comprehensive database further includes:
[0089] Establishing a time series database based on the patient's CINV incidence data, the emetogenicity level of chemotherapy drugs, patient risk factors, and the use of antiemetic drugs;
[0090] The CINV occurrence data is decomposed into time series using an autoregressive integrated moving average model to predict the probability of CINV occurrence in the future; wherein the autoregressive integrated moving average model is ARIMA(p, d, q), where p is the order of the autoregressive term, d is the order of the difference term, and q is the order of the moving average term;
[0091] The patient's age, gender, previous vomiting history, and anxiety level were used as input features, and a gradient boosting decision tree was used to train a GBDT classification model to predict the risk level of CINV.
[0092] Comprehensive CINV warning information is constructed based on the CINV occurrence probability predicted by the ARIMA model and the CINV risk level predicted by the GBDT classification model.
[0093] In this example, the ARIMA model and the GBDT model were combined to fully leverage the temporal relationships of time series data and the classification capabilities of patient risk factors, improving the reliability of CINV prediction. The GBDT model uses methods such as feature importance analysis to understand the impact of each risk factor on CINV risk level, increasing the model's interpretability and helping clinicians understand and trust the model's predictions.
[0094] Step S103, based on the CINV warning information, the comprehensive database, the patient physiological parameter baseline database, the patient's drug allergy history and drug interaction information, automatically match and recommend personalized antiemetic drug treatment plans through a knowledge graph reasoning algorithm; wherein, the antiemetic drug treatment plan includes the type of antiemetic drug, dosage form, dosage, route of administration, first administration time, subsequent administration frequency and course of treatment.
[0095] In this embodiment, the knowledge graph reasoning algorithm automatically analyzes complex medical data, discovers the associations hidden in the data, reduces the time of manual data review and judgment, and by considering drug interactions and patient physiological parameters as well as optimization goals (minimizing CINV risk, adverse reaction incidence and treatment costs), the recommended plan further reduces the occurrence of adverse reactions and avoids unnecessary drug use.
[0096] Specifically, the knowledge graph nodes (entities) include chemotherapy drugs (such as cisplatin and cyclophosphamide), antiemetics (such as ondansetron, dexamethasone, and aprepitant), patient diseases (such as breast cancer and lung cancer), patient risk factors (such as age, gender, history of vomiting, and anxiety level), adverse reactions (such as constipation and headache), and physiological indicators (such as liver function and renal function), among others. Edges (relationships) include chemotherapy drugs (with an emetogenicity rating of "high"), ondansetron (used to treat "chemotherapy-induced nausea and vomiting"), ondansetron (dose of "8 mg"), ondansetron (route of administration of "intravenous"), and dexamethasone (may cause "hyperglycemia"). Ondansetron and dexamethasone exhibit synergistic effects, and renal insufficiency requires adjustment of ondansetron dosage. The knowledge graph is stored in a graph database (such as Neo4j). Using a patient's CINV warning information, drug allergy history, and drug interaction information as a starting point, the knowledge graph is searched for the optimal antiemetic treatment plan. Constraints are imposed to exclude medications to which the patient is allergic and to avoid combinations with adverse drug interactions. The search objective is to find a drug combination that effectively reduces the risk of CINV, minimizes side effects, and meets the patient's physiological parameters and treatment needs. Different pathways are scored based on the weights of the edges in the knowledge graph (e.g., the efficacy of the antiemetic drug, the incidence of adverse reactions, and the strength of its interactions with other medications). The highest-scoring pathway is selected as the recommended solution.
[0097] For example, consider Patient A, a 55-year-old female diagnosed with breast cancer, who is receiving a highly emetogenic chemotherapy regimen (cisplatin). She experiences mild anxiety but no history of vomiting. Her liver and kidney function are normal, and she has no history of drug allergies. The CINV early warning module predicts that Patient A is at high risk for CINV, with a high probability of occurrence and an expected onset of CINV 6-8 hours after chemotherapy. Using "cisplatin (highly emetogenic)," "breast cancer," "female," "55 years old," "mild anxiety," and "normal liver and kidney function" as starting points, knowledge graph reasoning reveals that ondansetron + dexamethasone is an effective antiemetic regimen for highly emetogenic chemotherapy. Excluding the patient's history of drug allergies and drug interactions, the optimal ondansetron dose calculated based on the patient's condition is 8 mg and dexamethasone 4 mg. Considering the patient's anxiety, a short-term low-dose lorazepam is added. Recommended regimen: Ondansetron 8 mg intravenously, 30 minutes before chemotherapy. Dexamethasone 4 mg intravenously, 30 minutes before chemotherapy. Ondansetron 8mg orally every 8 hours after chemotherapy for 3 days. Lorazepam 0.5mg orally, as needed, to relieve anxiety. Show the recommended regimen to the doctor, who will confirm and send it to the nursing mobile app.
[0098] In an optional embodiment, the automatic matching and recommendation of personalized antiemetic drug treatment plans by the knowledge graph reasoning algorithm further includes:
[0099] The nodes of the knowledge graph are entities including chemotherapy drugs, antiemetic drugs, and patient diseases, and the edges of the knowledge graph are relationships between entities;
[0100] According to the patient's CINV warning information, drug allergy history, and drug interaction information, the best antiemetic drug treatment plan is searched in the knowledge graph using the PathRanking algorithm.
[0101] In this embodiment, the knowledge graph processes complex interactions between drugs (such as synergistic effects and antagonistic effects), thereby avoiding irrational drug combinations. The reasoning process provides a basis for recommendation results through path display, which increases the medical staff's trust in the recommended plan.
[0102] For example, entities in the knowledge graph include: chemotherapy drugs such as cisplatin and doxorubicin; antiemetics such as ondansetron, dexamethasone, and aprepitant; patient diseases such as breast cancer and lung cancer; CINV symptoms such as nausea, vomiting, and diarrhea; adverse drug reactions such as constipation and headache; patient risk factors such as age, gender, previous history of vomiting, and anxiety; and gene mutations related to drug metabolism. Relationships include treatment, prevention, cause, interaction, contraindications, and dosage adjustment. A query target is constructed based on a patient's CINV warning information, drug allergy history, and drug interaction history. For example, the starting point is the patient (including all of their attributes: age, gender, disease, allergy history, and CINV risk level). The endpoint is the antiemetic drug. Constraints include avoiding allergic drugs and avoiding drug interactions. A path ranking algorithm is used to search the knowledge graph for all possible paths from the starting point to the endpoint, where each path represents a potential antiemetic treatment option. The Path Ranking algorithm ranks these paths based on their quality. Path quality evaluation metrics include: path length (shorter paths are generally more reliable); relationship type (e.g., prevention is more important than relief); and entity type (e.g., guideline-recommended medications are more important than over-the-counter medications). The algorithm returns the top N antiemetic treatment options as recommendations. For example, Path 1: Patient A -> "Currently receiving chemotherapy" -> cisplatin -> "Cause" -> CINV -> "Prevention" -> ondansetron. Path 2: Patient A -> "Currently receiving chemotherapy" -> cisplatin -> "Cause" -> CINV -> "Prevention" -> dexamethasone. Path 3: Patient A -> "Currently receiving chemotherapy" -> cisplatin -> "Cause" -> CINV -> "Prevention" -> aprepitant. Path 4: Patient A -> "Allergy" -> metoclopramide. The Path Ranking algorithm ranks paths based on their quality. Paths 1, 2, and 3 have higher quality because ondansetron, dexamethasone, and aprepitant are all antiemetics recommended by clinical guidelines. Path 4 has lower quality because metoclopramide is an allergy medication.
[0103] In an optional embodiment, the method for recommending a personalized antiemetic drug treatment regimen further comprises:
[0104] Three optimization objectives are defined, wherein the optimization objectives include minimizing the CINV risk score S(t), minimizing the incidence of adverse drug reactions P, and minimizing the treatment cost C total ;
[0105] The multi-objective problem is transformed into a single-objective optimization through the ε constraint method. The optimization formula is:
[0106] min(S(t)+max(P adverse -P max ,0)+max(C total -C max ,0))
[0107] Among them, P adverse is the actual observed or predicted incidence of adverse drug reactions; P max is the maximum acceptable threshold for the incidence of adverse drug reactions; C max is the maximum acceptable threshold for total treatment cost.
[0108] In this embodiment, by setting P max and C max The threshold value can be adjusted to accommodate the tolerance for adverse reactions and treatment costs, thus flexibly adapting to the needs of different patients. For example, for patients with better economic conditions, the C max , achieving better CINV control, even at a higher cost. Each term in the objective function represents a clear clinical consideration, making the final recommendation interpretable and helpful for physicians to understand and accept.
[0109] In an optional manner, the minimization of the CINV risk score S(t) satisfies the Bellman equation, which is:
[0110] S(t)=min u(t) L(x(t),u(t),t)+E[S(t+1)|x(t),u(t)]
[0111] where u(t) is the antiemetic treatment regimen taken at time t; x(t) is the patient's state at time t; L(x(t),u(t),t) is the immediate CINV risk resulting from treatment u(t) at time t when the patient's state is x(t); and E[S(t+1)|x(t),u(t)] is the expected optimal CINV risk score at time t+1, conditional on the state x(t) and treatment u(t) at time t.
[0112] In this embodiment, the Bellman equation can take into account the patient's state changes at different time points to dynamically evaluate and optimize the antiemetic drug treatment plan. In addition, it can also minimize the CINV risk score, thereby improving the patient's treatment effect and quality of life.
[0113] For example, Plan 1: Ondansetron alone. Predicted CINV risk score = 5, adverse reaction incidence = 0.05, and treatment cost = 5000 yuan. Plan 2: Aprepitant + ondansetron combination. Predicted CINV risk score = 3, adverse reaction incidence = 0.15, and treatment cost = 6000 yuan. Plan 3: Dexamethasone + ondansetron. Predicted CINV risk score = 4, adverse reaction incidence = 0.08, and treatment cost = 2000 yuan.
[0114] Based on the dynamic optimization results of the Bellman equation, regimen 2 (combination of aprepitant and ondansetron) was recommended because it performed best in minimizing the CINV risk score, incidence of adverse reactions, and treatment costs.
[0115] In an optional manner, the minimization of the incidence of adverse drug reactions P and the minimization of treatment costs C total The constraint is embedded into the single-objective optimization function through the Lagrangian relaxation method. The optimization formula of the single-objective optimization function is:
[0116] min u S(t)+λ1(P adverse (u)-P max ) + +λ2(C total (u)-C max ) +
[0117] Where u is the antiemetic drug treatment regimen; (.) + is a positive function; λ1 is the penalty factor when the adverse drug reaction violates the constraint; λ2 is the penalty factor when the treatment cost violates the constraint.
[0118] In this embodiment, the Lagrangian relaxation method allows complex constraints to be converted into penalty terms in the objective function, thereby simplifying the optimization process. If the optimized solution still violates the constraints, the corresponding penalty factor is increased. If the optimized solution is too conservative (e.g., the incidence of adverse reactions is much lower than P max ), then appropriately reduce the corresponding penalty factor. Iteratively adjust the penalty factor until a solution is found that can effectively reduce CINV risk while meeting adverse reaction and cost constraints.
[0119] In an optional manner, the constraint embedding process of the Lagrangian relaxation method further includes:
[0120] Construct a nonlinear constraint function for the incidence of adverse drug reactions and convert the continuous probability density function into discrete constraints through piecewise linearization.
[0121] The multi-objective optimization problem is transformed into a constrained stochastic differential equation, and the expected optimal solution of the nonlinear constraint function is calculated using the Ito formula.
[0122] In this example, the multi-objective optimization problem was transformed into a stochastic differential equation, accounting for the randomness and uncertainty in the CINV process, allowing the final antiemetic regimen to adapt to individual patient differences and environmental changes. Piecewise linearization simplifies the complex probability density function into multiple linear segments, and the stochastic differential equation is solved using Ito's formula, which, to a certain extent, avoids the difficulties of direct numerical solution.
[0123] Step S104, sending the antiemetic drug treatment plan to the nursing mobile terminal to remind the clinical nurse to give the antiemetic drug and record the administration time and route; and providing feedback on the patient's vomiting condition, nausea level and adverse reaction information.
[0124] In an optional manner, the method further includes:
[0125] Assigning time-sensitive weights to the historical CINV probability based on the patient's real-time physiological data;
[0126] A two-stage parameter optimization objective function is constructed, which consists of the first and second stages. In the first stage, a genetic algorithm is used to search for the minimum parameter combination (p, d, q) that satisfies the AIC criterion. In the second stage, the Bayesian Information Criterion is used to regularize the candidate model. The constraint condition of the Bayesian Information Criterion is:
[0127] BIC=-2lnL+ln(n)·(1+ω(t))
[0128] Where L is the model likelihood function; n is the sample size; ω(t) is the weight function; ω(t) = e -Δt ·[1+α·Var(x physio (t))]; Δt is the data time interval; α is the physiological parameter fluctuation adjustment factor; Var(x physio (t)) is the variance of the patient's real-time physiological parameters at time t; x physio (t) is the patient's real-time physiological parameter value at time t;
[0129] When it is detected that the change rate of chemotherapy drug infusion rate is greater than the preset threshold, the ARIMA model is forced to switch to the low-order configuration.
[0130] In this example, a two-stage parameterization approach was used, leveraging the global search capabilities of a genetic algorithm to identify potential optimal parameters. Regularization was then performed using the Bayesian Information Criterion, enabling the model to perform better with new patient data. When the chemotherapy drug infusion rate changes dramatically, a forced switch to a low-order ARIMA model is implemented, reducing the model's overreaction to unexpected events and ensuring its stability, enabling it to handle unexpected situations during chemotherapy (such as infusion rate adjustments).
[0131] According to the solution provided by the present invention, a comprehensive database containing the emetogenicity level of chemotherapy drugs, patient risk factors and antiemetic drug information is constructed, wherein the antiemetic drug information includes drug type, dosage form, dosage, route of administration, administration time, course of treatment, half-life, liver and kidney function requirements, adverse reactions and drug interactions, and the patient risk factors include age, gender, previous vomiting history and anxiety level; a patient physiological parameter baseline database containing the patient's blood routine, liver and kidney function, electrolytes and electrocardiogram indicators is established; the patient's CINV early warning information is dynamically predicted based on the comprehensive database, wherein the early warning information includes CINV risk level, CINV risk profile, and CINV risk level. rate, severity and expected time of occurrence; based on the CINV warning information, the comprehensive database, the patient physiological parameter baseline database, the patient's drug allergy history and drug interaction information, automatically match and recommend personalized antiemetic drug treatment plans through the knowledge graph reasoning algorithm; wherein, the antiemetic drug treatment plan includes the type of antiemetic drug, dosage form, dosage, route of administration, first administration time, subsequent administration frequency and course of treatment; the antiemetic drug treatment plan is sent to the nursing mobile terminal to remind clinical nurses to give antiemetic drugs and record the administration time and route; and the recorded patient vomiting, nausea level and adverse reaction information are fed back. The present invention matches appropriate antiemetic drugs and provides antiemetic treatment plans based on therapeutic principles and drug properties, and promptly reminds clinical nurses to give the optimal interval time of antiemetic drugs to ensure the best efficacy of antiemetic drugs. Specifically, by constructing a comprehensive database and a patient physiological parameter baseline database and combining the patient's specific situation, the patient's CINV warning information is dynamically predicted, and the knowledge graph reasoning algorithm is used to automatically match and recommend personalized antiemetic drug treatment plans, reducing unnecessary drug use and adverse reactions. The multi-objective problem is converted into a single-objective optimization through the ε constraint method, ensuring that the optimal treatment plan is found while meeting multiple goals. The antiemetic drug treatment plan is sent to the nursing mobile terminal to remind clinical nurses to give antiemetic drugs and record the time and route of administration. The best antiemetic drug treatment plan is searched in the knowledge graph through the Path Ranking algorithm, further improving the accuracy of the recommended plan. The historical CINV occurrence probability is assigned a time-sensitive weight based on the patient's real-time physiological data, and a two-stage parameter optimization objective function is constructed. The model is optimized using the Bayesian Information Criterion to ensure the adaptability of the prediction model. When the change rate of the chemotherapy drug infusion rate is detected to be greater than the preset threshold, the ARIMA model is forced to switch to a low-order configuration so that the model can remain accurate under dynamic changes.
[0132] Figure 2 The schematic diagram of the framework of the antiemetic medication management device for cancer chemotherapy patients according to an embodiment of the present invention is shown. The antiemetic medication management device for cancer chemotherapy patients includes:
[0133] Data construction module 210 is used to construct a comprehensive database containing the emetogenicity level of chemotherapy drugs, patient risk factors, and antiemetic drug information, wherein the antiemetic drug information includes drug type, dosage form, dosage, route of administration, administration time, treatment course, half-life, liver and kidney function requirements, adverse reactions and drug interactions; patient risk factors include age, gender, previous vomiting history, and anxiety level; and establish a patient physiological parameter baseline database including the patient's blood routine, liver and kidney function, electrolytes, and electrocardiogram indicators;
[0134] A CINV warning module 220 is used to dynamically predict the patient's CINV warning information based on the comprehensive database, wherein the warning information includes CINV risk level, CINV risk probability, severity and expected occurrence time;
[0135] A regimen recommendation module 230 is configured to automatically match and recommend a personalized antiemetic drug treatment regimen based on the CINV warning information, the comprehensive database, the patient physiological parameter baseline database, the patient's drug allergy history, and drug interaction information using a knowledge graph reasoning algorithm; wherein the antiemetic drug treatment regimen includes the type of antiemetic drug, dosage form, dosage, route of administration, time of first administration, frequency of subsequent administration, and course of treatment;
[0136] The execution and feedback module 240 is used to send the antiemetic drug treatment plan to the nursing mobile terminal to remind the clinical nurse to give the antiemetic drug and record the administration time and route; and to provide feedback on the recorded patient vomiting, nausea level and adverse reaction information.
[0137] Figure 3 The schematic diagram of the structure of the computing device embodiment of the present invention is shown. The specific embodiment of the present invention does not limit the specific implementation of the computing device.
[0138] like Figure 3 As shown, the computing device may include: a processor 302 , a communications interface 304 , a memory 306 , and a communication bus 308 .
[0139] Processor 302, communication interface 304, and memory 306 communicate with each other via communication bus 308. Communication interface 304 is used to communicate with other devices, such as client devices or other server network elements. Processor 302 is used to execute program 310, which may specifically perform the steps described in the aforementioned embodiment of the method for managing antiemetic medication for cancer chemotherapy patients.
[0140] Specifically, the program 310 may include program codes, which include computer operation instructions.
[0141] Processor 302 may be a central processing unit (CPU), an application-specific integrated circuit (ASIC), or one or more integrated circuits configured to implement embodiments of the present invention. The one or more processors included in a computing device may be processors of the same type, such as one or more CPUs, or processors of different types, such as one or more CPUs and one or more ASICs.
[0142] The memory 306 is used to store the program 310. The memory 306 may include a high-speed RAM memory, and may also include a non-volatile memory (non-volatile memory), such as at least one disk memory.
[0143] According to the solution provided by the present invention, a comprehensive database containing the emetogenicity level of chemotherapy drugs, patient risk factors and antiemetic drug information is constructed, wherein the antiemetic drug information includes drug type, dosage form, dosage, route of administration, administration time, course of treatment, half-life, liver and kidney function requirements, adverse reactions and drug interactions, and the patient risk factors include age, gender, previous vomiting history and anxiety level; a patient physiological parameter baseline database containing the patient's blood routine, liver and kidney function, electrolytes and electrocardiogram indicators is established; the patient's CINV early warning information is dynamically predicted based on the comprehensive database, wherein the early warning information includes CINV risk level, CINV risk profile, and CINV risk level. rate, severity and expected time of occurrence; based on the CINV warning information, the comprehensive database, the patient physiological parameter baseline database, the patient's drug allergy history and drug interaction information, automatically match and recommend personalized antiemetic drug treatment plans through the knowledge graph reasoning algorithm; wherein, the antiemetic drug treatment plan includes the type of antiemetic drug, dosage form, dosage, route of administration, first administration time, subsequent administration frequency and course of treatment; the antiemetic drug treatment plan is sent to the nursing mobile terminal to remind clinical nurses to give antiemetic drugs and record the administration time and route; and the recorded patient vomiting, nausea level and adverse reaction information are fed back. The present invention matches appropriate antiemetic drugs and provides antiemetic treatment plans based on therapeutic principles and drug properties, and promptly reminds clinical nurses to give the optimal interval time of antiemetic drugs to ensure the best efficacy of antiemetic drugs. Specifically, by constructing a comprehensive database and a patient physiological parameter baseline database and combining the patient's specific situation, the patient's CINV warning information is dynamically predicted, and the knowledge graph reasoning algorithm is used to automatically match and recommend personalized antiemetic drug treatment plans, reducing unnecessary drug use and adverse reactions. The multi-objective problem is converted into a single-objective optimization through the ε constraint method, ensuring that the optimal treatment plan is found while meeting multiple goals. The antiemetic drug treatment plan is sent to the nursing mobile terminal to remind clinical nurses to give antiemetic drugs and record the time and route of administration. The best antiemetic drug treatment plan is searched in the knowledge graph through the Path Ranking algorithm, further improving the accuracy of the recommended plan. The historical CINV occurrence probability is assigned a time-sensitive weight based on the patient's real-time physiological data, and a two-stage parameter optimization objective function is constructed. The model is optimized using the Bayesian Information Criterion to ensure the adaptability of the prediction model. When the change rate of the chemotherapy drug infusion rate is detected to be greater than the preset threshold, the ARIMA model is forced to switch to a low-order configuration so that the model can remain accurate under dynamic changes.
[0144] Those skilled in the art will appreciate that modules in the devices of the embodiments may be adaptively modified and deployed in one or more devices different from the embodiments. Modules, units, or components in the embodiments may be combined into a single module, unit, or component, and furthermore, they may be divided into multiple submodules, subunits, or subcomponents. All features disclosed in this specification (including the accompanying claims, abstract, and drawings), as well as all processes or units of any method or device disclosed therein, may be combined in any combination, except where at least some of such features and / or processes or units are mutually exclusive. Unless expressly stated otherwise, each feature disclosed in this specification (including the accompanying claims, abstract, and drawings) may be replaced by an alternative feature that provides the same, equivalent, or similar purpose. Furthermore, those skilled in the art will appreciate that although some embodiments herein include certain features included in other embodiments but not others, combinations of features from different embodiments are intended to be within the scope of the present invention and to form different embodiments. For example, in the claims below, any of the claimed embodiments may be used in any combination. The present invention may be implemented using hardware comprising a number of different elements and using a suitably programmed computer. In a unit claim that lists several means, several of these means may be embodied by the same hardware item. Unless otherwise specified, the steps in the above embodiments should not be understood as limiting the order of execution.
Claims
1. A method for managing antiemetic medication for patients undergoing chemotherapy for tumors, characterized in that: include: Construct a comprehensive database containing the emetogenicity level of chemotherapy drugs, patient risk factors, and antiemetic drug information. The antiemetic drug information includes drug type, dosage form, dosage, route of administration, administration time, treatment course, half-life, liver and kidney function requirements, adverse reactions, and drug interactions. The patient risk factors include age, gender, previous history of vomiting, and anxiety level. A baseline database of patient physiological parameters including blood routine, liver and kidney function, electrolytes, and electrocardiogram indicators is established. Dynamically predicting CINV warning information for the patient based on the comprehensive database, wherein the warning information includes CINV risk level, CINV risk probability, severity, and expected time of occurrence; Based on the CINV warning information, the comprehensive database, the patient physiological parameter baseline database, the patient's drug allergy history, and drug interaction information, a knowledge graph reasoning algorithm is used to automatically match and recommend a personalized antiemetic drug treatment plan; wherein the antiemetic drug treatment plan includes the type of antiemetic drug, dosage form, dosage, route of administration, time of first administration, frequency of subsequent administration, and course of treatment; The antiemetic drug treatment plan is sent to the nursing mobile terminal to remind clinical nurses to give antiemetic drugs and record the administration time and route; and the recorded patient vomiting, nausea level and adverse reaction information are fed back.
2. The antiemetic medication management method for tumor chemotherapy patients according to claim 1, characterized in that: The method of dynamically predicting the patient's CINV warning information based on the comprehensive database further includes: Establishing a time series database based on the patient's CINV incidence data, the emetogenicity level of chemotherapy drugs, patient risk factors, and the use of antiemetic drugs; The CINV occurrence data is decomposed into time series using an autoregressive integrated moving average model to predict the probability of CINV occurrence in the future; wherein the autoregressive integrated moving average model is ARIMA(p, d, q), where p is the order of the autoregressive term, d is the order of the difference term, and q is the order of the moving average term; The patient's age, gender, previous vomiting history, and anxiety level were used as input features, and a gradient boosting decision tree was used to train a GBDT classification model to predict the risk level of CINV. Comprehensive CINV warning information is constructed based on the CINV occurrence probability predicted by the ARIMA model and the CINV risk level predicted by the GBDT classification model.
3. The antiemetic medication management method for tumor chemotherapy patients according to claim 1, characterized in that: The automatic matching and recommendation of personalized antiemetic drug treatment plans through the knowledge graph reasoning algorithm further includes: The nodes of the knowledge graph are entities including chemotherapy drugs, antiemetic drugs, and patient diseases, and the edges of the knowledge graph are relationships between entities; According to the patient's CINV warning information, drug allergy history, and drug interaction information, the best antiemetic drug treatment plan is searched in the knowledge graph using a PathRanking algorithm.
4. The antiemetic medication management method for tumor chemotherapy patients according to claim 1, characterized in that: The method for recommending a personalized antiemetic drug treatment regimen further includes: Three optimization objectives are defined, wherein the optimization objectives include minimizing the CINV risk score S(t), minimizing the incidence of adverse drug reactions P, and minimizing the treatment cost C total ; The multi-objective problem is transformed into a single-objective optimization through the ε constraint method. The optimization formula is: min(S(t)+max(P adverse -P max ,0)+max(C total -C max ,0)) Among them, P adverse is the actual observed or predicted incidence of adverse drug reactions; P max is the maximum acceptable threshold for the incidence of adverse drug reactions; C max is the maximum acceptable threshold for total treatment cost.
5. The antiemetic medication management method for tumor chemotherapy patients according to claim 4, characterized in that: The minimized CINV risk score S(t) satisfies the Bellman equation, which is: S(t)=min u(t) L(x(t),u(t),t)+E[S(t+1)|x(t),u(t)] where u(t) is the antiemetic treatment regimen taken at time t; x(t) is the patient's state at time t; L(x(t),u(t),t) is the immediate CINV risk resulting from treatment u(t) at time t when the patient's state is x(t); and E[S(t+1)|x(t),u(t)] is the expected optimal CINV risk score at time t+1, conditional on the state x(t) and treatment u(t) at time t.
6. The antiemetic medication management method for tumor chemotherapy patients according to claim 4, characterized in that: The minimized incidence of adverse drug reactions P and the minimized treatment costs C total The constraint is embedded into the single-objective optimization function through the Lagrangian relaxation method. The optimization formula of the single-objective optimization function is: min u S(t)+λ1(P adverse (u)-P max ) + +λ2(C total (u)-C max ) + Where u is the antiemetic drug treatment regimen; (.) + is a positive function; λ1 is the penalty factor when the adverse drug reaction violates the constraint; λ2 is the penalty factor when the treatment cost violates the constraint.
7. The antiemetic medication management method for tumor chemotherapy patients according to claim 1, characterized in that: The method further comprises: Assigning time-sensitive weights to the historical CINV probability based on the patient's real-time physiological data; A two-stage parameter optimization objective function is constructed, which consists of the first and second stages. In the first stage, a genetic algorithm is used to search for the minimum parameter combination (p, d, q) that satisfies the AIC criterion. In the second stage, the Bayesian Information Criterion is used to regularize the candidate model. The constraint condition of the Bayesian Information Criterion is: BIC=-2lnL+ln(n)·(1+ω(t)) Where L is the model likelihood function; n is the sample size; ω(t) is the weight function; ω(t) = e -Δt ·[1+α·Var(x physio (t))]; Δt is the data time interval; α is the physiological parameter fluctuation adjustment factor; Var(x physio (t)) is the variance of the patient's real-time physiological parameters at time t; x physio (t) is the patient's real-time physiological parameter value at time t; When it is detected that the change rate of chemotherapy drug infusion rate is greater than the preset threshold, the ARIMA model is forced to switch to the low-order configuration.
8. The antiemetic medication management method for tumor chemotherapy patients according to claim 6, characterized in that: The constraint embedding process of the Lagrangian relaxation method further includes: Construct a nonlinear constraint function for the incidence of adverse drug reactions and convert the continuous probability density function into discrete constraints through piecewise linearization. The multi-objective optimization problem is transformed into a constrained stochastic differential equation, and the expected optimal solution of the nonlinear constraint function is calculated using the Ito formula.
9. A device for managing antiemetic medication for patients undergoing chemotherapy for tumors, characterized in that: include: A data construction module is used to construct a comprehensive database containing the emetogenicity level of chemotherapy drugs, patient risk factors, and antiemetic drug information, wherein the antiemetic drug information includes drug type, dosage form, dosage, route of administration, administration time, treatment course, half-life, liver and kidney function requirements, adverse reactions and drug interactions. The patient risk factors include age, gender, previous vomiting history, and anxiety level; and establish a patient physiological parameter baseline database including the patient's blood routine, liver and kidney function, electrolytes, and electrocardiogram indicators; A CINV early warning module, configured to dynamically predict CINV early warning information of a patient based on the comprehensive database, wherein the early warning information includes CINV risk level, CINV risk probability, severity, and expected occurrence time; a regimen recommendation module for automatically matching and recommending personalized antiemetic drug treatment regimens based on the CINV warning information, the comprehensive database, the patient physiological parameter baseline database, the patient's drug allergy history, and drug interaction information through a knowledge graph reasoning algorithm; wherein the antiemetic drug treatment regimen includes the type of antiemetic drug, dosage form, dosage, route of administration, time of first administration, frequency of subsequent administration, and course of treatment; The execution and feedback module is used to send the antiemetic drug treatment plan to the nursing mobile terminal to remind clinical nurses to give antiemetic drugs and record the administration time and route; and to provide feedback on the recorded patient vomiting, nausea level and adverse reaction information.
10. A computing device comprising: A processor, a memory, a communication interface, and a communication bus, wherein the processor, the memory, and the communication interface communicate with each other via the communication bus; The memory is used to store at least one executable instruction, and the executable instruction enables the processor to execute operations corresponding to the above-mentioned method for managing antiemetic medication for patients undergoing tumor chemotherapy.