Functional gastrointestinal disease acupuncture scheme recommendation method and system based on big data

Abstract

The invention provides a functional gastrointestinal disease acupuncture scheme recommendation method and system based on big data, and the method comprises the steps: obtaining patient symptom information and patient physique information, inputting the patient symptom information into a gastrointestinal disease judgment model, obtaining a gastrointestinal disease judgment result, and carrying out the calculation of the patient physique information and the gastrointestinal disease judgment result, inputting the acupuncture scheme prediction model to obtain the use effect of the acupuncture scheme prediction result, screening the acupuncture scheme prediction result according to the ranking of the use effect of the acupuncture scheme prediction result, and inputting the screened acupuncture scheme into the acupuncture scheme verification model to obtain the physique verification information and the gastrointestinal disease verification result. According to a comparison result of the physique verification information and the patient physique information and a comparison result of the gastrointestinal disease verification result and the gastrointestinal disease judgment result, an acupuncture scheme recommendation result is determined; according to the method, the acupuncture scheme recommendation result can be determined, and the situation that an improper acupuncture scheme is determined for a patient due to insufficient personal experience of a doctor is avoided.

Description

Technical Field

[0001] This application relates to the field of big data technology, and in particular to a method and system for recommending acupuncture treatment plans for functional gastrointestinal diseases based on big data. Background Technology

[0002] Functional gastrointestinal disorders are a group of diseases characterized primarily by gastrointestinal dysfunction without any clear organic lesions. These mainly include functional dyspepsia, irritable bowel syndrome, functional constipation, functional bloating, functional diarrhea, and functional abdominal pain syndrome. Acupuncture can be used to relieve symptoms or treat functional gastrointestinal disorders.

[0003] Currently, in determining acupuncture plans for functional gastrointestinal disorders, due to the subjectivity and limitations of doctors' personal experience and technical level, it is possible that inappropriate acupuncture plans will be formulated and implemented for patients, resulting in unsatisfactory treatment effects for different patients. Therefore, it is crucial to provide reasonable acupuncture plans for patients with functional gastrointestinal disorders to assist doctors in formulating acupuncture plans for patients. Summary of the Invention

[0004] In view of the shortcomings of the prior art described above, this application provides a method and system for recommending acupuncture programs for functional gastrointestinal diseases based on big data, in order to solve the above-mentioned technical problems.

[0005] According to one aspect of the embodiments of this application, a method for recommending acupuncture programs for functional gastrointestinal diseases based on big data is provided. The method includes: acquiring patient symptom information and patient constitution information; inputting the patient symptom information into a gastrointestinal disease judgment model to obtain a gastrointestinal disease judgment result; inputting the patient constitution information and the gastrointestinal disease judgment result into an acupuncture program prediction model to obtain acupuncture program prediction result and the usage effect of the acupuncture program prediction result; the acupuncture program prediction model is trained on a pre-established prediction model using first sample data; the acupuncture program prediction results are filtered according to the usage effect ranking of the acupuncture program prediction results to obtain a filtered acupuncture program; the filtered acupuncture program is input into an acupuncture program verification model to obtain constitution verification information and gastrointestinal disease verification result; the acupuncture program verification model is trained on a pre-established verification model using second sample data; and acupuncture program recommendation result is determined based on the comparison result of the constitution verification information and the patient constitution information, and the comparison result of the gastrointestinal disease verification result and the gastrointestinal disease judgment result.

[0006] In one embodiment of this application, the process of training a pre-established prediction model using first sample data to obtain the acupuncture pattern prediction model includes: inputting the first sample data into the pre-established prediction model to obtain a predicted acupuncture pattern and the effect of the predicted acupuncture pattern; the first sample data includes: patient's historical diagnosis results, patient's historical physical condition information, acupuncture pattern used by the patient, the treatment effect of the acupuncture pattern used by the patient, the correspondence between the combination of the patient's historical diagnosis results and the patient's historical physical condition information and the acupuncture pattern used by the patient; the correspondence between the acupuncture pattern used by the patient and the treatment effect of the acupuncture pattern used by the patient; and adjusting the parameters in the pre-established prediction model with the goal of minimizing the gap between the predicted acupuncture pattern and the acupuncture pattern used by the patient, and the gap between the effect of the predicted acupuncture pattern and the treatment effect of the acupuncture pattern used by the patient, to obtain the acupuncture pattern prediction model.

[0007] In one embodiment of this application, the difference between the predicted acupuncture plan and the acupuncture plan used by the patient, and the difference between the effect of the predicted acupuncture plan and the treatment effect of the acupuncture plan used by the patient, are characterized by a loss function, the expression of which is as follows: ,in, Represents the loss function. The weights represent the cross-entropy loss. Represents cross-entropy loss, This indicates the weight of the mean squared error loss. This represents the mean squared error loss. This represents the total number of categories of acupuncture treatments. Indicates the patient's use of the first Category tags corresponding to each acupuncture treatment plan Indicates the first Predicted probabilities of acupuncture treatments Represents the total number of samples. Indicates the first The therapeutic effects of acupuncture on patients in a sample Indicates the first The predicted effectiveness of the acupuncture treatment obtained from the predicted sample was, in the ... The therapeutic effects of acupuncture on patients in the sample, and the treatment outcomes after the first... The effectiveness of the predicted acupuncture treatment plan obtained from each sample prediction is characterized by either a coding or numerical form. .

[0008] In one embodiment of this application, if the pre-established prediction model includes: a first input layer, a first multilayer perceptron layer, a first attention mechanism layer, a first feature fusion layer, a first fully connected layer, and a first output layer, then the process of inputting the first sample data into the pre-established prediction model to obtain the predicted acupuncture scheme and the predicted effect of the acupuncture scheme includes: inputting the first sample data into the first input layer to obtain a first encoding vector; the first input layer is used to perform normalization and encoding operations on the patient's historical diagnosis results, the patient's historical physical condition information, the acupuncture scheme used by the patient, and the treatment effect of the acupuncture scheme used by the patient; inputting the first encoding vector into the first multilayer perceptron layer to obtain first nonlinear feature information; the first multilayer perceptron layer is used to perform nonlinear feature extraction and feature dimension transformation on the first encoding vector; inputting the first encoding vector into the first attention mechanism layer to obtain attention weight information; the first attention mechanism layer is used to calculate the attention weight between features in the first encoding vector; and inputting the first nonlinear feature... The information and the attention weight information are input into the first feature fusion layer to obtain first fused feature information; the first feature fusion layer is used to perform feature concatenation and feature dimension transformation on the first nonlinear feature information and the attention weight information; the first fused feature information is input into the first fully connected layer to obtain the predicted acupuncture scheme and the effect of the predicted acupuncture scheme; the first fully connected layer is used to map the first fused feature information to a first preset spatial dimension to obtain a first preset spatial dimension feature; and activate the first preset spatial dimension feature; the predicted acupuncture scheme and the effect of the predicted acupuncture scheme are input into the first output layer to obtain the difference between the predicted acupuncture scheme and the acupuncture scheme used by the patient, and the difference between the effect of the predicted acupuncture scheme and the treatment effect of the acupuncture scheme used by the patient; the first output layer is used to calculate the difference between the predicted acupuncture scheme and the acupuncture scheme used by the patient, and the difference between the effect of the predicted acupuncture scheme and the treatment effect of the acupuncture scheme used by the patient.

[0009] In one embodiment of this application, the process of training a pre-established verification model using second sample data to obtain the acupuncture protocol verification model includes: inputting the second sample data into the pre-established verification model to obtain constitution prediction information and gastrointestinal disease prediction results; the second sample data includes: patient's historical diagnosis results, patient's historical constitution information, acupuncture protocol used by the patient, and the correspondence between the combination of the acupuncture protocol used by the patient and the patient's historical diagnosis results and the patient's historical constitution information; with the goal of minimizing the gap between the constitution prediction information and the patient's historical constitution information, and the gap between the gastrointestinal disease prediction results and the patient's historical diagnosis results, adjusting the parameters in the pre-established verification model to obtain the acupuncture protocol verification model.

[0010] In one embodiment of this application, if the pre-established verification model includes: a second input layer, a second multilayer perceptron layer, a second feature fusion layer, a second fully connected layer, and a second output layer, then the process of inputting the second sample data into the pre-established verification model to obtain physical fitness prediction information and gastrointestinal disease prediction results includes: inputting the second sample data into the second input layer to obtain a second encoding vector; the second input layer is used to perform normalization and encoding operations on the second sample data; inputting the second encoding vector into the second multilayer perceptron layer to obtain second nonlinear feature information; the second multilayer perceptron layer is used to perform nonlinear feature extraction and feature dimension transformation on the second encoding vector; inputting the second nonlinear feature information into the second feature fusion layer to obtain second fused feature information; and the second feature fusion layer is used to perform feature concatenation on the second nonlinear feature information. The second fused feature information is input into the second fully connected layer to obtain the constitution prediction information and the gastrointestinal disease prediction result; the second fully connected layer is used to map the second fused feature information to a second preset spatial dimension to obtain the second preset spatial dimension feature; and to activate the second preset spatial dimension feature; The physical condition prediction information and the gastrointestinal disease prediction results are input into the second output layer to obtain the gap between the physical condition prediction information and the patient's historical physical condition information, as well as the gap between the gastrointestinal disease prediction results and the patient's historical diagnosis results. The second output layer is used to calculate the gap between the physical condition prediction information and the patient's historical physical condition information, as well as the gap between the gastrointestinal disease prediction results and the patient's historical diagnosis results.

[0011] In one embodiment of this application, the process of filtering the acupuncture plan prediction results according to the ranking of the usage effects of the acupuncture plan prediction results to obtain the filtered acupuncture plans includes: if the number of acupuncture plans in the acupuncture plan prediction results is less than a preset number threshold, then the acupuncture plan prediction results are used as the filtered acupuncture plans; if the number of acupuncture plans in the acupuncture plan prediction results is greater than or equal to the preset number threshold, then the usage effects of the acupuncture plan prediction results are ranked in ascending order to obtain an ascending sorted sequence; and a preset number of acupuncture plans ranked at the bottom in the ascending sorted sequence are used as the filtered acupuncture plans; or, the usage effects of the acupuncture plan prediction results are ranked in descending order to obtain a descending sorted sequence; and a preset number of acupuncture plans ranked at the top in the descending sorted sequence are used as the filtered acupuncture plans.

[0012] In one embodiment of this application, the process of determining the recommended acupuncture plan based on the comparison results of the physical fitness verification information and the patient's physical fitness information, and the comparison results of the gastrointestinal disease verification results and the gastrointestinal disease judgment results, includes: if the physical fitness verification information includes the patient's physical fitness information and the gastrointestinal disease verification results include the gastrointestinal disease judgment results, then the screened acupuncture plan is taken as the recommended acupuncture plan; if the physical fitness verification information does not include the patient's physical fitness information, and / or, the gastrointestinal disease verification results do not include the gastrointestinal disease judgment results, then no recommended plan is taken as the recommended acupuncture plan.

[0013] In one embodiment of this application, before obtaining patient symptom information and patient physical condition information, the method further includes: inputting the patient symptom information and patient physical condition information through an interactive device; or, inputting patient historical diagnosis information through the interactive device and extracting the patient symptom information and patient physical condition information from the patient historical diagnosis information; the patient historical diagnosis information includes: patient diagnosis information and patient treatment information.

[0014] According to one aspect of the embodiments of this application, a functional gastrointestinal disease acupuncture program recommendation system based on big data is provided, including: an information collection module for acquiring patient symptom information and patient physical condition information; The system comprises the following modules: a disease judgment module, which inputs the patient's symptom information into a gastrointestinal disease judgment model to obtain a gastrointestinal disease judgment result; a treatment plan prediction module, which inputs the patient's constitution information and the gastrointestinal disease judgment result into an acupuncture treatment plan prediction model to obtain an acupuncture treatment plan prediction result and the effectiveness of the predicted result; the acupuncture treatment plan prediction model is trained on a pre-established prediction model using first sample data; a treatment plan screening module, which ranks the acupuncture treatment plan prediction results according to their effectiveness to obtain screened acupuncture treatment plans; a treatment plan verification module, which inputs the screened acupuncture treatment plans into an acupuncture treatment plan verification model to obtain constitution verification information and gastrointestinal disease verification results; the acupuncture treatment plan verification model is trained on a pre-established verification model using second sample data; and an information comparison module, which determines the recommended acupuncture treatment plan based on the comparison results between the constitution verification information and the patient's constitution information, and between the gastrointestinal disease verification results and the gastrointestinal disease judgment result.

[0015] The beneficial effects of this application are as follows: This application obtains patient symptom information and patient physical condition information, inputs the patient symptom information into a gastrointestinal disease judgment model to obtain gastrointestinal disease judgment results, inputs the patient physical condition information and gastrointestinal disease judgment results into an acupuncture plan prediction model to obtain acupuncture plan prediction results and the effectiveness of the acupuncture plan prediction results, ranks the acupuncture plan prediction results according to the effectiveness of the acupuncture plan prediction results, and obtains selected acupuncture plans, inputs the selected acupuncture plans into an acupuncture plan verification model to obtain physical condition verification information and gastrointestinal disease verification results, and determines the recommended acupuncture plan results based on the comparison results of physical condition verification information and patient physical condition information, as well as the comparison results of gastrointestinal disease verification results and gastrointestinal disease judgment results. The above process can determine the recommended acupuncture plan results based on the patient's physical condition information and gastrointestinal disease judgment results, which can be used by doctors or patients to determine the final acupuncture plan. This helps to avoid the situation where the doctor's personal experience and technical level limit the determination of an unsuitable acupuncture plan for the patient, resulting in unsatisfactory treatment effects.

[0016] It should be understood that the above general description and the following detailed description are exemplary and explanatory only, and do not limit this application. Attached Figure Description

[0017] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this application and, together with the description, serve to explain the principles of this application. It is obvious that the drawings described below are merely some embodiments of this application, and those skilled in the art can obtain other drawings based on these drawings without any inventive effort. In the drawings: Figure 1This is a schematic diagram illustrating an exemplary system architecture as shown in an exemplary embodiment of this application; Figure 2 This is a flowchart illustrating an exemplary embodiment of the present application of a method for recommending acupuncture protocols for functional gastrointestinal diseases based on big data; Figure 3 This is a schematic diagram illustrating the structure of a pre-established prediction model, as shown in an exemplary embodiment of this application. Figure 4 This is a schematic diagram illustrating the structure of a pre-established verification model, as shown in an exemplary embodiment of this application. Figure 5 This is a block diagram illustrating an exemplary embodiment of the present application of a big data-based acupuncture protocol recommendation system for functional gastrointestinal diseases. Detailed Implementation

[0018] The following specific examples illustrate the implementation of this application. Those skilled in the art can easily understand other advantages and effects of this application from the content disclosed in this specification. This application can also be implemented or applied through other different specific embodiments. Various details in this specification can also be modified or changed based on different viewpoints and applications without departing from the spirit of this application. In the absence of conflict, the following embodiments and features in the embodiments can be combined with each other.

[0019] It should be noted that the illustrations provided in the following embodiments are only schematic representations of the basic concept of this application. The drawings only show the components related to this application and are not drawn according to the actual number, shape and size of the components in the actual implementation. In the actual implementation, the form, quantity and proportion of each component can be arbitrarily changed, and the layout of the components may also be more complex.

[0020] In the following description, numerous details are explored to provide a more thorough explanation of embodiments of the present application. However, it will be apparent to those skilled in the art that embodiments of the present application may be practiced without these specific details. In other embodiments, well-known structures and devices are shown in block diagram form rather than in detail to avoid obscuring embodiments of the present application.

[0021] Figure 1 This is a schematic diagram illustrating an exemplary system architecture as shown in an exemplary embodiment of this application.

[0022] Reference Figure 1As shown, the system architecture may include a data acquisition device 101 and a data processing device 102. The data processing device 102 may be at least one of a desktop graphics processing unit (GPU) computer, a GPU computing cluster, or a neural network computer. Technical personnel can use this data processing device 102 to acquire patient symptom information and patient constitution information. The patient symptom information is input into a gastrointestinal disease judgment model to obtain a gastrointestinal disease judgment result. The patient constitution information and the gastrointestinal disease judgment result are input into an acupuncture plan prediction model to obtain acupuncture plan prediction results and their effectiveness. The acupuncture plan prediction results are ranked according to their effectiveness, and then filtered to obtain selected acupuncture plans. These selected acupuncture plans are input into an acupuncture plan verification model to obtain constitution verification information and gastrointestinal disease verification results. Based on the comparison results between the constitution verification information and the patient's constitution information, and the comparison results between the gastrointestinal disease verification results and the gastrointestinal disease judgment results, a recommended acupuncture plan is determined. The data acquisition device 101 is used to collect patient symptom information and patient constitution information and provide them to the data processing device 102 for processing.

[0023] In a schematic manner, after acquiring patient symptom information and patient physical condition information from acquisition device 101, data processing device 102 inputs the patient symptom information into a gastrointestinal disease judgment model to obtain a gastrointestinal disease judgment result. It then inputs the patient physical condition information and the gastrointestinal disease judgment result into an acupuncture plan prediction model to obtain acupuncture plan prediction results and their effectiveness. Based on the ranking of the effectiveness of the acupuncture plan prediction results, the prediction results are filtered to obtain selected acupuncture plans. These selected acupuncture plans are then input into an acupuncture plan verification model to obtain physical condition verification information and gastrointestinal disease verification results. Based on the comparison results between the physical condition verification information and the patient's physical condition information, and between the gastrointestinal disease verification results and the gastrointestinal disease judgment results, a recommended acupuncture plan is determined. This process allows for the determination of a recommended acupuncture plan based on the patient's physical condition information and the gastrointestinal disease judgment results, providing a reference for doctors or patients to determine the final acupuncture plan. This helps avoid situations where doctors, due to limitations in their personal experience and technical level, determine inappropriate acupuncture plans for patients, leading to unsatisfactory treatment results.

[0024] The implementation details of the technical solutions in the embodiments of this application are described in detail below: Figure 2 This is a flowchart illustrating an exemplary embodiment of the present application of a method for recommending acupuncture protocols for functional gastrointestinal disorders based on big data. (Refer to...) Figure 2 As shown, this method for recommending acupuncture protocols for functional gastrointestinal disorders based on big data includes at least steps S210 to S260, which are detailed below: In step S210, patient symptom information and patient constitution information are obtained. In one embodiment of this application, patient symptom information includes symptoms such as stomach pain, acid reflux, bloating, belching, diarrhea, and constipation. Patient constitution information covers types in traditional Chinese medicine constitution classification, such as balanced constitution, qi deficiency constitution, yang deficiency constitution, yin deficiency constitution, phlegm-dampness constitution, damp-heat constitution, blood stasis constitution, qi stagnation constitution, and special constitution. Patient symptom information and patient constitution information can be obtained through a health questionnaire filled out by the patient on an interactive device or electronic terminal, or through preliminary detection and data collection using advanced intelligent medical equipment, followed by verification and improvement by professional medical personnel.

[0025] In step S220, patient symptom information is input into the gastrointestinal disease diagnosis model to obtain the gastrointestinal disease diagnosis result. In one embodiment of this application, the gastrointestinal disease diagnosis model is a functional gastrointestinal disease diagnostic model built based on the Rome criteria (e.g., Rome IV criteria).

[0026] In step S230, the patient's physical condition information and the judgment results of gastrointestinal diseases are input into the acupuncture plan prediction model to obtain the acupuncture plan prediction results and the effect of the acupuncture plan prediction results. In one embodiment of this application, the acupuncture plan prediction model is obtained by training a pre-established prediction model with first sample data. The first sample data includes: the patient's historical diagnosis results, the patient's historical physical condition information, the acupuncture plan used by the patient, the treatment effect of the acupuncture plan used by the patient, the correspondence between the combination of the patient's historical diagnosis results and the patient's historical physical condition information and the acupuncture plan used by the patient, and the correspondence between the acupuncture plan used by the patient and the treatment effect of the acupuncture plan used by the patient. The pre-established prediction model includes: a first input layer, a first multilayer perceptron layer, a first attention mechanism layer, a first feature fusion layer, a first fully connected layer, and a first output layer. The process of training the pre-established prediction model with the first sample data to obtain the acupuncture plan prediction model can fully learn the complex relationship between the patient's physical condition, gastrointestinal diseases, acupuncture plan, and the effect of acupuncture plan use, as well as how these relationships affect the selection of acupuncture plan, thereby improving the accuracy and reliability of acupuncture plan prediction and providing patients with more personalized and accurate acupuncture plan recommendation results.

[0027] In one embodiment of this application, the acupuncture pattern prediction result can be multiple acupuncture patterns or a single acupuncture pattern.

[0028] In step S240, the acupuncture plan prediction results are ranked according to their effectiveness, and then filtered to obtain filtered acupuncture plans. In one embodiment of this application, the effectiveness of the acupuncture plan prediction results can be evaluated by terms such as excellent, good, average, and poor, or by evaluation scores (e.g., 60, 70, and 80 points). If there is only one acupuncture plan in the prediction results, then that plan is selected as the filtered acupuncture plan; if there are multiple acupuncture plans in the prediction results, then the process of filtering the prediction results involves selecting the acupuncture plan with better effectiveness from the prediction results.

[0029] In step S250, the selected acupuncture plan is input into the acupuncture plan verification model to obtain constitution verification information and gastrointestinal disease verification results. In one embodiment of this application, the acupuncture plan verification model is obtained by training a pre-established verification model with second sample data. The second sample data includes: the patient's historical diagnosis results, the patient's historical constitution information, the acupuncture plan used by the patient, and the correspondence between the combination of the acupuncture plan used by the patient and the patient's historical diagnosis results and patient's historical constitution information. The pre-established verification model includes: a second input layer, a second multilayer perceptron layer, a second feature fusion layer, a second fully connected layer, and a second output layer. The process of training the pre-established verification model with the second sample data to obtain the acupuncture plan verification model can fully learn the complex relationship between the selected acupuncture plan and the patient's constitution and gastrointestinal diseases, as well as the applicability and effectiveness of the selected acupuncture plan in terms of the patient's constitution and gastrointestinal diseases, ensuring that the recommended acupuncture plan can achieve the expected therapeutic effect in practical applications, thereby further enhancing the reliability and practicality of the acupuncture plan recommendation results.

[0030] In step S260, the recommended acupuncture plan is determined based on the comparison results of the physical fitness verification information and the patient's physical fitness information, as well as the comparison results of the gastrointestinal disease verification results and the gastrointestinal disease judgment results. In one embodiment of this application, the selected acupuncture plan is verified based on the comparison results of the physical fitness verification information and the patient's physical fitness information, as well as the comparison results of the gastrointestinal disease verification results and the gastrointestinal disease judgment results, thereby improving the accuracy and effectiveness of the recommended acupuncture plan.

[0031] In one embodiment of this application, a recommended acupuncture plan can be determined based on the patient's physical condition information and the results of the gastrointestinal disease assessment. This plan can be used by doctors or patients to determine the final acupuncture plan, which helps to avoid situations where unsuitable acupuncture plans are determined for patients due to the limitations of doctors' personal experience and technical level, resulting in unsatisfactory treatment effects.

[0032] In one embodiment of this application, the process of training a pre-established prediction model using first sample data to obtain an acupuncture treatment prediction model includes: The first sample data is input into a pre-established prediction model to obtain the predicted acupuncture plan and the predicted effect of the acupuncture plan. In one embodiment of this application, the first sample data includes: the patient's historical diagnosis results, the patient's historical physical condition information, the acupuncture plan used by the patient, the therapeutic effect of the acupuncture plan used by the patient, the correspondence between the combination of the patient's historical diagnosis results and the patient's historical physical condition information and the acupuncture plan used by the patient; and the correspondence between the acupuncture plan used by the patient and the therapeutic effect of the acupuncture plan used by the patient. The pre-established prediction model includes: a first input layer, a first multilayer perceptron layer, a first attention mechanism layer, a first feature fusion layer, a first fully connected layer, and a first output layer. The process of inputting the first sample data into the pre-established prediction model to obtain the predicted acupuncture plan and the predicted effect of the acupuncture plan includes: inputting the first sample data into the first input layer to obtain a first encoding vector; inputting the first encoding vector into the first multilayer perceptron layer to obtain a first encoding vector; and inputting the first encoding vector into the first multilayer perceptron layer to obtain a first encoding vector. The first nonlinear feature information is obtained by inputting the first encoding vector into the first attention mechanism layer; the first nonlinear feature information and the attention weight information are input into the first feature fusion layer to obtain the first fused feature information; the first fused feature information is input into the first fully connected layer to obtain the predicted acupuncture plan and the predicted effect of the acupuncture plan; the predicted acupuncture plan and the predicted effect of the acupuncture plan are input into the first output layer to obtain the difference between the predicted acupuncture plan and the acupuncture plan used by the patient, and the difference between the effect of the predicted acupuncture plan and the treatment effect of the patient using the acupuncture plan. The above process can accurately capture the intrinsic relationship between the patient's constitution, gastrointestinal diseases and the acupuncture plan and the effect of the acupuncture plan by deep learning the complex relationships in the first sample data.

[0033] With the goal of minimizing the gap between the predicted acupuncture plan and the acupuncture plan used by the patient, and the gap between the predicted effect of the acupuncture plan and the treatment effect of the patient using the acupuncture plan, the parameters in the pre-established prediction model are adjusted to obtain the acupuncture plan prediction model. In one embodiment of this application, during the adjustment of the parameters in the pre-established prediction model, if the number of parameter adjustments in the pre-established prediction model reaches a preset threshold (e.g., 500 times), or if the sum of the gap between the predicted acupuncture plan and the acupuncture plan used by the patient and the gap between the predicted effect of the acupuncture plan and the treatment effect of the patient using the acupuncture plan is less than or equal to a preset error threshold (e.g., 0.01), then the adjustment of the parameters in the pre-established prediction model is stopped, and the acupuncture plan prediction model is obtained.

[0034] In one embodiment of this application, the difference between the predicted acupuncture plan and the acupuncture plan used by the patient, and the difference between the predicted effect of the acupuncture plan and the treatment effect of the patient using the acupuncture plan, are characterized by a loss function, the expression of which is as follows: Equation (1) in, Represents the loss function. The weights represent the cross-entropy loss. Represents cross-entropy loss, This indicates the weight of the mean squared error loss. This represents the mean squared error loss. This represents the total number of categories of acupuncture treatments. Indicates the patient's use of the first Category tags corresponding to each acupuncture treatment plan Indicates the first The predicted probability of acupuncture treatment plan (i.e., the predicted acupuncture treatment plan). Represents the total number of samples. Indicates the first The therapeutic effects of acupuncture on patients in a sample Indicates the first The predicted effectiveness of the acupuncture treatment obtained from the predicted sample was, in the ... The treatment effects of acupuncture on patients in each sample (in text or numerical form), and the treatment outcomes after the first acupuncture session. The effectiveness of the predicted acupuncture treatment plan obtained from each sample (in text or numerical form) is characterized using either coded or numerical methods. .

[0035] In one embodiment of this application, if the pre-established prediction model includes: a first input layer, a first multilayer perceptron layer, a first attention mechanism layer, a first feature fusion layer, a first fully connected layer, and a first output layer, then the process of inputting the first sample data into the pre-established prediction model to obtain the predicted acupuncture scheme and the predicted effect of the acupuncture scheme includes: The first sample data is input into the first input layer to obtain the first encoding vector. In one embodiment of this application, the first input layer is used to perform normalization and encoding operations on the patient's historical diagnosis results, patient's historical physical condition information, the acupuncture plan used by the patient, and the treatment effect of the acupuncture plan used by the patient. The normalization operation can be performed using the Z-Score formula or the Min-Max scaling formula, without specific limitations. The process of encoding the patient's historical diagnosis results, patient's historical physical condition information, patient's acupuncture plan used by the patient, and the treatment effect of the acupuncture plan used by the patient is implemented by an encoder.

[0036] The first encoded vector is input into the first multilayer perceptron layer to obtain the first nonlinear feature information. In one embodiment of this application, the first multilayer perceptron layer is used to perform nonlinear feature extraction and feature dimension transformation on the first encoded vector. The first multilayer perceptron layer is beneficial for capturing complex feature patterns in the first encoded vector. Through nonlinear transformation, the first encoded vector is mapped to a lower-dimensional feature space (e.g., from 128 dimensions to 64 dimensions), thereby enhancing the ability of the pre-established prediction model to identify the inherent patterns of the first sample data. The first multilayer perceptron layer includes an input layer, multiple hidden neurons, and an output layer. Each neuron performs nonlinear combination of the first encoded vector through an activation function (e.g., ReLU, Sigmoid, etc.) to generate feature information with stronger expressive power, which helps to improve the modeling accuracy of the pre-established prediction model for the complex relationship between acupuncture treatment and patient constitution and gastrointestinal diseases.

[0037] The first encoding vector is input into the first attention mechanism layer to obtain attention weight information. In one embodiment of this application, the first attention mechanism layer is used to calculate the attention weights between features in the first encoding vector; the first attention mechanism layer dynamically assigns different importance weights to the features in the first encoding vector through a self-attention mechanism, so that the pre-established prediction model can focus on the feature dimensions that have a more significant impact on the predicted acupuncture scheme.

[0038] The first nonlinear feature information and attention weight information are input into the first feature fusion layer to obtain the first fused feature information. In one embodiment of this application, the first feature fusion layer is used to perform feature concatenation and feature dimension transformation on the first nonlinear feature information and attention weight information. The first feature fusion layer includes a concatenation operation module and a dimension transformation module. The concatenation operation module concatenates the first nonlinear feature information and attention weight information to obtain concatenated feature information, while the dimension transformation module adjusts the dimensions of the concatenated feature information to adapt to the input requirements of subsequent layers. This fusion method not only preserves diverse features but also strengthens the influence of key features through the attention mechanism, which helps to improve the ability of the pre-established prediction model to capture complex relationships.

[0039] The first fused feature information is input into the first fully connected layer to obtain the predicted acupuncture scheme and the effect of the predicted acupuncture scheme. In one embodiment of this application, the first fully connected layer is used to map the first fused feature information to a first preset spatial dimension to obtain the first preset spatial dimension feature; and to activate the first preset spatial dimension feature; the first preset spatial dimension is 128 dimensions, 256 dimensions, etc., and a soft activation function is used to activate the first preset spatial dimension feature.

[0040] In one embodiment of this application, the first fully connected layer maps the features of a first preset spatial dimension to the output space through a soft activation function, generating a preliminary probability distribution for predicting acupuncture schemes and their effects. The first fully connected layer integrates the complex features extracted by the first feature fusion layer through a fully connected structure, enabling the pre-established prediction model to output multi-dimensional prediction results based on comprehensive information about the patient's physical characteristics and gastrointestinal disease state. The selection of the activation function ensures the nonlinearity of the output values ​​and the rationality of the probability distribution, providing effective input for the loss calculation of subsequent output layers.

[0041] The predicted acupuncture plan and its predicted effect are input into the first output layer to obtain the difference between the predicted acupuncture plan and the acupuncture plan used by the patient, and the difference between the predicted effect of the acupuncture plan and the treatment effect of the patient using the acupuncture plan. In one embodiment of this application, the first output layer is used to calculate the difference between the predicted acupuncture plan and the acupuncture plan used by the patient, and the difference between the predicted effect of the acupuncture plan and the treatment effect of the patient using the acupuncture plan. The difference between the predicted acupuncture plan and the acupuncture plan used by the patient is calculated using the cross-entropy loss function, and the difference between the predicted effect of the acupuncture plan and the treatment effect of the patient using the acupuncture plan is calculated using the mean squared error loss formula. After obtaining the gap between the predicted acupuncture plan and the acupuncture plan used by the patient, and the gap between the effect of the predicted acupuncture plan and the treatment effect of the patient's acupuncture plan, the loss function value is calculated according to formula (1). Based on the loss function value, the parameters in the pre-established prediction model are continuously adjusted so that the predicted acupuncture plan gradually approaches the acupuncture plan actually used by the patient, and the effect of the predicted acupuncture plan gradually approaches the actual treatment effect of the patient's acupuncture plan. The above process can fully explore the potential patterns in the patient's historical diagnosis results, patient's historical physical information and other multi-dimensional data, so as to provide more accurate and personalized acupuncture plan recommendations for patients with functional gastrointestinal diseases. Moreover, the acupuncture plan prediction model will continue to expand the sample data. With the continuous increase of the sample data and the continuous optimization of the acupuncture plan prediction model, the accuracy and effectiveness of the acupuncture plan recommended by the optimized acupuncture plan prediction model will be further improved.

[0042] In one embodiment of this application, the process of training a pre-established verification model using second sample data to obtain an acupuncture scheme verification model includes: The second sample data is input into a pre-established validation model to obtain physical condition prediction information and gastrointestinal disease prediction results. In one embodiment of this application, the second sample data includes: the patient's historical diagnosis results, the patient's historical physical condition information, the acupuncture treatment plan used by the patient, and the correspondence between the combination of the acupuncture treatment plan used by the patient and the patient's historical diagnosis results and the patient's historical physical condition information. The pre-established verification model includes: a second input layer, a second multilayer perceptron layer, a second feature fusion layer, a second fully connected layer, and a second output layer. The second input layer is used to perform normalization and encoding operations on the second sample data. The second multilayer perceptron layer is used to perform nonlinear feature extraction and feature dimension transformation on the second encoded vector. The second feature fusion layer is used to perform feature concatenation on the second nonlinear feature information. The second fully connected layer is used to map the second fused feature information to a second preset spatial dimension to obtain the second preset spatial dimension feature and activate the second preset spatial dimension feature. The second output layer is used to calculate the gap between the physical condition prediction information and the patient's historical physical condition information, as well as the gap between the gastrointestinal disease prediction results and the patient's historical diagnosis results. The above process can accurately capture the intrinsic connection between the acupuncture treatment plan and the patient's physical condition and gastrointestinal diseases by deeply learning the complex relationships in the second sample data.

[0043] With the goal of minimizing the gap between predicted physical constitution information and the patient's historical physical constitution information, as well as the gap between predicted gastrointestinal disease results and the patient's historical diagnostic results, the parameters in a pre-established validation model are adjusted to obtain an acupuncture protocol validation model. In one embodiment of this application, during the adjustment of the parameters in the pre-established validation model, if the number of parameter adjustments in the pre-established validation model reaches a preset threshold (e.g., 500 times), or if the sum of the gaps between predicted physical constitution information and the patient's historical physical constitution information, and the gaps between predicted gastrointestinal disease results and the patient's historical diagnostic results is less than or equal to a preset error threshold (e.g., 0.01), then the adjustment of the parameters in the pre-established validation model is stopped, and the acupuncture protocol validation model is obtained.

[0044] In one embodiment of this application, the gap between the physical fitness prediction information and the patient's historical physical fitness information, as well as the gap between the gastrointestinal disease prediction results and the patient's historical diagnosis results, are characterized by a first loss function, the calculation formula of which is as follows: Equation (2) in, Denotes the first loss function. This represents the weight of the cross-loss term for physical fitness information. This indicates a cross-loss of physical fitness information. This indicates the weight of the cross-loss term representing the diagnostic results. This indicates crossover loss in diagnostic results. This represents the total number of categories of physical fitness information. Indicates the first Category labels corresponding to each patient's historical physical condition information Indicates the first The predictive probability of a patient’s historical physical condition information (i.e., physical condition prediction information). This represents the total number of categories of diagnostic results. Indicates the first Category labels corresponding to each patient's historical diagnosis results Indicates the first The predictive probability of a patient’s historical diagnoses (i.e., the predictive outcome of gastrointestinal diseases). .

[0045] In one embodiment of this application, if the pre-established verification model includes: a second input layer, a second multilayer perceptron layer, a second feature fusion layer, a second fully connected layer, and a second output layer, then the process of inputting the second sample data into the pre-established verification model to obtain physical fitness prediction information and gastrointestinal disease prediction results includes: The second sample data is input into the second input layer to obtain the second encoding vector. In one embodiment of this application, the second input layer is used to perform normalization and encoding operations on the second sample data; the normalization operation can be performed using the Z-Score formula or the Min-Max scaling formula, without specific limitations. The process of encoding the patient's historical diagnosis results, patient's historical physical condition information, and the acupuncture treatment plan used by the patient is implemented through an encoder.

[0046] The second encoded vector is input into the second multilayer perceptron layer to obtain the second nonlinear feature information. In one embodiment of this application, the second multilayer perceptron layer is used to perform nonlinear feature extraction and feature dimension transformation on the second encoded vector. The second multilayer perceptron layer is beneficial for capturing complex feature patterns in the second encoded vector. Through nonlinear transformation, the second encoded vector is mapped to a lower-dimensional feature space (e.g., from 128 dimensions to 64 dimensions), thereby enhancing the ability of the pre-established validation model to identify the inherent patterns of the second sample data. The second multilayer perceptron layer includes an input layer, multiple hidden neurons, and an output layer. Each neuron uses an activation function (e.g., ReLU, Sigmoid, etc.) to perform a nonlinear combination operation on the second encoded vector, thereby generating feature information with stronger expressive power, which helps to improve the modeling accuracy of the pre-established validation model for the complex relationship between acupuncture treatment and patient constitution and gastrointestinal diseases.

[0047] The second nonlinear feature information is input into the second feature fusion layer to obtain the second fused feature information. In one embodiment of this application, the second feature fusion layer is used to perform feature concatenation on the second nonlinear feature information. The second feature fusion layer combines different features output by the second multilayer perceptron layer into more comprehensive feature information through the concatenation operation.

[0048] The second fused feature information is input into the second fully connected layer to obtain physical condition prediction information and gastrointestinal disease prediction results. In one embodiment of this application, the second fully connected layer is used to map the second fused feature information to a second preset spatial dimension to obtain second preset spatial dimension features; and to activate the second preset spatial dimension features; the second preset spatial dimension is 128 dimensions, 256 dimensions, etc., and a soft activation function is used to activate the second preset spatial dimension features.

[0049] The physical fitness prediction information and the gastrointestinal disease prediction results are input into the second output layer to obtain the gap between the physical fitness prediction information and the patient's historical physical fitness information, as well as the gap between the gastrointestinal disease prediction results and the patient's historical diagnosis results. In one embodiment of this application, the second output layer is used to calculate the gap between the physical fitness prediction information and the patient's historical physical fitness information, and the gap between the gastrointestinal disease prediction results and the patient's historical diagnosis results. The gap between the physical fitness prediction information and the patient's historical physical fitness information is calculated using a cross-entropy loss function, and the gap between the gastrointestinal disease prediction results and the patient's historical diagnosis results is calculated using a cross-entropy loss function. After obtaining the gap between the physical condition prediction information and the patient's historical physical condition information, as well as the gap between the gastrointestinal disease prediction results and the patient's historical diagnosis results, the first loss function value is calculated according to formula (2). Based on the first loss function value, the parameters in the pre-established validation model are continuously adjusted so that the gastrointestinal disease prediction results are gradually close to the patient's historical diagnosis results. At the same time, the gap between the gastrointestinal disease prediction results and the patient's historical diagnosis results is also gradually close. The above process can fully explore the deep correlation between the patient's physical condition, gastrointestinal diseases and acupuncture plan in the second sample data, provide solid data support for the optimization of the pre-established validation model, and enhance the pre-established validation model's ability to process complex medical data through continuous iterative adjustment of parameters. It also provides a scientific basis for the personalized recommendation of subsequent acupuncture plans.

[0050] In one embodiment of this application, the process of filtering acupuncture treatment prediction results according to their effectiveness ranking to obtain the filtered acupuncture treatment plans includes: If the number of acupuncture patterns in the predicted acupuncture pattern is less than a preset threshold, the predicted acupuncture pattern is used as the selected acupuncture pattern. In one embodiment of this application, the preset threshold is set according to the actual situation; for example, the preset threshold is set to 1.

[0051] If the number of acupuncture treatment plans in the predicted acupuncture plan results is greater than or equal to a preset threshold, the effectiveness of the predicted acupuncture plan results is ranked in ascending order to obtain an ascending sorted sequence; and a preset number of acupuncture plans ranked lower in the ascending sorted sequence are used as selected acupuncture plans. Alternatively, the effectiveness of the predicted acupuncture plan results is ranked in descending order to obtain a descending sorted sequence; and a preset number of acupuncture plans ranked higher in the descending sorted sequence are used as selected acupuncture plans. In one embodiment of this application, the preset number is set according to the actual situation, and the selected acupuncture plans have a better therapeutic effect on the patient.

[0052] In one embodiment of this application, the process of determining the recommended acupuncture plan based on the comparison results of physical fitness verification information and patient physical fitness information, and the comparison results of gastrointestinal disease verification results and gastrointestinal disease judgment results, includes: If the physical fitness verification information includes the patient's physical fitness information, and the gastrointestinal disease verification result includes the gastrointestinal disease diagnosis result, then the selected acupuncture plan will be used as the recommended acupuncture plan. In one embodiment of this application, if the physical fitness verification information includes the patient's physical fitness information, and the gastrointestinal disease verification result includes the gastrointestinal disease diagnosis result, it indicates that the selected acupuncture plan can treat patients with both gastrointestinal disease diagnosis results and patient physical fitness information, and the treatment effect is good.

[0053] If the physical fitness verification information does not include the patient's physical fitness information, and / or the gastrointestinal disease verification result does not include the gastrointestinal disease judgment result, then no recommended treatment plan will be considered as the recommended acupuncture treatment plan. In one embodiment of this application, if the physical fitness verification information does not include the patient's physical fitness information, and / or the gastrointestinal disease verification result does not include the gastrointestinal disease judgment result, then it indicates that the selected acupuncture treatment plan is not suitable for treating patients with gastrointestinal disease judgment results and patient physical fitness information.

[0054] In one embodiment of this application, if there are multiple acupuncture treatment options, it is necessary to compare the physical condition verification information obtained from each selected acupuncture treatment option with the patient's physical condition information, and to compare the gastrointestinal disease verification results obtained from each selected acupuncture treatment option with the gastrointestinal disease judgment results. Based on the comparison results, the acupuncture treatment option suitable for the patient is determined. If there are multiple acupuncture treatment options suitable for the patient, the acupuncture treatment option with the best treatment effect is selected, or the acupuncture treatment option that is easiest to implement is selected.

[0055] In one embodiment of this application, the patient's physical condition information is verified through physical condition verification information, and the gastrointestinal disease judgment result is verified through gastrointestinal disease verification results. This ensures that the recommended acupuncture plan is highly matched with the patient's actual physical condition and gastrointestinal disease status. This dual verification mechanism not only improves the accuracy of the recommended plan, but also enhances the safety of the treatment process.

[0056] In one embodiment of this application, before obtaining patient symptom information and patient physical condition information, the method for recommending acupuncture treatment plans for functional gastrointestinal disorders based on big data further includes: Patient symptom information and patient physical condition information can be input via an interactive device, or patient historical diagnosis information can be input via an interactive device, and patient symptom information and patient physical condition information can be extracted from the patient historical diagnosis information; the patient historical diagnosis information includes: patient diagnosis information and patient treatment information. In one embodiment of this application, the process of extracting patient symptom information and patient physical condition information from the patient historical diagnosis information is performed by field matching, or it can be performed by a matching model based on machine learning, without specific limitation.

[0057] Figure 3 This is a schematic diagram illustrating the structure of a pre-established prediction model, as shown in an exemplary embodiment of this application. Figure 3 As shown, the pre-established prediction model includes: a first input layer, a first multilayer perceptron layer, a first attention mechanism layer, a first feature fusion layer, a first fully connected layer, and a first output layer. The first input layer is used to perform normalization and encoding operations on the patient's historical diagnosis results, the patient's historical physical condition information, the acupuncture plan used by the patient, and the treatment effect of the acupuncture plan used by the patient. The first multilayer perceptron layer is used to perform nonlinear feature extraction and feature dimension transformation on the first encoded vector. The first attention mechanism layer is used to calculate the attention weights between features in the first encoded vector. The first feature fusion layer is used to perform feature concatenation and feature dimension transformation on the first nonlinear feature information and attention weight information. The first fully connected layer is used to map the first fused feature information to a first preset spatial dimension to obtain the first preset spatial dimension features and activate the first preset spatial dimension features. The first output layer is used to calculate the difference between the predicted acupuncture plan and the acupuncture plan used by the patient, and the difference between the predicted effect of the acupuncture plan and the treatment effect of the acupuncture plan used by the patient.

[0058] Figure 4 This is a schematic diagram illustrating the structure of a pre-established verification model as shown in an exemplary embodiment of this application, such as... Figure 4As shown, the pre-established verification model includes: a second input layer, a second multilayer perceptron layer, a second feature fusion layer, a second fully connected layer, and a second output layer. The second input layer is used to perform normalization and encoding operations on the second sample data. The second multilayer perceptron layer is used to perform nonlinear feature extraction and feature dimension transformation on the second encoded vector. The second feature fusion layer is used to perform feature concatenation on the second nonlinear feature information. The second fully connected layer is used to map the second fused feature information to a second preset spatial dimension to obtain the second preset spatial dimension features and to activate the second preset spatial dimension features. The second output layer is used to calculate the gap between the physical condition prediction information and the patient's historical physical condition information, as well as the gap between the gastrointestinal disease prediction results and the patient's historical diagnosis results.

[0059] The following describes an embodiment of the apparatus described in this application, which can be used to execute the method for recommending acupuncture protocols for functional gastrointestinal disorders based on big data, as described in the above embodiments of this application. For details not disclosed in the apparatus embodiments of this application, please refer to the embodiments of the method for recommending acupuncture protocols for functional gastrointestinal disorders based on big data described in the above embodiments of this application.

[0060] Figure 5 This is a block diagram illustrating an exemplary embodiment of the present application of a big data-based acupuncture protocol recommendation system for functional gastrointestinal diseases.

[0061] like Figure 5 As shown, this exemplary big data-based acupuncture protocol recommendation system 500 for functional gastrointestinal disorders includes: The information collection module 501 is used to acquire patient symptom information and patient physical condition information.

[0062] The disease judgment module 502 is used to input patient symptom information into the gastrointestinal disease judgment model to obtain the gastrointestinal disease judgment result.

[0063] The treatment plan prediction module 503 is used to input the patient's physical condition information and the judgment results of gastrointestinal diseases into the acupuncture treatment plan prediction model to obtain the acupuncture treatment plan prediction results and the effect of the acupuncture treatment plan prediction results.

[0064] The scheme screening module 504 is used to rank the acupuncture scheme prediction results according to their effectiveness, and then screen the acupuncture scheme prediction results to obtain the screened acupuncture schemes.

[0065] The scheme verification module 505 is used to input the selected acupuncture schemes into the acupuncture scheme verification model to obtain the constitution verification information and gastrointestinal disease verification results.

[0066] The information comparison module 506 is used to determine the recommended acupuncture plan based on the comparison results of the physical condition verification information and the patient's physical condition information, as well as the comparison results of the gastrointestinal disease verification results and the gastrointestinal disease judgment results.

[0067] In one embodiment of this application, patient symptom information includes symptoms such as stomach pain, acid reflux, bloating, belching, diarrhea, and constipation. Patient constitution information covers types from traditional Chinese medicine constitution classifications, including balanced constitution, qi deficiency constitution, yang deficiency constitution, yin deficiency constitution, phlegm-dampness constitution, damp-heat constitution, blood stasis constitution, qi stagnation constitution, and special constitution. Patient symptom and constitution information can be obtained through health questionnaires self-filled by patients on interactive devices or electronic terminals, or through preliminary detection and data collection using advanced intelligent medical equipment, followed by verification and refinement by professional medical personnel.

[0068] In one embodiment of this application, the gastrointestinal disease assessment model is a functional gastrointestinal disease diagnostic model built based on the Rome criteria (e.g., the Rome IV criteria).

[0069] In one embodiment of this application, the acupuncture plan prediction model is obtained by training a pre-established prediction model with first sample data. The first sample data includes: the patient's historical diagnosis results, the patient's historical physical condition information, the acupuncture plan used by the patient, the treatment effect of the acupuncture plan used by the patient, the correspondence between the combination of the patient's historical diagnosis results and the patient's historical physical condition information and the acupuncture plan used by the patient, and the correspondence between the acupuncture plan used by the patient and the treatment effect of the acupuncture plan used by the patient. The pre-established prediction model includes: a first input layer, a first multilayer perceptron layer, a first attention mechanism layer, a first feature fusion layer, a first fully connected layer, and a first output layer. The process of training the pre-established prediction model with the first sample data to obtain the acupuncture plan prediction model can fully learn the complex relationship between the patient's physical condition, gastrointestinal diseases, acupuncture plan, and the effect of acupuncture plan use, as well as how these relationships affect the selection of acupuncture plan, thereby improving the accuracy and reliability of acupuncture plan prediction and providing patients with more personalized and accurate acupuncture plan recommendation results.

[0070] In one embodiment of this application, the acupuncture pattern prediction result can be multiple acupuncture patterns or a single acupuncture pattern.

[0071] In one embodiment of this application, the effectiveness of the acupuncture pattern prediction results can be evaluated using criteria such as excellent, good, average, and poor, or by evaluation scores (e.g., 60, 70, and 80 points). If there is only one acupuncture pattern in the prediction results, it is used as the selected acupuncture pattern; if there are multiple acupuncture patterns in the prediction results, the process of selecting the acupuncture pattern with better effectiveness is to select the acupuncture pattern from the prediction results.

[0072] In one embodiment of this application, the acupuncture protocol verification model is obtained by training a pre-established verification model with second sample data. The second sample data includes: the patient's historical diagnosis results, the patient's historical physical condition information, the acupuncture protocol used by the patient, and the correspondence between the combination of the acupuncture protocol used by the patient and the patient's historical diagnosis results and historical physical condition information. The pre-established verification model includes: a second input layer, a second multilayer perceptron layer, a second feature fusion layer, a second fully connected layer, and a second output layer. The process of training the pre-established verification model with the second sample data to obtain the acupuncture protocol verification model can fully learn and screen the complex relationship between acupuncture protocols and the patient's physical condition and gastrointestinal diseases, as well as screen the applicability and effectiveness of acupuncture protocols in terms of the patient's physical condition and gastrointestinal diseases. This ensures that the recommended acupuncture protocol is not only theoretically feasible, but also achieves the expected therapeutic effect in practical application, thereby further enhancing the reliability and practicality of the acupuncture protocol recommendation results.

[0073] In one embodiment of this application, the acupuncture treatment plan is verified based on the comparison results of physical fitness verification information and patient physical fitness information, as well as the comparison results of gastrointestinal disease verification results and gastrointestinal disease judgment results, thereby improving the accuracy and effectiveness of the acupuncture treatment plan recommendation results.

[0074] In one embodiment of this application, a recommended acupuncture plan can be determined based on the patient's physical condition information and the results of the gastrointestinal disease assessment. This plan can be used by doctors or patients to determine the final acupuncture plan, which helps to avoid situations where unsuitable acupuncture plans are determined for patients due to the limitations of doctors' personal experience and technical level, resulting in unsatisfactory treatment effects.

[0075] It should be noted that the functional gastrointestinal disease acupuncture plan recommendation system based on big data provided in the above embodiments and the functional gastrointestinal disease acupuncture plan recommendation method based on big data provided in the above embodiments belong to the same concept. The specific operation methods of each module and unit have been described in detail in the method embodiments and will not be repeated here. In practical applications, the functional gastrointestinal disease acupuncture plan recommendation system based on big data provided in the above embodiments can be assigned to different functional modules as needed, that is, the internal structure of the device can be divided into different functional modules to complete all or part of the functions described above. This is not a limitation here.

[0076] The above embodiments are merely illustrative of the principles and effects of this application and are not intended to limit this application. Any person skilled in the art can modify or alter the above embodiments without departing from the spirit and scope of this application. Therefore, all equivalent modifications or alterations made by those skilled in the art without departing from the spirit and technical concept disclosed in this application should still be covered by the claims of this application.

Claims

1. A method for recommending acupuncture treatment plans for functional gastrointestinal disorders based on big data, characterized in that: The method includes: Obtain patient symptom information and patient physical condition information; The patient's symptom information is input into the gastrointestinal disease diagnosis model to obtain the gastrointestinal disease diagnosis result; The patient's physical condition information and the gastrointestinal disease diagnosis results are input into the acupuncture plan prediction model to obtain the acupuncture plan prediction results and the effect of the acupuncture plan prediction results; the acupuncture plan prediction model is obtained by training a pre-established prediction model with the first sample data; Based on the ranking of the usage effects of the predicted acupuncture treatment plan, the predicted acupuncture treatment plan is filtered to obtain the filtered acupuncture treatment plan. The selected acupuncture plan is input into the acupuncture plan verification model to obtain constitution verification information and gastrointestinal disease verification results; the acupuncture plan verification model is obtained by training a pre-established verification model with second sample data; Based on the comparison results of the physical condition verification information and the patient's physical condition information, as well as the comparison results of the gastrointestinal disease verification results and the gastrointestinal disease judgment results, the recommended acupuncture plan is determined.

2. The method for recommending acupuncture protocols for functional gastrointestinal disorders based on big data according to claim 1, characterized in that, The process of training the pre-established prediction model using the first sample data to obtain the acupuncture scheme prediction model includes: The first sample data is input into the pre-established prediction model to obtain the predicted acupuncture plan and the predicted effect of the acupuncture plan. The first sample data includes: the patient's historical diagnosis results, the patient's historical physical condition information, the acupuncture plan used by the patient, the treatment effect of the acupuncture plan used by the patient, the correspondence between the combination of the patient's historical diagnosis results and the patient's historical physical condition information and the acupuncture plan used by the patient, and the correspondence between the acupuncture plan used by the patient and the treatment effect of the acupuncture plan used by the patient. With the goal of minimizing the gap between the predicted acupuncture plan and the acupuncture plan used by the patient, and the gap between the effect of the predicted acupuncture plan and the treatment effect of the acupuncture plan used by the patient, the parameters in the pre-established prediction model are adjusted to obtain the acupuncture plan prediction model.

3. The method for recommending acupuncture protocols for functional gastrointestinal disorders based on big data according to claim 2, characterized in that, The difference between the predicted acupuncture plan and the acupuncture plan used by the patient, and the difference between the effect of the predicted acupuncture plan and the treatment effect of the acupuncture plan used by the patient, are characterized by a loss function, the expression of which is as follows: ， in, Represents the loss function. The weights represent the cross-entropy loss. Represents cross-entropy loss, This indicates the weight of the mean squared error loss. This represents the mean squared error loss. This represents the total number of categories of acupuncture treatments. Indicates the patient's use of the first Category tags corresponding to each acupuncture treatment plan Indicates the first Predicted probabilities of acupuncture treatments Represents the total number of samples. Indicates the first The therapeutic effects of acupuncture on patients in a sample Indicates the first The predicted effectiveness of the acupuncture treatment obtained from the predicted sample was, in the ... The therapeutic effects of acupuncture on patients in the sample, and the treatment outcomes after the first... The effectiveness of the predicted acupuncture treatment plan obtained from each sample prediction is characterized by either a coding or numerical form. .

4. The method for recommending acupuncture protocols for functional gastrointestinal disorders based on big data according to claim 2, characterized in that, If the pre-established prediction model includes: a first input layer, a first multilayer perceptron layer, a first attention mechanism layer, a first feature fusion layer, a first fully connected layer, and a first output layer, then the process of inputting the first sample data into the pre-established prediction model to obtain the predicted acupuncture scheme and the effect of the predicted acupuncture scheme includes: The first sample data is input into the first input layer to obtain the first encoding vector; the first input layer is used to perform normalization and encoding operations on the patient's historical diagnosis results, the patient's historical physical condition information, the acupuncture plan used by the patient, and the treatment effect of the acupuncture plan used by the patient. The first encoded vector is input into the first multilayer perceptron layer to obtain the first nonlinear feature information; the first multilayer perceptron layer is used to perform nonlinear feature extraction and feature dimension transformation on the first encoded vector; The first encoding vector is input into the first attention mechanism layer to obtain attention weight information; the first attention mechanism layer is used to calculate the attention weights between features in the first encoding vector. The first nonlinear feature information and the attention weight information are input into the first feature fusion layer to obtain the first fused feature information; the first feature fusion layer is used to perform feature concatenation and feature dimension transformation on the first nonlinear feature information and the attention weight information. The first fused feature information is input into the first fully connected layer to obtain the predicted acupuncture scheme and the effect of the predicted acupuncture scheme; the first fully connected layer is used to map the first fused feature information to a first preset spatial dimension to obtain the first preset spatial dimension feature; and to activate the first preset spatial dimension feature; The predicted acupuncture plan and its effect are input into the first output layer to obtain the difference between the predicted acupuncture plan and the acupuncture plan used by the patient, and the difference between the effect of the predicted acupuncture plan and the treatment effect of the patient using the acupuncture plan; the first output layer is used to calculate the difference between the predicted acupuncture plan and the acupuncture plan used by the patient, and the difference between the effect of the predicted acupuncture plan and the treatment effect of the patient using the acupuncture plan.

5. The method for recommending acupuncture protocols for functional gastrointestinal disorders based on big data according to claim 1, characterized in that, The process of training the pre-established verification model using second sample data to obtain the acupuncture scheme verification model includes: The second sample data is input into the pre-established validation model to obtain constitution prediction information and gastrointestinal disease prediction results; the second sample data includes: patient's historical diagnosis results, patient's historical constitution information, acupuncture program used by the patient, and the correspondence between the acupuncture program used by the patient and the combination of the patient's historical diagnosis results and the patient's historical constitution information; With the goal of minimizing the gap between the predicted physical constitution information and the patient's historical physical constitution information, as well as the gap between the predicted gastrointestinal disease results and the patient's historical diagnostic results, the parameters in the pre-established validation model are adjusted to obtain the validation model of the acupuncture scheme.

6. The method for recommending acupuncture protocols for functional gastrointestinal disorders based on big data according to claim 5, characterized in that, If the pre-established validation model includes: a second input layer, a second multilayer perceptron layer, a second feature fusion layer, a second fully connected layer, and a second output layer, then the process of inputting the second sample data into the pre-established validation model to obtain physical fitness prediction information and gastrointestinal disease prediction results includes: The second sample data is input into the second input layer to obtain the second encoding vector; the second input layer is used to perform normalization and encoding operations on the second sample data. The second encoded vector is input into the second multilayer perceptron layer to obtain the second nonlinear feature information; the second multilayer perceptron layer is used to perform nonlinear feature extraction and feature dimension transformation on the second encoded vector; The second nonlinear feature information is input into the second feature fusion layer to obtain the second fused feature information; the second feature fusion layer is used to perform feature concatenation on the second nonlinear feature information; The second fused feature information is input into the second fully connected layer to obtain the constitution prediction information and the gastrointestinal disease prediction result; the second fully connected layer is used to map the second fused feature information to a second preset spatial dimension to obtain the second preset spatial dimension feature; and to activate the second preset spatial dimension feature; The physical condition prediction information and the gastrointestinal disease prediction results are input into the second output layer to obtain the gap between the physical condition prediction information and the patient's historical physical condition information, as well as the gap between the gastrointestinal disease prediction results and the patient's historical diagnosis results. The second output layer is used to calculate the gap between the physical condition prediction information and the patient's historical physical condition information, as well as the gap between the gastrointestinal disease prediction results and the patient's historical diagnosis results.

7. The method for recommending acupuncture protocols for functional gastrointestinal disorders based on big data according to any one of claims 1-6, characterized in that, The process of ranking the acupuncture treatment plans based on their predicted effectiveness and then filtering them to obtain the selected acupuncture treatment plans includes: If the number of acupuncture schemes in the acupuncture scheme prediction results is less than a preset number threshold, then the acupuncture scheme prediction results will be used as the selected acupuncture schemes. If the number of acupuncture schemes in the predicted acupuncture scheme results is greater than or equal to the preset number threshold, then the effectiveness of the predicted acupuncture scheme results is ranked in ascending order to obtain an ascending sorted sequence; and a preset number of acupuncture schemes ranked at the bottom in the ascending sorted sequence are used as the selected acupuncture schemes; or, the effectiveness of the predicted acupuncture scheme results is ranked in descending order to obtain a descending sorted sequence; and a preset number of acupuncture schemes ranked at the top in the descending sorted sequence are used as the selected acupuncture schemes.

8. The method for recommending acupuncture protocols for functional gastrointestinal disorders based on big data according to any one of claims 1-6, characterized in that, The process of determining the recommended acupuncture plan based on the comparison results of the physical fitness verification information and the patient's physical fitness information, and the comparison results of the gastrointestinal disease verification results and the gastrointestinal disease judgment results, includes: If the physical fitness verification information includes the patient's physical fitness information, and the gastrointestinal disease verification result includes the gastrointestinal disease judgment result, then the selected acupuncture plan will be used as the recommended acupuncture plan result. If the physical condition verification information does not include the patient's physical condition information, and / or the gastrointestinal disease verification result does not include the gastrointestinal disease judgment result, then no recommended plan will be taken as the recommended result of the acupuncture plan.

9. The method for recommending acupuncture protocols for functional gastrointestinal disorders based on big data according to any one of claims 1-7, characterized in that, Before obtaining patient symptom information and patient physical condition information, the method further includes: The patient's symptom information and physical condition information are input through an interactive device; Alternatively, the patient's historical diagnostic information can be input through the interactive device, and the patient's symptom information and physical condition information can be extracted from the patient's historical diagnostic information; the patient's historical diagnostic information includes: patient diagnostic information and patient treatment information.

10. A functional gastrointestinal disease acupuncture program recommendation system based on big data, characterized in that, include: The information collection module is used to acquire patient symptom information and patient physical condition information; The disease judgment module is used to input the patient's symptom information into the gastrointestinal disease judgment model to obtain the gastrointestinal disease judgment result. The treatment plan prediction module is used to input the patient's physical condition information and the judgment result of the gastrointestinal disease into the acupuncture treatment plan prediction model to obtain the acupuncture treatment plan prediction result and the effect of the acupuncture treatment plan prediction result; the acupuncture treatment plan prediction model is obtained by training a pre-established prediction model with the first sample data; The scheme filtering module is used to filter the acupuncture scheme prediction results according to the ranking of the usage effects of the acupuncture scheme prediction results, and obtain the filtered acupuncture schemes. The scheme verification module is used to input the selected acupuncture scheme into the acupuncture scheme verification model to obtain physical constitution verification information and gastrointestinal disease verification results; the acupuncture scheme verification model is obtained by training a pre-established verification model with second sample data; The information comparison module is used to determine the recommended acupuncture plan based on the comparison results between the physical fitness verification information and the patient's physical fitness information, as well as the comparison results between the gastrointestinal disease verification results and the gastrointestinal disease judgment results.

Images