A multimodal sensing-based integrated traditional Chinese and Western medicine diagnostic and treatment method and system
By using multimodal sensor arrays and model fusion technology, the problems of fragmented data collection and subjective diagnosis in traditional Chinese medicine diagnosis and treatment have been solved, realizing the bidirectional integration of traditional Chinese and Western medicine diagnosis and the closed loop of automated diagnosis and treatment, and ensuring safety in unattended scenarios.
Patent Information
- Application Number
- CN202611258132.3
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-08-18
- Publication Date
- 2026-11-10
AI Technical Summary
Existing TCM diagnosis and treatment techniques suffer from problems such as fragmented data collection, strong subjectivity in diagnosis, separation of TCM and Western medicine diagnosis, lack of automated closed-loop treatment plans, and lack of safety backup mechanisms in unattended scenarios.
A multimodal sensor array is used to simultaneously collect TCM diagnostic and treatment characteristic data and Western medicine basic physical signs data. Cross-validation and dynamic weight fusion are performed through a multi-feature fusion syndrome differentiation model and a Western medicine intelligent diagnostic model to output diagnostic and treatment conclusions and automatically match intervention plans. At the same time, safety verification and risk level determination are carried out.
It has achieved multi-dimensional synchronous collection and standardization of TCM diagnosis and treatment data, improved the objectivity and accuracy of syndrome differentiation, realized bidirectional integrated decision-making of TCM and Western medicine diagnosis, provided a complete automated diagnosis and treatment closed loop, and ensured safety in unattended scenarios.
Smart Images

Figure CN123072676A_ABST
Abstract
Description
Technical Field
[0001] This application relates to the field of intelligent diagnostic technology in traditional Chinese and Western medicine, specifically to a diagnostic and treatment method and system based on multimodal perception that integrates traditional Chinese and Western medicine. Background Technology
[0002] Traditional Chinese medicine pulse diagnosis centers on the nine pulse positions across the three wrist regions (cun, guan, chi). By sensing the pulse's depth, speed, and smoothness, it assesses the functional state of the internal organs, the state of qi and blood, and the nature of pathogenic factors, serving as a core basis for TCM syndrome differentiation and treatment. Facial and tongue diagnosis, on the other hand, observe the patient's complexion and tongue coating to aid in determining the deficiency or excess of internal organs and the depth of pathogenic factors. Traditional TCM diagnosis relies on physicians gathering information through the four diagnostic methods of observation, auscultation and olfaction, inquiry, and palpation, comprehensively applying theories such as the Eight Principles of Syndrome Differentiation and Zang-Fu Syndrome Differentiation to determine syndromes and prescribe medication.
[0003] In recent years, with the development of sensing technology, artificial intelligence, and big data technology, research on the modernization of traditional Chinese medicine (TCM) diagnosis and treatment has made some progress. In the area of digital pulse diagnosis, researchers have developed pulse image acquisition devices based on technologies such as pressure sensors and ultrasound detection, attempting to convert pulse characteristics into quantitative parameters. In the area of intelligent observation diagnosis, equipment such as tongue and face diagnostic instruments has achieved standardized acquisition of tongue and face images and preliminary AI-assisted recognition. Some research institutions are also exploring the integration and analysis of TCM's four diagnostic methods with Western medical laboratory indicators to construct an integrated TCM-Western medicine auxiliary diagnostic model, or attempting to apply infrared thermal imaging technology to TCM's differentiation of cold and heat patterns.
[0004] Despite numerous explorations in modernizing TCM diagnosis and treatment using existing technologies, the following major shortcomings still exist:
[0005] I. Data collection for the four diagnostic methods in Traditional Chinese Medicine is fragmented and lacks multimodal synchronous acquisition capabilities.
[0006] Current TCM diagnostic equipment largely focuses on a single diagnostic method or a single physical dimension of detection—pulse diagnostic instruments only collect pulse pressure waveforms, tongue diagnostic instruments only collect tongue images, and infrared thermal imagers only collect body surface temperature distribution. The separate collection and independent analysis of data from each diagnostic method violates the core diagnostic logic of TCM's "integrated diagnosis and treatment based on the four diagnostic methods." Furthermore, existing research is mostly limited to local observation (such as tongue or facial appearance), making it difficult to achieve a comprehensive perception of the multidimensional dynamic information of the entire body's appearance and demeanor, thus failing to meet the comprehensive needs of TCM's holistic diagnosis for multidimensional information.
[0007] Second, traditional Chinese medicine diagnosis relies on human experience, is highly subjective, and has a low degree of standardization.
[0008] In current technologies, pulse diagnosis still relies heavily on the physician's personal tactile experience, presenting a challenge in its transmission due to the difficulty of accurate pulse diagnosis. Furthermore, there is a lack of unified standards for the objective recording and standardized interpretation of pulse information. Similarly, facial and tongue diagnosis are limited by the physician's personal experience and observation environment, resulting in subjective and uncertain diagnoses. While some AI-powered constitution identification products have been deployed at the grassroots level, their diagnostic logic is largely based on knowledge graphs or questionnaires, failing to achieve true deep integration of data from the four diagnostic methods (inspection, auscultation, inquiry, and palpation), thus limiting diagnostic accuracy.
[0009] Third, the separation between TCM syndrome differentiation and Western medical diagnosis lacks a dual-dimensional integrated decision-making mechanism.
[0010] In current technologies, some research on the integration of traditional Chinese medicine (TCM) and Western medicine remains at the level of simple statistical correlation between Western medical test indicators and TCM syndrome types, lacking parallel reasoning and cross-validation mechanisms between TCM syndrome differentiation models and Western medical diagnostic models. While a few studies have attempted to integrate tongue and pulse diagnosis data with Western medical test indicators to construct diagnostic models, these are mostly single-model designs for specific diseases, failing to form a universal decision-making framework for parallel reasoning, cross-validation, and dynamic weight fusion between TCM and Western medicine. When contradictions arise between TCM syndrome conclusions and Western medical indicator interpretations, existing systems lack the ability to correct conflicts and automatically identify difficult cases.
[0011] Fourth, the diagnostic and treatment plans are limited and lack a complete automated closed loop from diagnosis to treatment.
[0012] Existing intelligent TCM diagnostic and treatment devices mostly remain at the level of detection and assessment, ending the process after outputting a constitution identification report or health advice. They lack the ability to automatically match personalized intervention plans such as TCM prescriptions and spectrum therapy parameters based on the syndrome differentiation conclusions. Although there are acupuncture diagnostic and treatment instruments with therapeutic functions, the matching of their treatment plans still requires manual selection, failing to achieve a fully automated integrated closed loop of "syndrome differentiation-prescription-therapy parameter matching".
[0013] V. Lack of security verification and manual backup mechanisms in unattended scenarios
[0014] Existing AI-assisted diagnosis and treatment systems for Traditional Chinese Medicine (TCM) generally lack automatic interception mechanisms for high-risk cases, special populations, and prescription safety. In clinical practice, there have been reports of patients delaying their treatment due to over-reliance on AI consultations—AI cannot proactively inquire about key medical history information, cannot assume medical responsibility, and lacks a safety net mechanism for risk grading and cloud-based physician review. In particular, when it comes to contraindications in TCM prescriptions (such as the "Eighteen Incompatibilities" and "Nineteen Antagonisms") and medication safety for special populations, existing systems lack rigorous automatic verification and manual intervention triggering mechanisms.
[0015] In conclusion, existing technologies have not yet solved the fundamental problems of "subjectivity, fragmentation, lack of quantification, and difficulty in closing the loop" in TCM diagnosis and treatment. Summary of the Invention
[0016] To address these issues, this application provides a multimodal perception-based integrated traditional Chinese and Western medicine diagnosis and treatment method and system to solve the problems of fragmented TCM diagnosis and treatment data collection, strong subjectivity in syndrome differentiation, separation of TCM and Western medicine diagnosis, lack of automated closed-loop treatment plans, and lack of safety fallback mechanisms in unattended scenarios in existing technologies.
[0017] To achieve the above objectives, this application provides the following technical solution:
[0018] Firstly, a multimodal perception-based integrated traditional Chinese and Western medicine diagnostic and treatment method includes:
[0019] Step 1: In a closed, soundproof, and light-proof environment, simultaneously collect TCM diagnostic and treatment characteristic data and Western medicine basic physical signs data of the target object through a multimodal sensor array; wherein, the TCM diagnostic and treatment characteristic data includes: four diagnostic methods characteristic data and meridian heat state characteristic data;
[0020] Step 2: Perform data preprocessing on the TCM diagnostic and treatment feature data and the Western medicine basic physical sign data respectively to obtain TCM syndrome feature vector and Western medicine physical sign feature vector;
[0021] Step 3: Input the TCM syndrome feature vector into the pre-trained multi-feature fusion syndrome differentiation model and output the TCM syndrome determination result;
[0022] Step 4: Input the Western medicine physical signs feature vector into the pre-trained Western medicine intelligent diagnostic model and output the Western medicine disease screening results;
[0023] Step 5: Cross-validate and dynamically weight the TCM syndrome determination results with the Western medicine symptom screening results, and output the fused diagnosis and treatment conclusion;
[0024] Step 6: Based on the fused diagnostic conclusions, automatically match and output an intervention plan; the intervention plan includes: traditional Chinese medicine prescription, spectrum therapy parameters, and health management suggestions;
[0025] Step 7: Perform a security check on the intervention plan and determine the risk level; if the determination result is a high-risk level, involves special populations or special medications, automatically intercept and push it to the cloud-based practicing physician terminal for manual review.
[0026] Optionally, in step 1, the four diagnostic features data include tongue appearance features data, facial appearance features data, pulse appearance features data, and breath features data;
[0027] The pulse characteristic data is acquired through sonar pulse detection technology and high-intensity cold light transmission imaging technology. The sonar pulse detection technology generates a continuous dynamic pulse waveform diagram that reflects the pulse position, pulse rate, pulse shape, and pulse strength by emitting high-frequency ultrasound waves to the cun, guan, and chi positions of the wrist and receiving the echoes. The high-intensity cold light transmission imaging technology uses low-heat, high-penetration cold light to irradiate the peripheral blood vessels of the wrist and limbs in a directional manner, and images the subcutaneous radial artery and microcirculation blood flow status in real time, extracting data on blood flow filling, blood flow velocity, changes in blood vessel diameter, and peripheral blood flow perfusion status.
[0028] The meridian thermal characteristic data is obtained through infrared thermal spectrum imaging technology: by collecting the thermal radiation spectrum of the corresponding reflection areas of meridians, acupoints and internal organs on the human body surface, a global thermal spectrum distribution map is generated, and the temperature value, thermal radiation intensity and cold-heat cross-sectional features of each zone are extracted.
[0029] Optionally, in step 2, the data preprocessing includes:
[0030] Environmental calibration: Based on the temperature, humidity and light parameters of the collected environment, automatically correct the color deviation of the tongue feature data and the facial feature data, as well as the temperature and humidity deviation of the breath feature data;
[0031] Outlier removal: Transient outlier data is removed using the 3σ criterion;
[0032] Normalization processing: The parameters of different dimensions in the TCM diagnostic and treatment feature data and the Western medicine basic physical sign data are standardized and unified to generate the TCM syndrome feature vector and the Western medicine physical sign feature vector.
[0033] Optionally, in step 5, the cross-validation specifically includes:
[0034] When the results of TCM syndrome determination and Western medicine disease screening are mutually corroborated, the confidence of the final diagnosis is increased.
[0035] When there is a conflict between the TCM syndrome determination results and the Western medicine disease screening results, a weighted correction is performed based on the preset disease type and the confidence level of the indicators, and low confidence determinations are eliminated.
[0036] When the TCM syndrome determination result seriously conflicts with the Western medicine disease screening result, it is automatically marked as a difficult case and triggers cloud-based manual review.
[0037] Optionally, in step 5, the dynamic weight fusion specifically includes:
[0038] The weights of TCM syndrome differentiation and Western medicine diagnosis are dynamically allocated according to different disease types:
[0039] Sub-health and constitution conditioning category: Traditional Chinese medicine diagnosis accounts for 60% of the weight, while Western medicine diagnosis accounts for 40% of the weight.
[0040] Chronic diseases and mild organic lesions: Western medicine diagnosis accounts for 60% of the weight, and traditional Chinese medicine syndrome differentiation accounts for 40% of the weight.
[0041] Critical and high-risk illnesses: Western medicine risk assessment is the core, with traditional Chinese medicine only serving as a supplementary reference.
[0042] Optionally, in step 6, the matching method for the traditional Chinese medicine prescription is as follows:
[0043] Based on the fused diagnostic and treatment conclusions, the pre-built compliant traditional Chinese medicine prescription matching model and the database of monarch-minister-assistant-guide compatibility rules are invoked to automatically screen suitable prescriptions and perform the Eighteen Incompatibilities and Nineteen Antagonisms Contraindication Verification. The dosage is standardized and adjusted according to the target subject's age, constitution, allergy history, and severity of illness.
[0044] Optionally, in step 6, the matching method for the spectral therapy parameters is as follows:
[0045] Based on the TCM syndrome differentiation results, the far-infrared spectrum therapy mode is automatically matched:
[0046] Cold syndrome, Yang deficiency, rheumatic pain: Matched with far-infrared continuous warming mode;
[0047] Moist heat, local inflammation: Match short-pulse low-energy mode;
[0048] Qi and blood stagnation, meridian blockage: Matching pulse and continuous composite spectrum mode;
[0049] The spectrum therapy parameters include at least: irradiation site, working mode, output power, irradiation duration, and treatment frequency.
[0050] Optionally, in step 6, the method for generating the health management recommendations is as follows:
[0051] When the fused diagnosis and treatment conclusion includes Western medicine disease screening results, a Western medicine intervention plan is output. The Western medicine intervention plan includes at least dietary restrictions, exercise recommendations, indicator control standards, and a long-term follow-up plan.
[0052] When the integrated diagnosis and treatment conclusion includes the results of TCM syndrome determination, a personalized health care plan is output by combining the principles of seasonal health care, meridian flow, and body constitution matching. The personalized health care plan includes at least dietary therapy, acupoint massage guidance, work and rest adjustment plan, and exercise adaptation suggestions.
[0053] Optionally, in step 7, the security verification further includes:
[0054] Safety protection verification is performed on the spectrum therapy parameters: real-time monitoring of skin temperature at the irradiation site and irradiation distance; when the skin temperature exceeds the safety threshold or the irradiation distance exceeds the safety range, the therapy is automatically terminated and an alarm is triggered.
[0055] When the high-risk level is determined to be Level 3 high risk, the diagnosis and treatment data and reasoning logs are encrypted and stored on the blockchain throughout the entire process to achieve tamper-proof, full-process medical traceability.
[0056] Secondly, a multimodal perception-based integrated traditional Chinese and Western medicine diagnosis and treatment system includes:
[0057] The multimodal data acquisition module is used to synchronously acquire TCM diagnostic and treatment characteristic data and Western medicine basic physical signs data of the target object through a multimodal sensor array in a closed, soundproof, and light-proof environment; wherein, the TCM diagnostic and treatment characteristic data includes: four diagnostic methods characteristic data and meridian heat state characteristic data;
[0058] The data preprocessing module is used to preprocess the TCM diagnostic and treatment feature data and the Western medicine basic physical sign data respectively to obtain TCM syndrome feature vectors and Western medicine physical sign feature vectors.
[0059] The TCM intelligent syndrome differentiation module is used to input the TCM syndrome feature vector into a pre-trained multi-feature fusion syndrome differentiation model and output the TCM syndrome determination result.
[0060] The Western medicine intelligent diagnosis module is used to input the Western medicine vital sign feature vector into the pre-trained Western medicine intelligent diagnosis model and output the Western medicine disease screening results.
[0061] The integrated TCM and Western medicine decision-making module is used to cross-validate and dynamically weight the TCM syndrome determination results and the Western medicine disease screening results, and output the integrated diagnosis and treatment conclusion.
[0062] An intervention plan generation module is used to automatically match and output intervention plans based on the fused diagnostic and treatment conclusions; the intervention plans include: traditional Chinese medicine prescriptions, spectrum physiotherapy parameters, and health management recommendations;
[0063] The security verification and fallback module is used to verify the security of the intervention plan and determine the risk level. If the determination result is a high-risk level, involves special populations or special medications, it is automatically intercepted and pushed to the cloud-based physician terminal for manual review.
[0064] Compared with the prior art, this application has at least the following beneficial effects:
[0065] This application provides a multimodal sensing-based integrated traditional Chinese and Western medicine diagnostic and treatment method, including:
[0066] The system synchronously collects TCM diagnostic and treatment characteristic data and Western medicine basic physical sign data of the target object through a multimodal sensor array, and performs data preprocessing to obtain TCM syndrome feature vectors and Western medicine physical sign feature vectors. The TCM syndrome feature vectors are input into a pre-trained multi-feature fusion syndrome differentiation model to output TCM syndrome determination results. The Western medicine physical sign feature vectors are input into a pre-trained Western medicine intelligent diagnostic model to output Western medicine symptom screening results. The TCM syndrome determination results and Western medicine symptom screening results are cross-validated and dynamically weighted to output the fused diagnosis and treatment conclusion. Based on the fused diagnosis and treatment conclusion, an intervention plan is automatically matched and output. The intervention plan is subjected to safety verification and risk level determination. If the determination result is high-risk, involves special populations or special medications, it is automatically intercepted and pushed to the cloud-based physician terminal for manual review. This application systematically solves the problems of fragmented data collection, strong subjectivity in diagnosis, separation of traditional Chinese and Western medicine diagnosis, lack of automated closed-loop treatment plans, and lack of safety backup mechanism in unattended scenarios by using a complete closed loop of multimodal sensor array synchronous acquisition, dual-model parallel inference, cross-validation of traditional Chinese and Western medicine and dynamic weight fusion, automatic matching and output of intervention plans, risk classification interception and cloud-based manual backup. Attached Figure Description
[0067] To more intuitively illustrate the prior art and this application, exemplary drawings are provided below. It should be understood that the specific shapes and structures shown in the drawings should not generally be regarded as limiting conditions for implementing this application; for example, based on the technical concept disclosed in this application and the exemplary drawings, those skilled in the art are able to easily make conventional adjustments or further optimizations to the addition / reduction / classification, specific shapes, positional relationships, connection methods, size ratios, etc. of certain units (components).
[0068] Figure 1 This is a flowchart of a multimodal sensing-based integrated traditional Chinese and Western medicine diagnosis and treatment method provided in Embodiment 1 of this application. Detailed Implementation
[0069] The present application will be further described in detail below with reference to the accompanying drawings and specific embodiments.
[0070] In the description of this application: unless otherwise stated, "a plurality of" means two or more. The terms "first," "second," "third," etc., in this application are intended to distinguish the objects referred to and do not have any special meaning in terms of technical connotation (e.g., they should not be construed as an emphasis on importance or order). Expressions such as "including," "comprising," and "having" also mean "not limited to" (certain units, components, materials, steps, etc.).
[0071] The terms used in this application, such as "upper," "lower," "left," "right," and "middle," are generally used to indicate the general relative positional relationship for the purpose of intuitive understanding by referring to the accompanying drawings, and are not absolute limitations on the positional relationship in the actual product.
[0072] Example 1
[0073] Please see Figure 1 This embodiment provides a multimodal perception-based integrated traditional Chinese and Western medicine diagnosis and treatment method, including:
[0074] S1: In a closed, soundproof, and light-proof environment, the TCM diagnostic and treatment characteristic data and Western medicine basic physical signs data of the target object are collected synchronously through a multimodal sensor array; among them, the TCM diagnostic and treatment characteristic data include: four diagnostic methods characteristic data and meridian heat state characteristic data;
[0075] Specifically, the four diagnostic features include: tongue appearance features, facial appearance features, pulse appearance features, and breath features; the multimodal sensor array includes: tongue and facial diagnosis integrated machine, bionic pulse diagnosis instrument, breath sensor, millimeter wave non-contact vital sign monitoring device, and Western medicine vital sign acquisition terminals such as body temperature, heart rate, and blood oxygen.
[0076] Among them, the tongue and face diagnosis integrated machine is used to collect tongue image feature data and facial image feature data. During tongue diagnosis, the tongue and face diagnosis integrated machine is based on a deep learning image segmentation model to accurately segment the tongue body, tongue coating, tongue shape, and tongue state, and extract 28 core features such as tongue body color, tongue coating thickness, moisture, dryness, greasiness, teeth marks, and cracks. It automatically avoids interference from light color difference and oral shadows, and the feature recognition accuracy rate is ≥96%. During face diagnosis, the tongue and face diagnosis integrated machine uses facial key point positioning technology to extract 16 facial syndrome features such as red, yellow, white, blue, and black complexion, complexion moisture, spots, and flushing. It combines the theory of facial division corresponding to the five internal organs in traditional Chinese medicine to associate the state of the internal organs and qi and blood.
[0077] Bionic pulse diagnosis instruments (such as the three-finger bionic pulse diagnosis instrument) are used to collect pulse characteristic data: The three-finger bionic pulse diagnosis instrument simulates the manual pulse diagnosis at the cun, guan, and chi positions, and collects 24 pulse parameters such as pulse frequency, rhythm, depth, slowness and speed, strength and weakness, and smoothness and roughness. Through time domain and frequency domain dual analysis algorithms, it filters out interference from limb tremors and heart rate fluctuations, and accurately matches the TCM pulse diagnosis standards.
[0078] To make the measured pulse characteristic data more accurate, this embodiment can obtain pulse characteristic data through sonar pulse detection technology and strong cold light transmission imaging technology: Sonar pulse detection technology emits high-frequency ultrasound waves to the cun, guan, and chi positions of the wrist and receives the echoes to generate a continuous dynamic pulse waveform map reflecting the pulse position, pulse rate, pulse shape, and pulse strength; Strong cold light transmission imaging technology uses low-heat, high-penetration cold light to irradiate the peripheral blood vessels of the wrist and limbs in a directional manner, and images the subcutaneous radial artery and microcirculation blood flow status in real time, extracting data on blood flow filling, blood flow velocity, changes in blood vessel diameter, and peripheral blood perfusion status.
[0079] More specifically, sonar pulse detection technology focuses on the main radial, cun, guan, chi, and radial arteries at the wrist. Through high-frequency ultrasonic pulse transmission and echo reception, it captures dynamic waveform data of the pulse, achieving full-dimensional quantification of the pulse in Traditional Chinese Medicine, replacing the doctor's tactile perception. Its technical principle is as follows: a high-frequency, low-sensitivity sonar beam precisely locks onto the cun, guan, and chi points at the wrist, collecting the amplitude, frequency, period, slope, and damping value of the pulse in real time, generating a continuous dynamic pulse waveform. It is non-invasive, radiation-free, and suitable for detecting superficial blood vessels in the human body.
[0080] Precise mapping between sonar waveforms and TCM symptom diagnosis:
[0081] - Floating pulse: High amplitude of sonar waveform, strong superficial echo, and rapid superficial diffusion of pulsation, corresponding to external pathogenic syndrome;
[0082] - Deep pulse: Sonar echoes are deeply focused, with weak superficial amplitude and inward-curving waveform, corresponding to internal symptoms and Qi deficiency;
[0083] - Slow pulse: The waveform period is long and the frequency of pulsation per unit time is low, which corresponds to cold syndrome and yang deficiency;
[0084] - Rapid pulse: The waveform period is shortened, the pulsation is dense, and the frequency is high, which corresponds to heat syndrome and yin deficiency;
[0085] - Slippery pulse: The waveform is smooth and rounded, the fluctuation is continuous and the amplitude is uniform, which corresponds to phlegm-dampness, food stagnation and pregnancy;
[0086] - A choppy pulse: The waveform is irregular with spiky lines, obvious pauses and breaks, and uneven rhythm, corresponding to blood stasis and deficiency of qi and blood;
[0087] - Wiry pulse: The waveform is steep and stiff, with rapid rise and fall and extremely high tension, corresponding to liver qi stagnation and pain.
[0088] More specifically, the high-intensity cold light transmission imaging technology differs from ordinary optical imaging. It uses low-heat, high-penetration high-intensity cold light to irradiate the main pulse of the wrist and peripheral blood vessels of the limbs. Utilizing the differences in light transmittance of biological tissues, it can image the subcutaneous radial artery and microcirculation blood flow in real time, directly corresponding to the traditional Chinese medicine theory of "qi and blood deficiency, blood vessel unblocking and stasis, and peripheral qi mechanism." Its technical principle is as follows: the wavelength of the high-intensity cold light is adapted to penetrate human soft tissue without thermal damage or interference with the original state of blood flow. Through a high-definition, high-speed imaging module, it captures changes in the diameter of the cun, guan, and chi vessels in the wrist, blood flow velocity, blood flow fullness, and peripheral blood perfusion status, outputting real-time dynamic blood flow transmission video.
[0089] High-intensity cold light transmission imaging and precise mapping of symptoms in traditional Chinese medicine:
[0090] - Sufficient Qi and Blood: Blood vessels are full and plump, blood flow is uniform and continuous, the extremities are evenly translucent, and perfusion is sufficient;
[0091] - Qi and blood deficiency: thin blood vessels, thin blood flow, slow flow rate, intermittent peripheral blood flow, and dim translucency;
[0092] - Blood stasis: Local blood flow blockage, uneven flow velocity, local vascular filling defects, and dark areas of blood stasis.
[0093] - Cold congeals the meridians: insufficient blood flow to the extremities, vasoconstriction in the wrist, and slow blood flow;
[0094] - Excessive heat and blood stasis: vasodilation, excessively rapid blood flow, and bright and agitated blood flow images.
[0095] Strong cold light transmission imaging technology can realize the visual visualization of the traditional Chinese medicine concept of "qi and blood that cannot be seen or felt" and make the deficiency, excess, stagnation, and cold and heat states of qi and blood intuitively present, making up for the deficiency of pulse diagnosis which can only perceive pulsation and cannot observe the essence of blood flow.
[0096] Breath sensors are used to collect breath characteristic data: high-precision breath sensors collect data such as respiratory rate, breath temperature, and turbidity, and combine them with the TCM respiratory syndrome judgment standards to help judge the strength of lung qi, spleen and stomach function, and the state of damp heat in the body.
[0097] Millimeter-wave non-contact vital sign monitoring equipment and Western medicine vital sign acquisition terminals for body temperature, heart rate, blood oxygen, etc. are used to collect basic Western medicine vital sign data: relying on millimeter-wave non-contact monitoring equipment and embedded sensor terminals, basic Western medicine vital sign data are collected without contact, and the data is standardized and converted according to national Western medicine clinical laboratory standards; the collection dimensions include: core vital sign indicators such as body temperature, resting heart rate, blood oxygen saturation, blood pressure, respiratory rate, body fat, and basal metabolism, while also recording structured basic Western medicine vital sign data such as user age, gender, height and weight, past medical history, allergy history, and medication history.
[0098] Meridian thermal characteristic data are obtained through infrared thermal spectrum imaging technology: by collecting the thermal radiation spectrum of the corresponding reflection areas of meridians, acupoints and internal organs on the human body surface, a global thermal spectrum distribution map is generated, and the temperature value, thermal radiation intensity and cold-heat cross-sectional features of each zone are extracted.
[0099] More specifically, infrared thermal spectrum technology captures the temperature field distribution and thermal radiation spectrum of meridians, acupoints, and pulse areas on the human body surface. Combined with AI training using a large-sample TCM diagnostic database, it transforms surface thermal imaging data into TCM conclusions on Yin-Yang, cold-heat, deficiency-excess, and Zang-Fu organ pathologies, thus establishing a connection between "pulse pattern - meridian temperature - Zang-Fu organ pathogenesis." Its technical principle is as follows: a high-precision infrared thermal imaging module collects the thermal radiation spectrum of the wrist's Cun, Guan, and Chi pulse areas, the meridians of the limbs, and the corresponding reflex areas of the Zang-Fu organs. It distinguishes between normal thermal balance, localized heat accumulation, and localized coldness and heat deficiency, generating a global thermal spectrum distribution map.
[0100] Precise mapping between infrared thermal spectrum and TCM symptoms:
[0101] - Excessive heat syndrome: The overall high temperature radiation and high thermal spectrum in the Cun, Guan and Chi pulse areas correspond to excessive internal heat and external heat pathogens;
[0102] - Deficiency-cold syndrome: Low temperature in the wrist pulse area and extremities, insufficient heat radiation, and low spectrum, corresponding to yang deficiency, qi deficiency, and cold coagulation;
[0103] - Localized visceral heat: Localized heat spots accumulate at a single point (cun / guan / chi), corresponding to localized heat in the heart and lungs, spleen and stomach, liver and kidneys;
[0104] - Meridian blockage: The flow of blood vessels and meridians shows a discontinuity of alternating hot and cold temperatures and uneven thermal spectrum, which corresponds to qi stagnation, blood stasis, and meridian obstruction.
[0105] This embodiment relies on the undisturbed environment of a closed treatment cabin in the clinic, and uses a dedicated algorithm to accurately extract the four-dimensional core features of traditional Chinese medicine: tongue, face, pulse, and breath, adapting to the needs of high-precision diagnosis and treatment. During the data collection process, the algorithm automatically filters abnormal data, performs multiple sampling and fitting on data with deviations from a single collection, and outputs effective vital sign data that meet clinical standards, providing objective evidence for Western medicine disease screening and constitution assessment.
[0106] S2: Perform data preprocessing on the TCM diagnostic and treatment characteristic data and the Western medicine basic physical sign data respectively to obtain the TCM syndrome feature vector and the Western medicine physical sign feature vector;
[0107] Specifically, in response to interference factors such as the clinic environment, equipment operation errors, and fluctuations in human condition, this embodiment uses algorithms to perform preprocessing on TCM diagnostic and treatment characteristic data and Western medicine basic physical sign data, including data cleaning, outlier removal, environmental calibration, and feature normalization. This eliminates the influence of external factors such as light, temperature, noise, equipment errors, and human movement interference, ensuring the accuracy of input data.
[0108] More specifically, environmental calibration: based on the temperature, humidity, and light parameters of the collected environment, automatically corrects color deviations in tongue and facial feature data, as well as temperature and humidity deviations in breath feature data; outlier removal: removes transient outlier data using the 3σ criterion; normalization processing: standardizes parameters of different dimensions in TCM diagnostic feature data and Western medicine basic physical sign data to generate TCM syndrome feature vectors and Western medicine physical sign feature vectors, providing data support for subsequent fusion reasoning.
[0109] S3: Input the TCM syndrome feature vector into the pre-trained multi-feature fusion syndrome differentiation model and output the TCM syndrome determination result;
[0110] Specifically, this embodiment pre-constructs a multi-feature fusion syndrome differentiation model based on the core theories of "Standards for Diagnosis and Efficacy of Diseases and Syndromes in Traditional Chinese Medicine" and "Traditional Chinese Medicine Syndrome Differentiation," combined with training on a massive dataset of primary-level clinical cases. This multi-feature fusion syndrome differentiation model enables standardized and traceable TCM syndrome differentiation, avoiding the problems of strong subjectivity and inconsistent standards in traditional manual syndrome differentiation.
[0111] This multi-feature fusion syndrome differentiation model uses the four diagnostic methods (inspection, auscultation and olfaction, palpation, and olfaction) as its core, combining the three-dimensional systems of visceral differentiation, eight-principle differentiation, and qi-blood-body fluid differentiation. Through a weighted correlation algorithm, it matches syndrome combination patterns. The model incorporates thousands of common TCM syndrome combination rules, covering frequently occurring syndrome types at the grassroots level, such as spleen and stomach diseases, lung diseases, liver and kidney deficiency, damp-heat, phlegm-dampness, blood stasis, qi deficiency, and yang deficiency.
[0112] The dialectical process of the multi-feature fusion syndrome differentiation model is as follows:
[0113] Summary of Four Diagnostic Features: Integrating all effective feature data of tongue, face, pulse, and breath;
[0114] Feature weight assignment: Based on the correlation between symptoms, differentiated weights are assigned to core features and auxiliary features;
[0115] Syndrome matching reasoning: Perform similarity matching with the standard syndrome database and output the top 3 high-probability syndromes;
[0116] Syndrome merging and deduplication: Eliminate conflicting syndromes and determine the final core syndrome and concurrent syndromes based on the user's physical characteristics;
[0117] Traditional Chinese Medicine Etiology and Pathogenesis Analysis: Automatically generates results on etiology, pathogenesis, physical constitution, and severity of illness.
[0118] This embodiment, through a multi-feature fusion syndrome differentiation model, can ultimately output the user's TCM constitution type, core syndrome, concurrent symptoms, etiology and pathogenesis, TCM treatment principles (strengthening the body's resistance, eliminating pathogens, clearing heat, removing dampness, tonifying qi and nourishing blood, etc.), contraindications, and other TCM syndrome determination results, providing theoretical support for TCM prescriptions, physiotherapy plans, and health maintenance.
[0119] S4: Input the Western medicine physical signs feature vector into the pre-trained Western medicine intelligent diagnostic model and output the Western medicine disease screening results;
[0120] Specifically, this embodiment pre-trains a machine learning model based on the "Guidelines for the Diagnosis and Treatment of Common Diseases in Western Medicine" and big data on physical signs to obtain a Western medicine intelligent diagnostic model. This Western medicine intelligent diagnostic model performs intelligent screening and preliminary diagnosis of common and chronic diseases that are prevalent at the grassroots level. It can focus on three core scenarios: chronic disease management, mild symptom screening, and risk warning. For example, it covers high-frequency symptoms at the grassroots level such as respiratory, digestive, cardiovascular, and metabolic diseases, including common mild symptoms and chronic diseases such as colds, pharyngitis, bronchitis, hypertension, hyperlipidemia, hyperglycemia, indigestion, and sub-health syndromes.
[0121] This Western medicine intelligent diagnostic model uses objective physical signs and indicators from Western medicine as its core, combined with users' basic health data and medical history data. It employs a dual algorithm of threshold determination and case similarity matching to complete preliminary disease screening. Unlike models with a single threshold determination, this model can dynamically adjust the judgment criteria based on differences in age, gender, and physical condition, reducing the probability of misdiagnosis and missed diagnosis.
[0122] S5: Cross-validate and dynamically weight the results of TCM syndrome determination with the results of Western medicine disease screening, and output the fused diagnosis and treatment conclusion.
[0123] Specifically, the two-way cross-validation of TCM and Western medicine diagnosis results is as follows: When the TCM syndrome determination result and the Western medicine symptom screening result corroborate each other, the confidence of the final diagnosis is increased. For example, when the TCM syndrome is determined to be damp-heat accumulation in the spleen, and the Western medicine signs show high blood lipids and metabolic abnormalities, the two-way cross-validation of the diagnosis results improves the accuracy of diagnosis and treatment. When the TCM syndrome determination result and the Western medicine symptom screening result conflict, a weighted correction is performed based on the preset symptom type and the confidence of the indicators, and low-confidence determinations are eliminated. When the TCM syndrome determination result and the Western medicine symptom screening result seriously conflict, it is automatically marked as a difficult case and triggers cloud-based manual review.
[0124] In this embodiment, the principles of dynamic weight fusion are as follows: Western medicine determines indicators, symptoms, and risks; traditional Chinese medicine determines pathogenesis, constitution, and treatment plan; objective indicators are given priority for error correction, and overall syndrome is optimized in a coordinated manner; mild cases are mainly treated with traditional Chinese medicine and supplemented by Western medicine, chronic diseases are treated with coordinated intervention of traditional Chinese and Western medicine, and severe cases are treated with Western medicine and supplemented by traditional Chinese medicine.
[0125] Based on the above principles, the weight of traditional Chinese medicine syndrome differentiation and the weight of Western medicine diagnosis can be dynamically allocated for different disease types: for sub-health and constitution conditioning categories, the weight of traditional Chinese medicine syndrome differentiation accounts for 60%, and the weight of Western medicine diagnosis accounts for 40%; for chronic diseases and mild organic diseases, the weight of Western medicine diagnosis accounts for 60%, and the weight of traditional Chinese medicine syndrome differentiation accounts for 40%; for critical risk diseases, Western medicine risk judgment is completely taken as the core, and traditional Chinese medicine is only used as an auxiliary reference; dynamic weights are adapted to different diagnosis and treatment scenarios to maximize the rationality of diagnosis and treatment.
[0126] S6: Automatically match and output an intervention plan based on the fused diagnosis and treatment conclusion; the intervention plan includes: traditional Chinese medicine prescriptions, spectrum physiotherapy parameters, and health management suggestions;
[0127] Specifically, in this embodiment, an intervention plan and a standardized "Intelligent Integrated Traditional Chinese and Western Medicine Diagnosis and Treatment Report" are automatically generated based on the fused diagnosis and treatment conclusion, including: traditional Chinese medicine prescriptions, spectrum physiotherapy parameters, and health management suggestions, such as: preliminary Western medicine diagnosis and index interpretation, traditional Chinese medicine constitution and syndrome judgment, comprehensive etiology analysis, integrated traditional Chinese and Western medicine treatment plan, traditional Chinese medicine / food therapy prescription, physiotherapy conditioning plan, work and rest and exercise guidance, reexamination cycle and risk warning, etc.
[0128] Wherein, the matching method of the traditional Chinese medicine prescription is as follows:
[0129] Based on the fused diagnosis and treatment conclusion, the pre-constructed compliant traditional Chinese medicine prescription matching model and the database of monarch, minister, assistant and guide compatibility rules are invoked to automatically screen compatible prescriptions, perform the Eighteen Incompatibilities and Nineteen Antagonisms contraindication verification, carry out standardized dosage adjustment according to the age, constitution, allergy history and severity of the condition of the target object, and can realize original prescription output, modification based on syndrome differentiation, and standardized dosage adjustment. All prescriptions comply with primary medical medication specifications and support secondary review and calibration by cloud-based physicians. Wherein, the compliant traditional Chinese medicine prescription matching model is constructed based on the Pharmacopoeia of the People's Republic of China and the Manual of Appropriate Traditional Chinese Medicine Technologies for Primary Care.
[0130] The matching method of the spectrum physiotherapy parameters is as follows: automatically match the far-infrared spectrum physiotherapy mode according to the traditional Chinese medicine syndrome judgment result: for cold syndrome, yang deficiency, and wind-dampness arthralgia, the far-infrared continuous warming mode is matched; for damp-heat and local inflammation, the short-pulse low-energy mode is matched; for qi and blood stasis and blocked meridians, the pulse and continuous composite spectrum mode is matched.
[0131] The spectrum physiotherapy parameters at least include: irradiation site, working mode, output power, irradiation duration, and treatment course frequency.
[0132] The generation method of the health management suggestions is as follows:
[0133] When the fused diagnosis and treatment conclusion includes a Western medicine disease screening result, a Western medicine intervention plan is output, and the Western medicine intervention plan at least includes dietary contraindications, exercise suggestions, index control standards, and long-term follow-up plans;
[0134] When the integrated diagnosis and treatment conclusion includes the results of TCM syndrome determination, a personalized health care and conditioning plan is output by combining the principles of seasonal health preservation, meridian flow, and body constitution adaptation. The personalized health care and conditioning plan includes at least dietary therapy, acupoint massage guidance, work and rest adjustment plan, and exercise adaptation suggestions, so as to realize the integrated service of diagnosis and treatment and conditioning.
[0135] S7: Perform security verification on the intervention plan and determine the risk level; if the determination result is high risk level, involves special populations or special medications, automatically intercept and push to the cloud-based physician terminal for manual review.
[0136] Specifically, the risk levels are divided into three levels:
[0137] Level 1 (Low Risk): Mild symptoms, sub-health, slightly abnormal indicators; automatically generates diagnosis and treatment plans.
[0138] Level 2 (Medium Risk): Significantly abnormal indicators and unstable chronic disease; provide an intervention plan and remind patients to have regular check-ups.
[0139] Level 3 (High Risk): Suspected severe illness, severely elevated indicators, critical signs. Automatic interception of self-diagnosis and treatment, immediately triggering remote consultation with a cloud-based physician, and simultaneously pushing medical advice. When the high-risk level is determined to be Level 3 (High Risk), the entire process of diagnosis and treatment data and reasoning logs are encrypted and stored on the blockchain for evidence, achieving tamper-proof, end-to-end medical traceability.
[0140] This embodiment also sets up multiple interception rules: for difficult and complicated diseases, high-risk signs, serious conflicts between Chinese and Western medicine results, first-time visits to special groups (elderly, pregnant women, patients with serious chronic diseases), and special prescription medication scenarios, all of which automatically trigger remote support from cloud-based licensed physicians, and manual review of diagnostic results and prescription plans. Only after confirming that there are no errors can the report be output, completely avoiding the safety risks of unmanned diagnosis and treatment.
[0141] Safety verification also includes: performing safety protection verification on spectrum therapy parameters: real-time monitoring of skin temperature at the irradiation site and irradiation distance; when the skin temperature exceeds the safety threshold or the irradiation distance exceeds the safety range, the therapy is automatically terminated and an alarm is triggered.
[0142] In this embodiment, all data throughout the entire diagnosis and treatment process (collected data, reasoning logs, diagnostic results, and prescription records) is timestamped and encrypted on the blockchain, with local encrypted storage and cloud dual backup. It is tamper-proof, fully traceable, and fully complies with the "Regulations on Medical Record Management of Medical Institutions" and the "Specifications for Medical Data Security Management," meeting the requirements for health and family planning filing and daily supervision.
[0143] This embodiment can also automatically accumulate daily clinical cases, and continuously optimize the accuracy of diagnosis, diagnostic accuracy, and prescription suitability through incremental learning algorithms. Combined with the physical characteristics of grassroots areas and the characteristics of high-incidence diseases, the algorithm can be localized and iteratively upgraded to continuously improve the adaptability of clinic diagnosis and treatment.
[0144] This embodiment provides an adaptation of a multimodal perception-based integrated traditional Chinese and Western medicine diagnosis and treatment method. The standardized 24-hour unattended smart TCM clinic and prefabricated mobile modular clinic, by building a dual-dimensional fully intelligent diagnosis and treatment algorithm system combining TCM syndrome differentiation and Western medicine physical signs in the unattended smart TCM clinic and prefabricated mobile modular clinic, can replace the traditional manual consultation, physical examination and syndrome differentiation process, thereby realizing unattended data collection, standardized feature extraction, parallel reasoning of dual models, integrated diagnosis of TCM and Western medicine results, and output of personalized prescriptions and health plans. It is suitable for the routine diagnosis and treatment needs of community health service centers, township health centers, village clinics, community health centers, and health stations of enterprises and institutions. It supports two operation modes: fixed clinic routine diagnosis and treatment and mobile outreach clinics. At the same time, it fully meets the diagnosis and treatment standards, data security and medical traceability requirements of health departments.
[0145] The core advantages of this method are "dual-dimensional complementarity, precise noise reduction, and intelligent backup". It corrects the subjective diagnostic bias of traditional Chinese medicine by using objective physical signs data from Western medicine, and makes up for the limitations of single-indicator diagnosis and treatment in Western medicine by using the holistic syndrome system of traditional Chinese medicine. It solves the pain points of traditional AI diagnosis and treatment, such as vague syndrome differentiation, one-sided indicators, high misdiagnosis rate, and poor adaptability, and realizes standardized, intelligent, and compliant unmanned diagnosis and treatment combining traditional Chinese and Western medicine.
[0146] The integrated traditional Chinese and Western medicine diagnosis and treatment method based on multimodal perception provided in this embodiment has the following advantages:
[0147] 1. Standardized diagnosis and treatment: Avoiding differences in human experience, the diagnosis and treatment standards of integrated traditional Chinese and Western medicine are unified, replicable, and scalable;
[0148] 2. High safety and compliance: Multiple risk interception, manual backup, and full data traceability fully meet medical regulatory requirements;
[0149] 3. Highly adaptable: The algorithm is adapted to both fixed clinics and mobile shelters, and is suitable for the physical condition of the general public and common diseases.
[0150] 4. Fully automated process: No medical staff are required on duty; AI independently completes data collection, identification, diagnosis, prescription, and report output.
[0151] In summary, the method provided in this embodiment deeply matches the hardware system of unattended smart TCM clinics with the needs of primary healthcare services, constructing a complete intelligent diagnosis and treatment closed loop of "hardware perception - data preprocessing - two-dimensional reasoning - fusion decision-making - prescription output - security safeguards - iterative optimization". Through the complementarity of TCM and Western medicine technologies, it retains the core advantages of TCM's holistic diagnosis and constitution regulation while relying on objective physical data from Western medicine to ensure the accuracy and safety of diagnosis and treatment. It effectively addresses pain points such as insufficient manpower, inconsistent standards, and inadequate coverage in primary TCM services. It can be quickly implemented and replicated in batches, providing core technical support for the normalization of smart TCM services at the primary level. It systematically solves the problems of fragmented data collection, strong subjectivity in diagnosis, separation of TCM and Western medicine diagnosis, lack of automated closed-loop treatment plans, and lack of security safeguards in unattended scenarios found in existing technologies.
[0152] Example 2
[0153] This embodiment provides a multimodal sensing-based integrated traditional Chinese and Western medicine diagnosis and treatment system, including:
[0154] The multimodal data acquisition module is used to synchronously acquire TCM diagnostic and treatment characteristic data and Western medicine basic physical signs data of the target object through a multimodal sensor array in a closed, soundproof, and light-proof environment; wherein, the TCM diagnostic and treatment characteristic data includes: four diagnostic methods characteristic data and meridian heat state characteristic data;
[0155] The data preprocessing module is used to preprocess the TCM diagnostic and treatment feature data and the Western medicine basic physical sign data respectively to obtain TCM syndrome feature vectors and Western medicine physical sign feature vectors.
[0156] The TCM intelligent syndrome differentiation module is used to input the TCM syndrome feature vector into a pre-trained multi-feature fusion syndrome differentiation model and output the TCM syndrome determination result.
[0157] The Western medicine intelligent diagnosis module is used to input the Western medicine vital sign feature vector into the pre-trained Western medicine intelligent diagnosis model and output the Western medicine disease screening results.
[0158] The integrated TCM and Western medicine decision-making module is used to cross-validate and dynamically weight the TCM syndrome determination results and the Western medicine disease screening results, and output the integrated diagnosis and treatment conclusion.
[0159] An intervention plan generation module is used to automatically match and output intervention plans based on the fused diagnostic and treatment conclusions; the intervention plans include: traditional Chinese medicine prescriptions, spectrum physiotherapy parameters, and health management recommendations;
[0160] The security verification and fallback module is used to verify the security of the intervention plan and determine the risk level. If the determination result is a high-risk level, involves special populations or special medications, it is automatically intercepted and pushed to the cloud-based physician terminal for manual review.
[0161] For details on the specific implementation of each module in a multimodal perception-based integrated traditional Chinese and Western medicine diagnosis and treatment system, please refer to the above description of the limitations of a multimodal perception-based integrated traditional Chinese and Western medicine diagnosis and treatment method, which will not be repeated here.
[0162] The technical features of the above embodiments can be combined in any way (as long as there is no contradiction in the combination of these technical features). For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described; these embodiments not explicitly written should also be considered to be within the scope of this specification.
Claims
1. A diagnostic and treatment method integrating traditional Chinese and Western medicine based on multimodal perception, characterized in that, include: Step 1: In a closed, soundproof, and light-proof environment, simultaneously collect TCM diagnostic and treatment characteristic data and Western medicine basic physical signs data of the target object through a multimodal sensor array; wherein, the TCM diagnostic and treatment characteristic data includes: four diagnostic methods characteristic data and meridian heat state characteristic data; Step 2: Perform data preprocessing on the TCM diagnostic and treatment feature data and the Western medicine basic physical sign data respectively to obtain TCM syndrome feature vector and Western medicine physical sign feature vector; Step 3: Input the TCM syndrome feature vector into the pre-trained multi-feature fusion syndrome differentiation model and output the TCM syndrome determination result; Step 4: Input the Western medicine physical signs feature vector into the pre-trained Western medicine intelligent diagnostic model and output the Western medicine disease screening results; Step 5: Cross-validate and dynamically weight the TCM syndrome determination results with the Western medicine symptom screening results, and output the fused diagnosis and treatment conclusion; Step 6: Based on the fused diagnostic and treatment conclusions, automatically match and output an intervention plan; the intervention plan includes: traditional Chinese medicine prescription, spectrum physiotherapy parameters, and health management suggestions; Step 7: Perform a security check on the intervention plan and determine the risk level; if the determination result is a high-risk level, involves special populations or special medications, automatically intercept and push it to the cloud-based practicing physician terminal for manual review.
2. The integrated traditional Chinese and Western medicine diagnosis and treatment method based on multimodal perception according to claim 1, characterized in that, In step 1, the four diagnostic features data include tongue appearance features data, facial appearance features data, pulse appearance features data, and breath features data; The pulse characteristic data is acquired through sonar pulse detection technology and high-intensity cold light transmission imaging technology. The sonar pulse detection technology generates a continuous dynamic pulse waveform diagram that reflects the pulse position, pulse rate, pulse shape, and pulse strength by emitting high-frequency ultrasound waves to the cun, guan, and chi positions of the wrist and receiving the echoes. The high-intensity cold light transmission imaging technology uses low-heat, high-penetration cold light to irradiate the peripheral blood vessels of the wrist and limbs in a directional manner, and images the subcutaneous radial artery and microcirculation blood flow status in real time, extracting data on blood flow filling, blood flow velocity, changes in blood vessel diameter, and peripheral blood flow perfusion status. The meridian thermal characteristic data is obtained through infrared thermal spectrum imaging technology: by collecting the thermal radiation spectrum of the corresponding reflection areas of meridians, acupoints and internal organs on the human body surface, a global thermal spectrum distribution map is generated, and the temperature value, thermal radiation intensity and cold-heat cross-sectional features of each zone are extracted.
3. The integrated traditional Chinese and Western medicine diagnosis and treatment method based on multimodal perception according to claim 2, characterized in that, In step 2, the data preprocessing includes: Environmental calibration: Based on the temperature, humidity and light parameters of the collected environment, automatically correct the color deviation of the tongue feature data and the facial feature data, as well as the temperature and humidity deviation of the breath feature data; Outlier removal: Transient outlier data is removed using the 3σ criterion; Normalization processing: The parameters of different dimensions in the TCM diagnostic and treatment feature data and the Western medicine basic physical sign data are standardized and unified to generate the TCM syndrome feature vector and the Western medicine physical sign feature vector.
4. The integrated traditional Chinese and Western medicine diagnosis and treatment method based on multimodal perception according to claim 1, characterized in that, In step 5, the cross-validation specifically includes: When the results of TCM syndrome determination and Western medicine disease screening are mutually corroborated, the confidence of the final diagnosis is increased. When there is a conflict between the TCM syndrome determination results and the Western medicine disease screening results, a weighted correction is performed based on the preset disease type and the confidence level of the indicators, and low confidence determinations are eliminated. When the TCM syndrome determination result seriously conflicts with the Western medicine disease screening result, it is automatically marked as a difficult case and triggers cloud-based manual review.
5. The integrated traditional Chinese and Western medicine diagnosis and treatment method based on multimodal perception according to claim 1, characterized in that, Step 5, the dynamic weight fusion specifically includes: The weights of TCM syndrome differentiation and Western medicine diagnosis are dynamically allocated according to different disease types: Sub-health and constitution conditioning category: Traditional Chinese medicine diagnosis accounts for 60% of the weight, while Western medicine diagnosis accounts for 40% of the weight. Chronic diseases and mild organic lesions: Western medicine diagnosis accounts for 60% of the weight, and traditional Chinese medicine syndrome differentiation accounts for 40% of the weight. Critical and high-risk illnesses: Western medicine risk assessment is the core, with traditional Chinese medicine only serving as a supplementary reference.
6. The integrated traditional Chinese and Western medicine diagnosis and treatment method based on multimodal perception according to claim 1, characterized in that, In step 6, the matching method for the traditional Chinese medicine prescription is as follows: Based on the fused diagnostic and treatment conclusions, the pre-built compliant traditional Chinese medicine prescription matching model and the database of monarch-minister-assistant-guide compatibility rules are invoked to automatically screen suitable prescriptions and perform the Eighteen Incompatibilities and Nineteen Antagonisms Contraindication Verification. The dosage is standardized and adjusted according to the target subject's age, constitution, allergy history, and severity of illness.
7. The integrated traditional Chinese and Western medicine diagnosis and treatment method based on multimodal perception according to claim 1, characterized in that, In step 6, the matching method for the spectrum therapy parameters is as follows: Based on the TCM syndrome differentiation results, the far-infrared spectrum therapy mode is automatically matched: Cold syndrome, Yang deficiency, rheumatic pain: Matched with far-infrared continuous warming mode; Moist heat, local inflammation: Match short-pulse low-energy mode; Qi and blood stagnation, meridian blockage: Matching pulse and continuous composite spectrum mode; The spectrum therapy parameters include at least: irradiation site, working mode, output power, irradiation duration, and treatment frequency.
8. The integrated traditional Chinese and Western medicine diagnosis and treatment method based on multimodal perception according to claim 1, characterized in that, In step 6, the method for generating the health management recommendations is as follows: When the fused diagnosis and treatment conclusion includes Western medicine disease screening results, a Western medicine intervention plan is output. The Western medicine intervention plan includes at least dietary restrictions, exercise recommendations, indicator control standards, and a long-term follow-up plan. When the integrated diagnosis and treatment conclusion includes the results of TCM syndrome determination, a personalized health care plan is output by combining the principles of seasonal health care, meridian flow, and body constitution matching. The personalized health care plan includes at least dietary therapy, acupoint massage guidance, work and rest adjustment plan, and exercise adaptation suggestions.
9. The integrated traditional Chinese and Western medicine diagnosis and treatment method based on multimodal perception according to claim 1, characterized in that, In step 7, the security verification also includes: Safety protection verification is performed on the spectrum therapy parameters: real-time monitoring of skin temperature at the irradiation site and irradiation distance; when the skin temperature exceeds the safety threshold or the irradiation distance exceeds the safety range, the therapy is automatically terminated and an alarm is triggered. When the high-risk level is determined to be Level 3 high risk, the diagnosis and treatment data and reasoning logs are encrypted and stored on the blockchain throughout the entire process to achieve tamper-proof, full-process medical traceability.
10. A multimodal sensing-based integrated traditional Chinese and Western medicine diagnostic and treatment system, characterized in that, include: The multimodal data acquisition module is used to synchronously acquire TCM diagnostic and treatment characteristic data and Western medicine basic physical signs data of the target object through a multimodal sensor array in a closed, soundproof, and light-proof environment; wherein, the TCM diagnostic and treatment characteristic data includes: four diagnostic methods characteristic data and meridian heat state characteristic data; The data preprocessing module is used to preprocess the TCM diagnostic and treatment feature data and the Western medicine basic physical sign data respectively to obtain TCM syndrome feature vectors and Western medicine physical sign feature vectors. The TCM intelligent syndrome differentiation module is used to input the TCM syndrome feature vector into a pre-trained multi-feature fusion syndrome differentiation model and output the TCM syndrome determination result. The Western medicine intelligent diagnosis module is used to input the Western medicine vital sign feature vector into the pre-trained Western medicine intelligent diagnosis model and output the Western medicine disease screening results. The integrated TCM and Western medicine decision-making module is used to cross-validate and dynamically weight the TCM syndrome determination results and the Western medicine disease screening results, and output the integrated diagnosis and treatment conclusion. An intervention plan generation module is used to automatically match and output intervention plans based on the fused diagnostic and treatment conclusions; the intervention plans include: traditional Chinese medicine prescriptions, spectrum physiotherapy parameters, and health management recommendations; The security verification and fallback module is used to verify the security of the intervention plan and determine the risk level. If the determination result is a high-risk level, involves special populations or special medications, it is automatically intercepted and pushed to the cloud-based physician terminal for manual review.
