Traditional Chinese medicine holographic data health suggestion generation method based on neural network

Abstract

According to the method, multi-modal data fusion is adopted, various holographic features of tongue images, face images, ear images, palm prints and sweat glands are comprehensively analyzed, each holographic feature corresponds to the traditional Chinese medicine second-level physique, the weight of each holographic feature is combined, the score of the second-level physique is calculated, and the score of the second-level physique is obtained. Three most possible second-level traditional Chinese medicine constitutions are obtained according to the maximum probability principle, and meanwhile personalized suggestions are given. The invention relates to the field of health service management, the final traditional Chinese medicine physique is obtained through model large fusion analysis, and the accuracy of model physique fusion is further improved, so that the suggestions provided for the user better meet the actual needs of the user, and the reliability of the whole fusion method is improved.

Description

Technical Field

[0001] The present invention discloses a method for using multimodal data fusion, comprehensively analyzing multiple holographic features such as tongue image, face image, ear image, palm print, sweat gland, etc., corresponding each holographic feature to a secondary constitution of traditional Chinese medicine, combining the weight of each holographic feature, calculating the score of the secondary constitution obtained, and obtaining the most likely three secondary constitutions of traditional Chinese medicine according to the principle of maximum probability, while giving personalized suggestions. The present invention belongs to the field of health service management. Background Art

[0002] Regarding the integration of TCM holograms in existing systems, the types of holograms are different from those in our system. Compared with other systems, our system integrates more holograms and comprehensively analyzes multiple holographic features, which can provide a more comprehensive understanding of the user's health status, thereby improving the accuracy and reliability of diagnosis.

[0003] Related technologies: Multi-classifier neural network model (CCNN), Panda, and weight calculation method are all mature technical methods.

[0004] Reference materials: "Classification and Determination of Constitution in Traditional Chinese Medicine", "Huangdi Neijing", "Diagnosis in Traditional Chinese Medicine", "Constitution in Traditional Chinese Medicine", and "Basic Theory of Traditional Chinese Medicine".

[0005] The key lies in how to integrate the whole picture of TCM (tongue image, ear image, palm print, facial image, sweat gland), comprehensively analyze the TCM constitution, and give personalized suggestions. Summary of the invention

[0006] Although the application of TCM holography combined with computer technology has made certain progress, there are still some shortcomings in practical applications. These shortcomings are mainly reflected in the following aspects: Data standardization issues. TCM data collection and collation often lack unified standards, which makes it difficult to effectively integrate and compare data from different sources. Data privacy and security issues. As more and more personal health data are collected for TCM research and diagnosis and treatment, protecting patient privacy and data security has become an important issue. The existing system does not integrate and analyze the TCM constitution corresponding to the five holograms of TCM (tongue image, facial image, ear image, palm print, sweat gland), and the resulting constitution characteristics lack comprehensiveness.

[0007] Compared with other systems, the present invention adopts multimodal data fusion, and gives personalized suggestions by comprehensively analyzing multiple holographic features of tongue image, face image, ear image, palm print, sweat gland, and weight analysis of each holographic feature; using big data technology and machine learning algorithms, a large number of TCM cases can be analyzed to discover the laws and patterns of diseases and provide support for clinical decision-making. The introduction of computer technology has promoted the cross-integration of TCM and modern science and technology, and provided a new way for the modernization and internationalization of TCM. Through standardized data collection and processing procedures, the consistency and reliability of each diagnosis are ensured. Comply with national and international data protection regulations to ensure the legal use and transmission of data. The system that integrates and analyzes tongue image, face image, ear image, palm print, and sweat gland to obtain TCM constitution has shown significant advantages in many aspects. In the future, with the continuous advancement of technology, the application prospects in this field will be broader.

[0008] The system includes:

[0009] Feature extraction module: This module aims to extract the features of four TCM holograms (tongue image, ear image, palm print, and facial image); sweat glands can directly derive TCM constitution through the sweat gland system, and a patent application has been submitted for the sweat gland system.

[0010] Characteristic constitution correspondence module: This module aims to correspond the features extracted from the four holograms (tongue image, ear image, palm print, and face image) to the first-level constitution of traditional Chinese medicine (the correspondence table of each holographic feature and constitution has been placed in other certification documents). According to the features extracted from the previous module, the weight of the hologram corresponds to the corresponding characteristic constitution table, and the first-level constitution of traditional Chinese medicine with a score is obtained. (The sweat glands are directly obtained through the sweat gland system and combined with the weight to obtain three first-level constitutions with scores)

[0011] TCM constitution refinement module: This module aims to refine the first-level constitution of TCM. The first-level constitution with scores obtained in the previous module is used to obtain the second-level constitution with scores from Table 1.

[0012] Fusion module: Calculate the score of each secondary constitution and output the top three secondary constitutions; and derive personalized clothing, food, housing, transportation and department recommendations from Table 2.

[0013] Feedback module: Feedback the results to the front end. See System Architecture for details. Figure 1 . System flow diagram Figure 2 . Table 1: Correspondence between first-level constitution and second-level constitution. Second level constitution recommendation comparison table 2. Correspondence table between first-level constitution and second-level constitution Table 1 Comparison table of recommended secondary constitutions Table 2 DETAILED DESCRIPTION

[0014] Step one: Integrate the entire system into the company’s health management system.

[0015] Step 2: Obtain features extracted from the four major holograms of traditional Chinese medicine (tongue image, ear image, palm print, and facial image).

[0016] Step 3: Input the features extracted from each TCM hologram into their respective neural network models (there is no need to extract features from sweat glands, as the first-level constitution can be directly derived through the sweat gland system). According to the principle of maximum probability, the three most likely first-level constitutions are obtained for each hologram, and a list of first-level constitutions most likely grouped according to the four major holograms is obtained.

[0017] Step 4: Count the extracted constitution list according to TCM grouping and correspond them to the secondary constitutions.

[0018] Step 4: Combine the numerical value of each TCM diagnosis with the weight of the TCM diagnosis to calculate the score for each constitution.

[0019] Step 5: Output the three second-level physiques with the highest scores.

[0020] Step 7: Read the second-level physical constitution recommendation comparison table 2 to obtain recommendations on food, clothing, housing, transportation and departments.

[0021] Step 8: Return the result to the front end. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] System Architecture Figure 1 .

[0023] System flow diagram Figure 2 .

Claims

1. The present invention uses a neural network model (CCNN), Panda, and a weight calculation method to fuse and analyze the TCM hologram (tongue image, ear image, palm print, facial image, sweat glands), and provides personalized suggestions for the current physical condition of the service recipient.

2. Establish a method for changing physical constitution and diet, and generate daily life guidance suggestions (i.e. Table 2).

Images