Tumor patient aromatherapy dose decision modeling method based on evidence-based data

Abstract

The invention discloses a tumor patient aromatherapy dose decision modeling method based on evidence-based data, and relates to the technical field of medicine and data science, and the method comprises the following steps: obtaining clinical data of tumor patients from multiple sources, the data including but not limited to the age, gender and gene characteristics of the patients; according to the method, the heterogeneity problem between clinical research and patient groups is solved through data standardization processing, the accuracy of a therapeutic dose decision model is improved, and it is ensured that patients receive the most appropriate aromatherapy dose. Through a personalized dose decision model, a customized treatment scheme is generated according to patient characteristics, the treatment effect is optimized, and side effects are reduced. In addition, by monitoring the physiological data of the patient in real time and dynamically adjusting the treatment dosage, continuous optimization of the treatment scheme is ensured, and the treatment safety and effect are improved to the maximum extent.

Description

Technical Field

[0001] The present invention relates to the field of medicine and data science technology, and in particular to a modeling method for aromatherapy dosage decision-making for tumor patients based on evidence-based data. Background Art

[0002] "Aromatherapy Dosage Decision Modeling for Cancer Patients Based on Evidence-Based Data" refers to the use of evidence-based medical data, through data analysis and modeling methods, to provide personalized dosage decision plans for aromatherapy treatments for cancer patients. Evidence-based medical data includes empirical data such as clinical trials, patient treatment responses, and disease development trends, which can provide a scientific basis for treatment decisions. This modeling method analyzes the historical treatment data of a large number of patients to construct a mathematical model to predict the optimal dosage of aromatherapy under different conditions, thereby tailoring the most appropriate treatment plan for each patient. This method aims to improve treatment efficacy, reduce side effects, and optimize the individualization of treatment plans, thereby improving patients' quality of life and treatment outcomes.

[0003] The existing technology has the following deficiencies: In the existing technology, in the process of modeling aromatherapy dosage decisions for cancer patients based on evidence-based data, data heterogeneity may lead to inaccurate dosage decisions. Due to differences in conditions across different clinical studies and patient groups (such as age, gender, genetic characteristics, etc.), data from different sources may have significant differences. This data heterogeneity will affect the model's accurate prediction of the true response of cancer patients under specific aromatherapy. When this problem is not identified and corrected in a timely manner, it may lead to patients receiving inappropriate treatment doses, which may affect the treatment effect and may also cause serious side effects, thereby negatively affecting the patient's health and treatment effect, and even delaying the optimal treatment time.

[0004] The above information disclosed in this Background section is only for enhancement of understanding of the background of the present disclosure and therefore it may contain information that does not form the prior art that is already known to a person of ordinary skill in the art. Summary of the Invention

[0005] The purpose of this invention is to provide a modeling method for aromatherapy dosage decision-making for cancer patients based on evidence-based data. This method, through data standardization, addresses heterogeneity across clinical studies and patient populations, improves the accuracy of the treatment dosage decision-making model, and ensures that patients receive the most appropriate aromatherapy dosage. Through a personalized dosage decision-making model, a customized treatment plan is generated based on patient characteristics, optimizing treatment efficacy and minimizing side effects. Furthermore, by monitoring patient physiological data in real time and dynamically adjusting treatment dosage, this method ensures continuous optimization of the treatment plan, maximizing treatment safety and effectiveness, and thus addressing the aforementioned issues in the background art.

[0006] To achieve the above objectives, the present invention provides the following technical solution: a method for modeling aromatherapy dosage decisions for cancer patients based on evidence-based data, comprising the following steps:

[0007] Obtain clinical data of cancer patients from multiple sources, including but not limited to the patients' age, gender, and genetic characteristics;

[0008] Standardize clinical data to eliminate the impact of data heterogeneity on model training and ensure that data from different sources are compared and analyzed under the same standards;

[0009] Build a data analysis model to identify key factors related to aromatherapy dosage by extracting features from standardized data;

[0010] Using data analysis models for training to obtain the optimal dosage prediction model, the model predicts the patient's response to aromatherapy based on their specific characteristics;

[0011] Generate personalized aromatherapy dosage decisions based on the trained dosage prediction model to ensure that each patient's treatment dosage matches their individual characteristics;

[0012] Based on the generated therapeutic dose, the dosage decision is continuously optimized and adjusted by real-time monitoring of the patient's treatment response to address changes in the patient's response.

[0013] Preferably, the specific steps of the standardization process are as follows:

[0014] Preprocess clinical data from different sources to remove outliers and noise data to ensure data quality;

[0015] Convert characteristics of different patient groups into uniform standardized units to eliminate errors caused by different measurement units or dimensions;

[0016] A normalization algorithm was used to process the patient's age, gender, and genetic characteristics so that they were within the same range, and each characteristic was weighted according to its importance;

[0017] Statistical analysis methods are used to detect the correlation between various features, and the data standardization strategy is further adjusted based on the correlation analysis results.

[0018] Preferably, the specific steps for establishing the data analysis model are as follows:

[0019] Collect processed and standardized data and input it into the machine learning algorithm;

[0020] Divide the training data, use cross-validation method to evaluate model performance, and select the optimal model structure;

[0021] The model parameters were adjusted through an optimization algorithm, and the objective function of error minimization was adopted to ensure the model's prediction accuracy for aromatherapy dosage decisions for cancer patients.

[0022] Evaluation metrics were used to validate the model's performance and ensure that the model could adapt to the data characteristics of different patient groups.

[0023] Preferably, the steps for generating a personalized aromatherapy dosage decision are as follows:

[0024] According to the trained dose prediction model, the clinical characteristics and standardized data of the tumor patients are input;

[0025] Based on the predicted values output by the model, a personalized treatment dose is generated for each patient, and the treatment dose is fine-tuned according to the patient's condition and treatment history;

[0026] Validate dose predictions through clinical expert feedback to ensure the dose maximizes therapeutic efficacy and minimizes side effects;

[0027] According to the patient's response during treatment, the treatment dose is adjusted in real time to optimize the drug effect during treatment.

[0028] Preferably, the real-time monitoring steps are as follows:

[0029] During treatment, clinical monitoring equipment is used to collect patients’ physiological data in real time;

[0030] Correlate patient physiological data with treatment dose and response to treatment to assess patient response to aromatherapy;

[0031] Based on real-time monitoring results, the most suitable aromatherapy dosage for the patient's current condition is calculated, and the patient's treatment plan is adjusted through feedback.

[0032] Preferably, the specific steps of the data analysis model training and optimization process are as follows:

[0033] Input standardized dataset X, X={x i}={x1,x2,……,x n}, where x i represents the clinical characteristics of the i-th patient, and n is the total number of patients;

[0034] The objective function is modeled for each patient's response value, and the goal is to minimize the error function of the model, which is as follows:

[0035]

[0036] , where L(θ) is the loss function, θ is the model parameter, and y i is the response value of the ith patient, is the predicted response value for the i-th patient;

[0037] Use the optimization algorithm to iteratively update the parameter θ until the loss function L(θ) converges to the minimum value;

[0038] In order to prevent overfitting, a regularization term λ‖θ‖ is added 2 , the optimization objectives are as follows:

[0039]

[0040] , where L reg (θ) is the regularized error function, λ is the regularization coefficient, ‖θ‖ 2 is the sum of the squares of the parameters.

[0041] Preferably, the therapeutic dose is adjusted as follows:

[0042] Adjust the dosage of aromatherapy according to the patient's side effects or response to treatment during treatment;

[0043] Combine patient feedback data and physiological changes to calculate and predict the next treatment dose to prevent overdose or underdose;

[0044] By setting critical values, the range of dosage adjustment is determined to ensure the safety and effectiveness of the treatment process.

[0045] Preferably, the real-time optimization process of the treatment dose decision model is as follows:

[0046] Input the patient's physiological data S, S={s j}={s1,s2,……,s m}, where s j represents the jth real-time physiological parameter of the patient, and m is the number of physiological parameter types;

[0047] For each group of patients' physiological data S j , calculate the predicted value of the therapeutic dose, the calculation expression is as follows:

[0048] D j =α1s1+α2s2+…+α m s m +β

[0049] , where α1, α2, α m They are s1, s2, and s m The weight coefficient reflects the influence of each physiological data item on the predicted treatment dose, and β is the bias term;

[0050] Optimize the prediction model by adjusting the weight coefficients to make it more adaptable to changes in real-time data and use optimization methods to minimize errors;

[0051] Based on the optimized weight coefficients, the patient's aromatherapy dosage is adjusted in real time to ensure that each patient's treatment plan matches their current physiological state, thereby maximizing the treatment effect and reducing the risk of side effects.

[0052] In the above technical solution, the technical effects and advantages provided by the present invention are:

[0053] This invention eliminates data heterogeneity from different clinical studies and patient populations through data normalization, ensuring consistency across different patient populations in the treatment dose decision model. This data normalization step effectively addresses data inconsistencies caused by differences in factors such as age, gender, and genetic characteristics, enabling the model to make predictions based on unified, standardized data. This optimization significantly improves the accuracy of treatment doses and reduces errors caused by data discrepancies, ensuring that patients receive the most appropriate aromatherapy dose and avoiding the negative impact of incorrect treatment doses on patient health.

[0054] This invention constructs a personalized dosage decision model to generate the most appropriate aromatherapy dosage based on each patient's characteristics, condition, and treatment response. This method incorporates clinical characteristics, such as age, gender, and genetic profile, to provide a customized treatment plan for each patient. Personalized dosage decisions not only improve treatment efficacy but also effectively minimize side effects and lower patient risks. This precise treatment plan allows patients to achieve better outcomes and improve their quality of life.

[0055] The present invention ensures continuous optimization of the treatment plan by monitoring the patient's physiological state in real time and dynamically adjusting the treatment dose based on the monitoring data. During the treatment process, the patient's physiological and therapeutic responses can be fed back into the model in real time, allowing the dosage to be adjusted at any time based on the patient's physical condition. This dynamic adjustment mechanism can promptly identify and correct side effects or adverse reactions that occur during treatment, thereby maximizing the safety and effectiveness of the treatment. Through this flexible adjustment method, the present invention significantly improves the individualization of treatment, allowing each patient to receive treatment at the most appropriate dose and reducing unnecessary risks. BRIEF DESCRIPTION OF THE DRAWINGS

[0056] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, a brief introduction to the drawings required for use in the embodiments will be given below. Obviously, the drawings described below are only some embodiments recorded in the present invention. For ordinary technicians in this field, other drawings can also be obtained based on these drawings.

[0057] Figure 1This is a flow chart of the method for modeling aromatherapy dosage decision-making for cancer patients based on evidence-based data of the present invention. DETAILED DESCRIPTION

[0058] Example embodiments will now be described more fully with reference to the accompanying drawings. However, example embodiments can be implemented in many forms and should not be construed as limited to the examples set forth herein; rather, these example embodiments are provided so that the description of this disclosure will be thorough and complete and will fully convey the concepts of the example embodiments to those skilled in the art.

[0059] The present invention provides Figure 1 The modeling approach for aromatherapy dosage decision-making in cancer patients based on evidence-based data includes the following steps:

[0060] Obtain clinical data of cancer patients from multiple sources, including but not limited to the patients' age, gender, and genetic characteristics;

[0061] Standardize clinical data to eliminate the impact of data heterogeneity on model training and ensure that data from different sources are compared and analyzed under the same standards;

[0062] The specific steps for standardization are as follows:

[0063] Preprocess clinical data from different sources to remove outliers and noise data to ensure data quality;

[0064] Convert characteristics of different patient groups into uniform standardized units to eliminate errors caused by different measurement units or dimensions;

[0065] A normalization algorithm was used to process the patient's age, gender, and genetic characteristics so that they were within the same range, and each characteristic was weighted according to its importance;

[0066] Statistical analysis methods are used to detect the correlation between various features, and the data standardization strategy is further adjusted based on the correlation analysis results.

[0067] Build a data analysis model to identify key factors related to aromatherapy dosage by extracting features from standardized data;

[0068] The specific steps for establishing a data analysis model are as follows:

[0069] Collecting processed and standardized data and inputting it into a machine learning algorithm, including but not limited to a decision tree, random forest, support vector machine (SVM), or deep learning network;

[0070] Divide the training data, use cross-validation method to evaluate model performance, and select the optimal model structure;

[0071] The model parameters were adjusted through an optimization algorithm, and the objective function of error minimization was adopted to ensure the model's prediction accuracy for aromatherapy dosage decisions for cancer patients.

[0072] Evaluation indicators such as accuracy, recall, and F1 value were used to verify the performance of the model and ensure that the model can adapt to the data characteristics of different patient groups.

[0073] Using data analysis models for training to obtain the optimal dosage prediction model, the model predicts the patient's response to aromatherapy based on their specific characteristics;

[0074] The steps for generating personalized aromatherapy dosage decisions are as follows:

[0075] According to the trained dose prediction model, the clinical characteristics and standardized data of the tumor patients are input;

[0076] Based on the predicted values output by the model, a personalized treatment dose is generated for each patient, and the treatment dose is fine-tuned according to the patient's condition and treatment history;

[0077] Validate dose predictions through clinical expert feedback to ensure the dose maximizes therapeutic efficacy and minimizes side effects;

[0078] According to the patient's response during treatment, the treatment dose is adjusted in real time to optimize the drug effect during treatment.

[0079] Generate personalized aromatherapy dosage decisions based on the trained dosage prediction model to ensure that each patient's treatment dosage matches their individual characteristics;

[0080] The real-time monitoring steps are as follows:

[0081] During treatment, clinical monitoring equipment is used to collect patients’ physiological data in real time, such as body temperature, blood pressure, heart rate, etc.

[0082] Correlate patient physiological data with treatment dose and response to treatment to assess patient response to aromatherapy;

[0083] Based on real-time monitoring results, the most suitable aromatherapy dosage for the patient's current condition is calculated, and the patient's treatment plan is adjusted through feedback.

[0084] The specific steps of the data analysis model training and optimization process are as follows:

[0085] Input standardized dataset X, X={x i}={x1,x2,……,x n}, where x i represents the clinical characteristics of the i-th patient, and n is the total number of patients;

[0086] The objective function modeling is performed for each patient's response value (the efficacy index of aromatherapy). The goal is to minimize the error function of the model. The formula is as follows:

[0087]

[0088] , where L(θ) is the loss function, which is used to measure the difference between the model prediction results and the actual results, θ is the model parameter, and y i is the response value of the i-th patient, usually an indicator of the efficacy of aromatherapy, is the predicted response value for the i-th patient;

[0089] Use an optimization algorithm (such as gradient descent) to iteratively update the parameter θ until the loss function L(θ) converges to a minimum;

[0090] In order to prevent overfitting, a regularization term λ‖θ‖ is added 2 , the optimization objectives are as follows:

[0091]

[0092] , where L reg (θ) is the regularized error function. The purpose of adding the regularization term is to make the optimization process not only pursue minimizing the training error, but also to avoid the model being too complex, which leads to overfitting of the training data. λ is the regularization coefficient (regularization parameter), which controls the weight of the regularization term. 2 It is the sum of squares of the parameters (also called L2 norm), which measures the size of the model parameters and is used to limit the situation where the parameters are too large.

[0093] This regularization term is used to penalize excessively large or unreasonable parameter values, thereby improving the generalization ability of the model.

[0094] Based on the generated therapeutic dose, the dosage decision is continuously optimized and adjusted by real-time monitoring of the patient's treatment response to address changes in the patient's response.

[0095] The steps for adjusting the therapeutic dose are as follows:

[0096] Adjust the dosage of aromatherapy according to the patient's side effects or response to treatment during treatment;

[0097] Combine patient feedback data and physiological changes to calculate and predict the next treatment dose to prevent overdose or underdose;

[0098] By setting critical values, the range of dosage adjustment is determined to ensure the safety and effectiveness of the treatment process.

[0099] The real-time optimization process of the treatment dose decision model is as follows:

[0100] Input the patient's physiological data S, S={s j}={s1,s2,……,s m}, where s j represents the jth real-time physiological parameter of the patient (such as heart rate, body temperature, etc.), and m is the number of physiological parameter types;

[0101] For each group of patients' physiological data S j , calculate the predicted value of the therapeutic dose, the calculation expression is as follows:

[0102] D j =α1s1+α2s2+…+α m s m +β

[0103] , where α1, α2, α m They are s1, s2, and s m The weight coefficient reflects the influence of each physiological data item on the predicted treatment dose. β is the bias term, which represents the basic value of the treatment dose when all physiological data items s are input to zero.

[0104] Optimize the prediction model by adjusting the weight coefficients to make it more adaptable to changes in real-time data, and use optimization methods (such as the least squares method) to minimize errors;

[0105] Based on the optimized weight coefficients, the patient's aromatherapy dosage is adjusted in real time to ensure that each patient's treatment plan matches their current physiological state, thereby maximizing the treatment effect and reducing the risk of side effects.

[0106] Implementation Method 1: In modern medical data analysis, data quality is often compromised by the diversity of data sources and patient populations. This is particularly true in evidence-based aromatherapy dosage decision-making modeling for cancer patients. Data from different clinical studies, patient populations, and even hospitals and laboratories can vary significantly, impacting modeling accuracy. To address this issue, this implementation method proposes a comprehensive data standardization and cleaning optimization process. This process utilizes a series of data preprocessing and processing techniques to ensure data consistency and reliability, providing a high-quality data foundation for subsequent modeling.

[0107] First, data cleaning is a core step in ensuring data quality. Clinical data typically comes from a variety of sources, including hospitals in different regions, clinical trials, and patients' personal records. Due to the differences in these data sources, various problems may exist, such as missing values, outliers, and duplicate records. To this end, we use automated outlier detection and data cleaning algorithms to gradually eliminate these invalid or erroneous data. Outlier detection is first performed using a rule-based approach. We set some rules, such as a patient's age should not exceed 100 years old, and genetic characteristics should be within a certain known reasonable range. Data outside this range will be marked as outliers. Detected outliers can be processed in a variety of ways, such as deletion, correction, or interpolation, depending on the specific situation. In addition, duplicate records are also a common problem. Data deduplication is performed using unique identifiers (such as patient IDs) to ensure that each patient's information appears only once, avoiding data redundancy.

[0108] Secondly, after data cleaning, we perform missing value processing. There are generally two approaches to handling missing values: deleting missing data or filling in missing data. Deleting missing data is suitable for cases with relatively few missing values, preventing large-scale data loss. For cases with more missing values, we use interpolation to fill in missing data. Common interpolation methods include mean interpolation, nearest neighbor interpolation, and regression interpolation. These methods fill in missing values to ensure that every piece of data is complete and avoid model inaccuracies caused by missing data.

[0109] Once data cleaning and missing value handling are complete, the next step is data standardization and normalization. Because clinical data from cancer patients often come from different hospitals or laboratories, the units and scales of each dataset may vary significantly. For example, a patient's weight may be expressed in kilograms in one dataset but in pounds in another, or different hospitals may use different genetic testing methods, resulting in different data distributions. To eliminate these differences, we standardize and normalize the data.

[0110] Standardization refers to converting data of different dimensions into a unified standard scale. The commonly used method is z-score standardization. Specifically, for each feature value, subtract the mean of the feature and divide it by the standard deviation of the feature. The data processed in this way has zero mean and unit variance, so that the numerical differences between different features can be eliminated. For example, after various clinical characteristics such as age, blood pressure, and weight are standardized, all features are in the same dimension, eliminating the deviation introduced by different units of different features. Standardization processing can not only eliminate the influence of data dimension, but also improve the computational efficiency and accuracy of the subsequent modeling process. Especially when using distance-based algorithms (such as KNN, SVM, etc.), standardization is an indispensable step.

[0111] Normalization is the process of scaling data to a uniform range, usually mapping the data to the interval [0,1]. Common normalization methods include min-max normalization, which is to obtain the normalized value by subtracting the minimum value from each feature and dividing it by the difference between the maximum and minimum values of the feature. The main function of normalization is to prevent certain features from having an unreasonable impact on model training due to excessively large or small values. For example, when modeling aromatherapy dosage decisions, gene expression levels and patient weight, age and other features may have large numerical differences. Directly inputting them into the model will cause certain features to dominate the calculation, while other important features will be ignored. Through normalization, all features are converted to the same range, avoiding this problem.

[0112] After standardization and normalization, the data enters the next stage - feature engineering. Feature engineering is to extract features from the raw data that are meaningful for predicting the results. Feature selection and feature extraction are the key to this step. Feature selection aims to select the features that are most relevant to the target variable (in this case, aromatherapy dose response), remove redundant features, reduce the computational burden, and improve the efficiency and accuracy of the model. Commonly used feature selection methods include univariate selection methods based on statistical tests, model-based selection methods (such as Lasso regression), and tree-based models (such as random forests).

[0113] Finally, we also need to consider the issue of data balance. For the dose decision modeling of aromatherapy for cancer patients, there may be large differences in the amount of data for different categories of patients (for example, there are fewer patients with certain types of cancer). In this case, the model may tend to predict the category with larger data volumes. To address this problem, we use data resampling techniques, such as undersampling and oversampling, to balance the amount of data for each category. In this way, we ensure that the data for each category of patients occupies a reasonable proportion in the training, so that the model can better adapt to different categories of patient groups and make more accurate predictions.

[0114] In summary, this implementation eliminates errors caused by data heterogeneity through systematic standardization, normalization, data cleaning, and missing value processing, ensuring data consistency, accuracy, and completeness. This processed data will serve as the foundation for subsequent modeling, providing high-quality input for aromatherapy dosage decisions for cancer patients and improving the model's accuracy and reliability. This data optimization process minimizes errors caused by data variability and provides solid data support for precision medicine.

[0115] Implementation Method 2: In this implementation, a personalized dosage decision model is trained using a machine learning algorithm. The goal is to provide each cancer patient with a customized aromatherapy dosage decision based on their clinical characteristics and treatment response. This process is based on the patient's historical treatment data and uses a variety of machine learning techniques, such as deep learning, decision trees, and support vector machines (SVMs), to analyze the relationship between various patient characteristics and treatment outcomes, thereby accurately predicting different patients' responses to specific aromatherapy treatments. The trained model can generate a personalized treatment dosage for each patient, ensuring the accuracy and effectiveness of treatment.

[0116] The first step in the personalized dosing decision model is to input standardized data. This data includes basic information and clinical characteristics of the cancer patient (such as age, gender, and weight), as well as the patient's medical history, genetic information, and lifestyle. This data has been standardized and cleaned and optimized in the previous stage to ensure data quality and consistency. To enable the model to better understand and process this multi-dimensional data, feature extraction is first required. Feature extraction is the process of extracting important information from the raw data to construct a feature set with high predictive power.

[0117] For example, genetic characteristics may significantly influence the effectiveness of aromatherapy, so various gene expression data (such as specific gene mutations and gene expression levels) are input into the model as important features. Information such as the patient's age and gender serves as basic clinical characteristics and is included in model training. The goal of feature extraction is to transform this clinical data into highly effective features that represent individual patient differences, thereby improving the model's predictive capabilities.

[0118] After feature extraction, selecting an appropriate machine learning algorithm for training is crucial. Considering the complexity of the data and individual patient variability, this implementation utilizes several commonly used machine learning algorithms, including deep learning, decision trees, and support vector machines (SVMs). These algorithms have demonstrated excellent results in processing medical data, predicting disease responses, and developing treatment plans.

[0119] Deep Learning: Deep learning methods, particularly neural networks, excel at automatically extracting features from large amounts of data and capturing complex nonlinear relationships. During training, deep neural networks learn the complex relationships between input features and treatment effects, thereby generating personalized treatment doses for each patient. Deep learning models have strong generalization capabilities and are capable of processing data from diverse patient populations.

[0120] Decision Tree: Decision tree algorithms can break down treatment decision-making problems into a series of decision nodes, each representing a characteristic judgment. This series of characteristic judgments ultimately leads to the treatment dose. The advantage of decision tree models is their ease of interpretation, helping clinicians understand the logic and reasoning behind treatment decisions.

[0121] Support Vector Machine (SVM): A support vector machine is a classification model based on maximizing the margin, suitable for high-dimensional data classification problems. In this scenario, the SVM can be used to distinguish different patients' responses to aromatherapy, helping the model find the optimal treatment dose across different patient groups.

[0122] These algorithms were chosen because they can effectively process high-dimensional data and capture the underlying complex patterns in the data. Each algorithm has its own strengths and weaknesses, so in practical applications, multiple models can be combined or integrated for learning, depending on the characteristics of the data, to improve the accuracy and robustness of the overall model.

[0123] Model training is a crucial step in machine learning. To improve model accuracy, we employ cross-validation and parameter tuning to optimize model performance. Cross-validation reduces overfitting and ensures generalization by splitting the dataset into multiple subsets and repeatedly training and testing the model. Specifically, the training data is divided into a training set and a validation set. The training set is used to train the model, while the validation set is used to test the model's performance. This ensures that the model's performance on new data is consistent with its performance on the training set.

[0124] During model training, adjusting parameters is key to improving model performance. Different machine learning algorithms have different hyperparameters, such as the learning rate and number of hidden layers in deep neural networks, and the maximum depth and minimum number of sample splits in decision trees. Using methods such as grid search or random search, we traverse various possible hyperparameter combinations and select the best-performing configuration. Through these adjustments, the model can better adapt to different types of patient data, ensuring the accuracy and reliability of treatment dosing decisions.

[0125] To enable the model to adapt to changing patient data and treatment circumstances, this implementation also proposes a dynamic adjustment mechanism. As treatment progresses, the patient's physical condition may change, such as weight changes, or whether their condition improves or worsens. These changes will affect the patient's response to aromatherapy, necessitating real-time adjustments to the treatment dosage.

[0126] The dynamic adjustment process relies on the relationship between the features captured during model training and the patient's response. Whenever a patient's treatment response or physiological state changes, the new data is fed into the trained model, which then adjusts the weights and biases to recalculate the optimal treatment dose for the current state. This allows treatment decisions to be optimized based on the patient's real-time condition, ensuring maximum treatment effectiveness.

[0127] Finally, to ensure the model's reliability in real-world applications, we rigorously evaluated and continuously optimized the model. This evaluation primarily used a series of metrics, including accuracy, recall, and F1 score. These metrics allowed us to comprehensively measure the model's performance, particularly in treating different patient types.

[0128] Continuous optimization means that during practical application, the model continuously learns from new data, gradually refining treatment dosing decisions. For example, as more patient data accumulates, the model is continuously updated based on this new data, correcting any deviations in its predictions. Whenever new patient data or treatment responses emerge, the model can be fine-tuned to better adapt to the new environment, thereby improving treatment accuracy and effectiveness.

[0129] In summary, this embodiment uses a machine learning algorithm to train a personalized dosage decision model, which can accurately predict a patient's response to aromatherapy based on individual differences and generate the most appropriate treatment dose. The combination of multiple algorithms, including deep learning, decision trees, and support vector machines, ensures the model's efficiency and accuracy. Furthermore, the use of cross-validation, parameter optimization, and dynamic adjustment techniques further enhances the model's performance and adaptability, ensuring personalized and real-time optimization of treatment plans. This embodiment provides cancer patients with a scientific and precise aromatherapy dosage decision-making tool, significantly improving treatment efficacy and reducing the occurrence of side effects.

[0130] Implementation method three: In the field of modern medicine, dynamic adjustment during the treatment process has become an important means to improve efficacy and reduce side effects. Especially in aromatherapy for cancer patients, the patient's physiological state and treatment response will change as the treatment progresses. Therefore, real-time monitoring and dynamic adjustment of treatment doses are key to ensuring treatment effectiveness and patient safety. This implementation method utilizes advanced real-time monitoring technology, combined with existing treatment models, to collect patients' physiological data in real time and dynamically adjust the aromatherapy treatment dose based on this data. In this way, it can not only ensure that patients always receive the most appropriate treatment dose, but also effectively deal with possible side effects and adverse reactions, maximize treatment effects, and protect patients' health.

[0131] In this embodiment, real-time monitoring of the patient's physiological state is fundamental. By installing various physiological data monitoring devices (such as heart rate monitors, blood pressure monitors, and thermometers), various physiological parameters of the patient during treatment are collected in real time. These physiological parameters are crucial for determining the patient's treatment response. For example, changes in the patient's body temperature may be related to the effects and side effects of aromatherapy, while fluctuations in heart rate and blood pressure may reflect the patient's physiological state and tolerance to the treatment.

[0132] In addition to traditional physiological data monitoring, advanced sensing technologies can also be combined to collect other relevant patient data in real time, such as oxygen saturation and blood sugar levels. Through this multi-dimensional physiological data, a comprehensive understanding of the patient's health status can be achieved, providing a rich basis for subsequent dosage decisions.

[0133] Furthermore, with the increasing prevalence of smart wearable devices, patients can also use smartwatches, smart wristbands, and other devices to continuously monitor their health status in their daily lives, transmitting this data in real time to hospitals or medical systems. This approach not only improves the real-time and accuracy of data but also reduces the burden on patients, making it easier for them to receive treatment.

[0134] Collected physiological data is transmitted via wireless networks to the hospital's central database or cloud platform for storage and processing. Data transmission utilizes efficient and stable communication protocols, ensuring that real-time data is quickly and accurately transmitted to the processing system. Data processing begins with data cleaning and formatting to ensure accuracy and consistency.

[0135] During data storage and processing, we implement strict privacy protection measures to ensure the security of patients' personal health data. We use encryption technology to protect data transmission and storage, prevent data leakage or tampering, and comply with relevant laws and regulations to ensure that patients' privacy is fully protected.

[0136] Next, the previously trained personalized dosage decision model is fed with the patient's real-time physiological data and historical treatment data for intelligent analysis and processing. The model compares this data with the patient's individual characteristics and calculates the most appropriate aromatherapy dosage for the patient.

[0137] Dynamic adjustment of therapeutic dosage is the core of this implementation. Based on real-time physiological data, the system automatically adjusts aromatherapy dosage, ensuring that the patient always receives a dose that matches their current physiological state. To achieve this, real-time data must first be input into a previously trained personalized decision-making model, which automatically calculates the optimal dosage based on the patient's physiological indicators and treatment response.

[0138] For example, if a patient experiences abnormal fluctuations in heart rate or blood pressure, this could indicate an adverse reaction to the current dose of aromatherapy, and the model would automatically reduce the treatment dose. Conversely, if the patient's physiological parameters remain stable and no side effects occur, the model can increase the treatment dose based on their response, thereby improving efficacy. Furthermore, if a patient experiences an elevated temperature or other clinical symptoms during treatment, the model can immediately adjust the dose to mitigate side effects or adjust the treatment plan.

[0139] This dynamic adjustment not only targets the aromatherapy dosage but may also involve changes in treatment methods. For example, if aromatherapy causes discomfort to the patient, the system may recommend pausing or changing the treatment strategy, or using other auxiliary treatment methods to alleviate the discomfort. Through this real-time adjustment, the patient's treatment plan is always in the optimal state, thereby maximizing the treatment effect and reducing the risk of side effects.

[0140] During treatment, side effects are a factor that cannot be ignored. Although aromatherapy is a relatively mild treatment method, it can also cause some side effects, such as dizziness, nausea, and allergic reactions. Therefore, this embodiment specifically designs a side effect monitoring and emergency response mechanism to address possible adverse treatment reactions.

[0141] By monitoring patients' physiological data in real time, the system can proactively identify early signs of side effects. For example, if a patient's blood pressure suddenly rises or their temperature exceeds the normal range, the system will immediately issue an alert and recommend adjusting the treatment dosage or taking emergency measures. Doctors or caregivers will also receive prompt notifications and take further action. This emergency response mechanism ensures that patients receive timely care and intervention during treatment, preventing the worsening of side effects.

[0142] The system also builds a personalized side effect prediction model based on the patient's historical treatment data and side effect records. By analyzing the patient's tolerance to aromatherapy, the model can predict potential side effects and take appropriate preventative measures. For example, some patients may be allergic to specific types of aromatherapy oils. The model can identify this in advance based on their historical data and recommend changing the treatment medication or adjusting the aromatherapy oil concentration.

[0143] Real-time feedback and decision optimization are another key step in ensuring that the patient's condition matches the treatment plan during treatment. During each treatment, the patient's physiological data and treatment response are continuously transmitted to the system, which uses this data to optimize treatment decisions. During treatment, patient feedback (such as discomfort, pain, fatigue, etc.) is also input into the system to further optimize the treatment plan.

[0144] Patient feedback can be recorded and entered into the system via specialized medical devices or directly by medical staff. The system analyzes this feedback to optimize treatment strategies. For example, if a patient reports extreme discomfort after aromatherapy, the system will immediately adjust the dosage or even suspend treatment based on the patient's physiological data and historical treatment records.

[0145] After each treatment, the system generates a treatment report that comprehensively evaluates the effectiveness of the treatment. Doctors can use these reports to further understand the patient's response and make further treatment decisions. Through continuous real-time feedback and optimization adjustments, treatment plans can be continuously refined based on the patient's real-time status and long-term treatment response.

[0146] Another key advantage of this approach is its personalized treatment optimization. Each patient exhibits a different physiological response to aromatherapy, and each patient's condition, tolerance, and treatment efficacy vary significantly. Therefore, a personalized treatment optimization strategy is crucial.

[0147] Through real-time monitoring and dynamic adjustments, patients can receive personalized aromatherapy dosages based on their specific physiological conditions and treatment responses. This optimization can be flexibly adjusted according to the patient's response, not only improving the effectiveness of the treatment, but also reducing the risk of side effects, ensuring the safety and comfort of the treatment process.

[0148] In summary, this implementation, through real-time monitoring of the patient's physiological state and dynamic adjustment of treatment dosage using an intelligent decision-making system, not only maximizes treatment efficacy but also promptly addresses side effects and adverse reactions, ensuring patient safety. During each treatment, the treatment plan is optimized in real time to ensure the patient always receives the most appropriate dosage, thereby enhancing the therapeutic effectiveness of aromatherapy and improving the patient's quality of life.

[0149] This invention eliminates data heterogeneity from different clinical studies and patient populations through data normalization, ensuring consistency across different patient populations in the treatment dose decision model. This data normalization step effectively addresses data inconsistencies caused by differences in factors such as age, gender, and genetic characteristics, enabling the model to make predictions based on unified, standardized data. This optimization significantly improves the accuracy of treatment doses and reduces errors caused by data discrepancies, ensuring that patients receive the most appropriate aromatherapy dose and avoiding the negative impact of incorrect treatment doses on patient health.

[0150] This invention constructs a personalized dosage decision model to generate the most appropriate aromatherapy dosage based on each patient's characteristics, condition, and treatment response. This method incorporates clinical characteristics, such as age, gender, and genetic profile, to provide a customized treatment plan for each patient. Personalized dosage decisions not only improve treatment efficacy but also effectively minimize side effects and lower patient risks. This precise treatment plan allows patients to achieve better outcomes and improve their quality of life.

[0151] The present invention ensures continuous optimization of the treatment plan by monitoring the patient's physiological state in real time and dynamically adjusting the treatment dose based on the monitoring data. During the treatment process, the patient's physiological and therapeutic responses can be fed back into the model in real time, allowing the dosage to be adjusted at any time based on the patient's physical condition. This dynamic adjustment mechanism can promptly identify and correct side effects or adverse reactions that occur during treatment, thereby maximizing the safety and effectiveness of the treatment. Through this flexible adjustment method, the present invention significantly improves the individualization of treatment, allowing each patient to receive treatment at the most appropriate dose and reducing unnecessary risks.

[0152] The above formulas are all dimensionless and numerical calculations. The formulas are obtained by collecting a large amount of data and performing software simulation to obtain the most recent real situation. The preset parameters in the formulas are set by technicians in this field according to actual conditions.

[0153] The above description is merely illustrative of certain exemplary embodiments of the present invention. It goes without saying that those skilled in the art will be able to modify the described embodiments in various ways without departing from the spirit and scope of the present invention. Therefore, the above drawings and description are illustrative in nature and should not be construed as limiting the scope of protection of the claims.

[0154] It should be noted that, in this document, if there are relational terms such as first and second, etc., they are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "comprises", "comprising" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device that includes a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, an element defined by the sentence "comprising a ..." does not exclude the presence of other identical elements in the process, method, article or device that includes the element.

[0155] It should be understood that in the various embodiments of the present application, the size of the serial numbers of the above-mentioned processes does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present application.

[0156] Those skilled in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0157] Those skilled in the art will clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0158] The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed across multiple network units. Some or all of these units may be selected to achieve the purpose of this embodiment according to actual needs.

[0159] In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit.

[0160] The above description is merely a specific embodiment of the present application, but the scope of protection of the present application is not limited thereto. Any changes or substitutions that can be easily conceived by a person skilled in the art within the technical scope disclosed in this application should be included in the scope of protection of this application. Therefore, the scope of protection of this application should be based on the scope of protection of the claims.

[0161] The above description is merely illustrative of certain exemplary embodiments of the present invention. It goes without saying that those skilled in the art will be able to modify the described embodiments in various ways without departing from the spirit and scope of the present invention. Therefore, the above drawings and description are illustrative in nature and should not be construed as limiting the scope of protection of the claims.

Claims

1. A modeling method for aromatherapy dosage decision-making for cancer patients based on evidence-based data, characterized by: The following steps are involved: Obtain clinical data of cancer patients from multiple sources, including but not limited to the patients' age, gender, and genetic characteristics; Standardize clinical data to eliminate the impact of data heterogeneity on model training and ensure that data from different sources are compared and analyzed under the same standards; Build a data analysis model to identify key factors related to aromatherapy dosage by extracting features from standardized data; Using data analysis models for training to obtain the optimal dosage prediction model, the model predicts the patient's response to aromatherapy based on their specific characteristics; Generate personalized aromatherapy dosage decisions based on the trained dosage prediction model to ensure that each patient's treatment dosage matches their individual characteristics; Based on the generated therapeutic dose, the dose decision is continuously optimized and adjusted by monitoring the patient's treatment response in real time to address changes in the patient's response; The real-time monitoring steps are as follows: During treatment, clinical monitoring equipment is used to collect patients’ physiological data in real time; Correlate patient physiological data with treatment dose and response to treatment to assess patient response to aromatherapy; Based on real-time monitoring results, the most suitable aromatherapy dosage for the patient's current condition is calculated, and the patient's treatment plan is adjusted through feedback; The real-time optimization process of the treatment dose decision model is as follows: Input the patient's physiological data S, S={s j }={s1,s2,……,s m }, where s j represents the jth real-time physiological parameter of the patient, and m is the number of physiological parameter types; For each group of patients' physiological data S j , calculate the predicted value of the therapeutic dose, the calculation expression is as follows: D j =α1s1+α2s2+…+α m s m +b Where α1, α2, α m They are s1, s2, and s m The weight coefficient reflects the influence of each physiological data item on the predicted treatment dose, and β is the bias term; Optimize the prediction model by adjusting the weight coefficients to make it more adaptable to changes in real-time data and use optimization methods to minimize errors; Based on the optimized weight coefficients, the patient's aromatherapy dosage is adjusted in real time to ensure that each patient's treatment plan matches their current physiological state, thereby maximizing the treatment effect and reducing the risk of side effects.

2. The aromatherapy dosage decision modeling method for cancer patients based on evidence-based data according to claim 1 is characterized in that: The specific steps for standardization are as follows: Preprocess clinical data from different sources to remove outliers and noise data to ensure data quality; Convert characteristics of different patient groups into uniform standardized units to eliminate errors caused by different measurement units or dimensions; A normalization algorithm was used to process the patient's age, gender, and genetic characteristics so that they were within the same range, and each characteristic was weighted according to its importance; Statistical analysis methods are used to detect the correlation between various features, and the data standardization strategy is further adjusted based on the correlation analysis results.

3. The aromatherapy dosage decision modeling method for cancer patients based on evidence-based data according to claim 1, characterized in that: The specific steps for establishing a data analysis model are as follows: Collect processed and standardized data and input it into the machine learning algorithm; Divide the training data, use cross-validation method to evaluate model performance, and select the optimal model structure; The model parameters were adjusted through an optimization algorithm, and the objective function of error minimization was adopted to ensure the model's prediction accuracy for aromatherapy dosage decisions for cancer patients. Evaluation metrics were used to validate the model's performance and ensure that the model could adapt to the data characteristics of different patient groups.

4. The aromatherapy dosage decision modeling method for cancer patients based on evidence-based data according to claim 1, characterized in that: The steps for generating personalized aromatherapy dosage decisions are as follows: According to the trained dose prediction model, the clinical characteristics and standardized data of the tumor patients are input; Based on the predicted values output by the model, a personalized treatment dose is generated for each patient, and the treatment dose is fine-tuned according to the patient's condition and treatment history; Validate dose predictions through clinical expert feedback to ensure the dose maximizes therapeutic efficacy and minimizes side effects; According to the patient's response during treatment, the treatment dose is adjusted in real time to optimize the drug effect during treatment.

5. The aromatherapy dosage decision modeling method for cancer patients based on evidence-based data according to claim 1, characterized in that: The specific steps of the data analysis model training and optimization process are as follows: Input standardized dataset X, X={x i }={x1,x2,……,x n }, where x i represents the clinical characteristics of the i-th patient, and n is the total number of patients; The objective function is modeled for each patient's response value, and the goal is to minimize the error function of the model, which is as follows: Where L(θ) is the loss function, θ is the model parameter, and y i is the response value of the ith patient, is the predicted response value for the i-th patient; Use the optimization algorithm to iteratively update the parameter θ until the loss function L(θ) converges to the minimum value; In order to prevent overfitting, a regularization term λ‖θ‖ is added 2 , the optimization objectives are as follows: Where, L reg (θ) is the regularized error function, λ is the regularization coefficient, ‖θ‖ 2 is the sum of the squares of the parameters.

6. The aromatherapy dosage decision modeling method for cancer patients based on evidence-based data according to claim 1, characterized in that: The steps for adjusting the therapeutic dose are as follows: Adjust the dosage of aromatherapy according to the patient's side effects or response to treatment during treatment; Combine patient feedback data and physiological changes to calculate and predict the next treatment dose to prevent overdose or underdose; By setting critical values, the range of dosage adjustment is determined to ensure the safety and effectiveness of the treatment process.

Images