Diagnostic system for arthritic cold and heat syndrome

Abstract

The disclosure provides a diagnostic system for arthralgia and cold-heat syndrome, comprising: an information collection module for obtaining current medical information of a patient, the current medical information comprising a tongue image and a joint image of the patient; a judgment module for obtaining a cold-heat syndrome judgment result based on a cold-heat syndrome model; the cold-heat syndrome model comprising a tongue image feature extraction network, a joint image feature extraction network and a cold-heat syndrome discrimination network; the tongue image feature extraction network is used for extracting tongue features from the tongue image, the joint image feature extraction network is used for extracting joint features, and the cold-heat syndrome discrimination network is used for obtaining a current cold-heat syndrome prediction result according to the combination of the tongue features and the joint features; the combination of the tongue features and the joint features is analyzed through the cold-heat syndrome discrimination network, which solves the problems of difficult identification of complex syndromes, easy missed diagnosis and misdiagnosis in traditional artificial diagnosis, and provides a reliable basis for timely adjustment of clinical treatment schemes.

Description

Technical Field

[0001] This disclosure relates to the field of analytical diagnostic technology, and in particular to a diagnostic system for arthritis with chills and fever. Background Technology

[0002] Arthritis is a chronic, systemic autoimmune disease characterized by erosive, symmetrical polyarthritis. Its condition is complex and prone to recurrence. Accurate diagnosis of the cold and heat syndrome is a key prerequisite for developing an effective treatment plan.

[0003] Currently, the clinical diagnosis of arthritis with alternating chills and fever patterns mainly relies on the physician's clinical experience, making subjective judgments by observing the patient's symptoms and signs (such as tongue appearance and the degree of joint swelling). This diagnostic method has significant limitations: on the one hand, the diagnostic results are greatly affected by the physician's experience level and subjective perception differences; different physicians may make different symptom classifications for the same patient, leading to poor diagnostic consistency; on the other hand, traditional diagnosis relies heavily on manual observation, making it difficult to accurately quantify and analyze tongue characteristics (such as the color, thickness, and moisture of the tongue coating) and joint characteristics (such as the degree of swelling and the extent of redness and swelling). This is especially true for complex syndromes such as alternating chills and fever patterns (where the patient's symptoms change between chills and fever) and mixed chills and fever patterns (where chills and fever symptoms coexist), where relying solely on manual judgment can easily lead to missed diagnoses or misdiagnoses.

[0004] While some related technologies offer auxiliary diagnostic methods based on a single feature (such as tongue appearance or joint symptoms alone), the cold and heat patterns of arthritis are often the result of the combined effects of tongue appearance and joint manifestations. Single feature analysis cannot fully reflect the patient's symptom status, thus limiting diagnostic accuracy and failing to meet the clinical need for precise diagnosis of complex symptom patterns. Summary of the Invention

[0005] In view of the shortcomings of the prior art described above, the purpose of this disclosure is to provide a diagnostic system for arthritis with chills and fever, and to solve the problems in the related art.

[0006] The first aspect of this disclosure provides a diagnostic system for arthritis with chills and fever, wherein the diagnostic system includes:

[0007] The information acquisition module is used to acquire the patient's current medical information, which includes the patient's tongue image and joint images;

[0008] The judgment module is used to obtain a judgment result of cold and heat syndrome based on the cold and heat syndrome model; the cold and heat syndrome model includes: a tongue image feature extraction network, a joint image feature extraction network, and a cold and heat syndrome discrimination network; the tongue image feature extraction network is used to extract tongue image features from the tongue image, the joint image feature extraction network is used to extract joint features, and the cold and heat syndrome discrimination network is used to obtain the current cold and heat syndrome prediction result based on the combination of tongue image features and joint features; the joint images include: infrared joint images and color joint images;

[0009] The tongue image feature extraction network adopts a convolutional neural network architecture and is trained using the tongue image.

[0010] The joint image feature extraction network includes a first joint detection network trained on the infrared joint image, a second joint detection network trained on the color joint image, and a third joint detection network trained collaboratively on the infrared joint image and the color joint image.

[0011] Therefore, the cold and heat syndrome discrimination network is trained by using training data composed of the tongue image and the joint image combined with cold and heat syndrome labels;

[0012] The predicted results of cold and heat syndromes include at least one of the following: cold syndrome, heat syndrome, cold-heat transition syndrome, and cold-heat alternation syndrome; the cold-heat transition syndrome is determined by comparing the current preliminary cold and heat syndrome prediction results obtained based on the tongue appearance characteristics and joint characteristics with the historical cold and heat syndrome prediction results when the patient has historical cold and heat syndrome prediction results in the diagnostic system, and discovering changes in cold / heat syndrome.

[0013] In an embodiment of the first aspect, the tongue features include tongue color parameters, tongue coating color parameters, tongue coating moisture and thickness; and / or, the joint features include the percentage of red and swollen area in the joint region, skin temperature distribution characteristics, degree of swelling and abnormal joint structure parameters.

[0014] In the first aspect of the embodiment, the information acquisition module is further configured to acquire interference information that affects the tongue image characteristics, the interference information including behaviors that occur within a preset time period before the tongue image acquisition time that cause changes in the tongue image; the body state information represented by the changed tongue image deviates from the patient's actual body state information.

[0015] The diagnostic system also includes an image confirmation module, which is used to confirm the availability of the tongue image based on the interference information, and outputs the confirmation to the judgment module when the tongue image is in a usable state.

[0016] In an embodiment of the first aspect, a prompting module is also included for generating acquisition guidance information when the tongue image and / or the joint image is unavailable.

[0017] In the first aspect of the embodiment, a storage module is also included, which stores the patient's historical cold and heat syndrome prediction results and the corresponding tongue image and joint image; when the judgment module generates the current cold and heat syndrome prediction result, it calls the historical cold and heat syndrome prediction results in the storage module for comparison and analysis.

[0018] In a first aspect embodiment, the judgment module calls the stored historical cold and heat syndrome prediction results, wherein the historical cold and heat syndrome prediction results are at least one cold or heat syndrome result generated by the system within a preset time interval for the patient;

[0019] The cold and heat syndrome discrimination network is used to extract the historical tongue features and historical joint features corresponding to the historical cold and heat syndrome prediction results, and compare them with the current tongue features and current joint features to obtain the change in the current tongue features and the change in the joint features.

[0020] Among them, when the current preliminary cold / heat syndrome prediction result and the historical cold / heat syndrome prediction result indicate different cold / heat syndromes, and the changes in tongue appearance characteristics and joint characteristics both exceed their respective preset thresholds, it is determined to be a cold / heat transition syndrome. The transition type of the cold / heat transition syndrome includes cold syndrome turning into heat syndrome or heat syndrome turning into cold syndrome.

[0021] In an embodiment of the first aspect, the cold-heat syndrome discrimination network is used to determine cold-heat syndrome based on at least one of the following conditions:

[0022] 1) Detect the presence of both cold and heat symptoms in the current tongue appearance;

[0023] 2) The proportion of redness and swelling area and the skin temperature distribution characteristics of the joints exceed the threshold of heat syndrome, and the abnormal joint morphology parameters meet the criteria of cold syndrome;

[0024] 3) The coexistence ratio of cold symptoms and heat symptoms exceeds the preset overlap threshold.

[0025] In an embodiment of the first aspect, the current medical information further includes clinical inflammatory marker information; the diagnostic system further includes a severity quantification module for chills and fever, used to obtain a severity index of the corresponding symptoms based on the tongue features and the joint features.

[0026] In an embodiment of the first aspect, a prescription generation module is further included, which is used to determine a prescription for treatment and the dosage of drugs in the prescription based on the results of the cold and heat syndrome assessment and the severity index.

[0027] In an embodiment of the first aspect, a prescription generation module is further configured to determine a prescription for treatment and the dosage of drugs in the prescription based on the diagnosis of cold and heat syndrome.

[0028] The beneficial effects of this disclosure are as follows: This application acquires the patient's tongue and joint images through an information acquisition module, and extracts image features using a feature extraction network in a cold and heat syndrome model. This reduces interference from subjective human judgment, avoids diagnostic biases caused by differences in physician experience, and significantly improves the objectivity of cold and heat syndrome diagnosis and the consistency of diagnosis among different physicians and in different scenarios. Simultaneously, combining tongue and joint features for symptom pattern judgment overcomes the limitations of single-feature analysis in existing technologies. By analyzing the combination of tongue and joint features through a cold and heat syndrome discrimination network, the problem of difficulty in identifying complex symptom patterns and the ease of missed or misdiagnosed diagnoses in traditional manual diagnosis is solved, providing a reliable basis for timely adjustments to clinical treatment plans. Attached Figure Description

[0029] Figure 1 A block diagram illustrating the overall structure of a diagnostic system for arthritis with chills and fever, according to one embodiment of this disclosure.

[0030] Figure 2 This invention presents a structural block diagram of an arthritis diagnosis system for cold and heat syndromes, including an image confirmation module, according to an embodiment of the present disclosure.

[0031] Figure 3 This illustration shows a block diagram of a diagnostic system for arthritis with cold and heat syndromes, including a storage module, according to an embodiment of the present disclosure.

[0032] Figure 4 The diagram illustrates the structural block diagram of a diagnostic system for arthritis with chills and fever syndromes, including a module for quantifying the severity of chills and fever syndromes, according to one embodiment of this disclosure. Detailed Implementation

[0033] The following specific examples illustrate the implementation of this disclosure. Those skilled in the art can easily understand other advantages and effects of this disclosure from the information disclosed herein. This disclosure can also be implemented or applied through other different specific embodiments, and various details in this disclosure can be modified or changed according to different viewpoints and application modules without departing from the spirit of this disclosure. It should be noted that, unless otherwise specified, the embodiments and features in the embodiments of this disclosure can be combined with each other.

[0034] The embodiments of this disclosure will now be described in detail with reference to the accompanying drawings, so that those skilled in the art to which this disclosure pertains can readily implement it. This disclosure may be embodied in many different forms and is not limited to the embodiments described herein.

[0035] In this disclosure, references to terms such as "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., indicate that a specific feature, structure, material, or characteristic represented in connection with that embodiment or example is included in at least one embodiment or example of this disclosure. Furthermore, the specific features, structures, materials, or characteristics represented may be combined in any suitable manner in any one or a group of embodiments or examples. Moreover, without contradiction, those skilled in the art can combine and integrate the different embodiments or examples represented in this disclosure, as well as the features of those different embodiments or examples.

[0036] Furthermore, the terms "first" and "second" are used for illustrative purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include at least one of that feature. In the representation of this disclosure, "a set" means two or more, unless otherwise explicitly specified.

[0037] For the purpose of clarity, devices unrelated to the description are omitted, and the same or similar components throughout the specification are given the same reference numerals.

[0038] Throughout this specification, when it is said that a device is "connected" to another device, this includes not only "direct connection" but also "indirect connection" by placing other components in between. Furthermore, when it is said that a device "comprises" a certain constituent element, unless otherwise stated otherwise, this does not exclude other constituent elements, but rather implies that other constituent elements may be included.

[0039] While the terms first, second, etc., are used in some examples herein to refer to various elements, these elements should not be limited by these terms. These terms are used only to distinguish one element from another. For example, first interface and second interface, etc., are used. Furthermore, as used herein, the singular forms “a,” “an,” and “the” are intended to also include the plural forms unless the context indicates otherwise. It should be further understood that the terms “comprising,” “including,” indicate the presence of features, steps, operations, elements, modules, items, kinds, and / or groups, but do not exclude the presence, occurrence, or addition of one or more other features, steps, operations, elements, modules, items, kinds, and / or groups. The terms “or” and “and / or” as used herein are interpreted as inclusive, or mean any one or any combination thereof. Thus, “A, B, or C” or “A, B, and / or C” means “any one of: A; B; C; A and B; A and C; B and C; A, B, and C.” Exceptions to this definition will only occur if the combination of elements, functions, steps, or operations is inherently mutually exclusive in some way.

[0040] The technical terms used herein are for reference only to specific embodiments and are not intended to limit the scope of this disclosure. The singular form used herein includes the plural form unless the statement explicitly indicates otherwise. The word "comprising" as used in this specification means to specify a particular characteristic, region, integer, step, operation, element, and / or component, and does not exclude the presence or addition of other characteristics, regions, integers, steps, operations, elements, and / or components.

[0041] Although not explicitly defined, all terms, including technical and scientific terms used herein, shall have the same meaning as commonly understood by one of ordinary skill in the art to which this disclosure pertains. Terms defined in commonly used dictionaries shall be further interpreted as having a meaning consistent with the relevant technical literature and the message of the present disclosure, and shall not be over-interpreted as having an ideal or overly formulaic meaning unless otherwise defined.

[0042] Arthritis is a chronic autoimmune disease primarily characterized by inflammation of the synovial membrane of the joints. Typical manifestations include symmetrical swelling, pain, and stiffness in the small joints of the hands and feet. As the disease progresses, joint deformities and limited mobility may occur. Some patients also experience systemic symptoms such as fever, fatigue, and weight loss. From the perspective of Traditional Chinese Medicine (TCM) syndrome differentiation, this disease often presents in different types, including cold syndrome, heat syndrome, cold-heat transition syndrome, or cold-heat alternating syndrome. Cold syndrome is often characterized by cold pain in the joints, aggravated by cold, and a white, greasy tongue coating. Heat syndrome presents with red, swollen, and hot joints, aggravated by heat, and a thick, yellow tongue coating. Cold-heat transition syndrome involves alternating patterns between cold and heat, while cold-heat alternating syndrome involves the simultaneous presence of both cold and heat symptoms. When joint pain flares up or worsens, patients often need to determine their current cold-heat syndrome pattern to adjust pain relief and treatment plans. This application can quickly assist in identifying the syndrome pattern, determining whether a treatment plan adjustment is necessary, and clarifying whether a cold-heat transition has occurred.

[0043] exist Figure 1 In this embodiment, the user terminal 200 can be conveniently used via an online mobile client. Simply take a picture of the tongue and joints and upload it to get the result of the diagnosis of cold and heat syndrome. This allows patients to monitor changes in their symptoms at home or outside of their appointment. It can also be used in a hospital setting, where it can be combined with professional medical equipment to collect more accurate image information. Physicians can then combine the system results to make a comprehensive diagnosis, improving the professionalism and reliability of the diagnosis.

[0044] One embodiment of this disclosure provides a diagnostic system 100 for arthritis with chills and fever, including: an information acquisition module 101 and a judgment module 102.

[0045] refer to Figure 1In this embodiment, the arthritis fever and chills diagnosis system 100 is communicatively connected to a user terminal 200, which can be implemented as a common imaging terminal or a professional imaging terminal. Common imaging terminals include, but are not limited to, smartphones, tablets, home digital cameras, and built-in cameras in laptops; professional imaging terminals include, but are not limited to, professional imaging cameras, infrared thermal imagers, and temperature detection devices.

[0046] In some embodiments, the specific implementation of the arthritis chills and fever diagnosis system 100 is as follows:

[0047] As an example: In the software system implementation, it is deployed as an application on mobile terminals, such as smartphones, tablets, or personal computers, specifically including apps, WeChat mini-programs, or web platforms. Users take images of their tongue and joints using the device's built-in camera. The system automatically analyzes the images, extracting features such as tongue color, coating color, area of ​​redness and swelling, and skin temperature distribution. Combining this with interference information and historical data, it intelligently determines the nature of cold or heat syndromes and generates a visual report. No dedicated hardware is required, making it low-cost, easy to promote, and suitable for scenarios such as home self-testing, remote health management, initial screening at primary healthcare institutions, and follow-up of chronic diseases using traditional Chinese medicine.

[0048] As an example: In a hardware-software integrated device implementation, the system is integrated into a dedicated medical-grade terminal, forming an integrated intelligent diagnostic device. This device includes an imaging module, such as a tongue image acquisition device under standard light sources, a medical infrared thermal imager, positioning aids such as a tongue positioning frame, a hand fixation bracket, a main control computing unit, and a touchscreen interface. The hardware ensures standardized and high-quality image acquisition, avoiding interference from factors such as lighting, angle, and occlusion; the software system is built into the device, possessing complete functions such as image preprocessing, feature extraction, cold / heat syndrome judgment, severity assessment, and prescription generation, supporting local operation or connection to hospital systems.

[0049] The information acquisition module 101 is used to acquire the patient's current medical information, which includes the patient's tongue image and joint image.

[0050] Specifically, in some embodiments, the tongue image may include a color image of the tongue as a whole and details of the tongue coating.

[0051] In practical applications, patients can choose different devices for image acquisition depending on the scenario. In medical institutions, patients can use professional imaging equipment, such as dedicated tongue image acquisition devices, cameras, and infrared thermal imaging devices, at specialized image acquisition points. Simultaneously, on-site positioning aids, such as hand fixation braces and joint supports, are provided to ensure that the position, angle, and lighting conditions of the tongue and joint images meet standardized acquisition requirements. In home or self-use scenarios, patients can complete the acquisition using smartphones or tablets, utilizing their built-in cameras to photograph the tongue and joint areas.

[0052] Preferably, high-precision professional imaging equipment is used to take multi-angle pictures of the patient's tongue protrusion to obtain a clear presentation of key features such as tongue color, tongue coating color, and tongue shape.

[0053] In some embodiments, the joint images may include images of painful joints such as the proximal interphalangeal joints of both hands, metacarpophalangeal joints, wrist joints, and knee joints, to obtain image information reflecting local redness, swelling, temperature changes, or abnormal morphology of the joints.

[0054] In some embodiments, during the data collection process, the diagnostic system 100 guides the patient to complete standard actions (such as naturally protruding the tongue and avoiding excessive force) through the human-computer interaction interface of the user terminal 200.

[0055] In some embodiments, the system automatically determines whether the image meets the analysis requirements. If there is overexposure, blurring, or occlusion, it prompts the user to re-acquire the image.

[0056] In some embodiments, when patients have their images captured offline, the hospital's image acquisition area is also equipped with positioning aids such as hand fixation braces or joint supports to ensure that the patient's position and angle meet the acquisition standards. All acquired tongue and joint images are appended with timestamps, patient identifiers, and acquisition parameter information, and are uploaded to the system database via a secure transmission protocol for analysis by the judgment module 102.

[0057] Optionally, the information acquisition module 101 is further used to acquire interference information that affects the tongue image characteristics. The interference information includes behaviors that cause changes in the tongue image within a preset time period before the tongue image acquisition time; the body state information represented by the changed tongue image deviates from the patient's actual body state information.

[0058] Specifically, in TCM clinical practice, tongue appearance is an important basis for syndrome differentiation and treatment, but its manifestation is easily interfered with by various external behaviors in the short period before collection. For example, if a patient drinks coffee, tea, or eats colored foods before tongue image collection, the tongue coating may be stained, presenting as a yellow or brown coating that is not due to the disease itself; brushing teeth or scraping the tongue can artificially remove the tongue coating, causing it to become thinner or even peel off, masking the true pathological state; smoking, taking certain medications, or having just engaged in strenuous exercise may also cause temporary changes in tongue color, shape, or moisture. These changes in tongue appearance caused by non-pathological factors can cause the physical state represented by the image to deviate from the patient's true cold or heat syndrome. If not identified and processed, this will directly affect the accuracy of subsequent syndrome differentiation. To ensure the reliability of the diagnostic results, the information collection module 101 not only acquires the tongue image but also simultaneously collects interfering information that may affect the tongue appearance within a preset time period before the collection time, such as diet, oral hygiene, smoking, medication, and physical activity, thereby avoiding misjudging temporary, non-disease-related appearance changes as cold or heat syndromes.

[0059] In some embodiments, the collection of interference information can be achieved through user interaction. When a user activates the tongue image capture function, the system automatically pops up a questionnaire, asking whether the user has engaged in any behaviors that might alter the tongue image within a preset time period before the data collection, such as eating, drinking water (especially colored or hot drinks), brushing teeth, smoking, or taking medication. The user completes the feedback by checking boxes or filling out a brief form. In some embodiments, the system can also incorporate natural language processing technology, supporting free input of voice or text, and automatically identifying and extracting key interference behaviors and their occurrence times.

[0060] exist Figure 2 In this embodiment, the diagnostic system 100 further includes an image confirmation module 104, which is used to confirm the availability of the tongue image based on the interference information, and output the confirmation to the judgment module 102 when the tongue image is in an available state.

[0061] Specifically, the image confirmation module 104 analyzes the type of interfering behavior and its impact on the tongue image according to preset rules. If the interference is determined to be significant and not eliminated, such as eating within 10 minutes before the image is collected, the tongue image is considered unusable, and the system will not use it or prompt for re-collection. If it is confirmed that there is no significant interference or the interference has been eliminated, the image is determined to be usable and is input into the judgment module 102 for analysis of cold and heat syndromes.

[0062] Optionally, in Figure 2 In this embodiment, the diagnostic system 100 further includes a prompting module 1041 for generating acquisition guidance information when the tongue image and / or the joint image are unavailable.

[0063] When the image confirmation module 104 detects interference in the tongue image (such as distortion caused by eating, brushing teeth, etc.), or when the information acquisition module 101 detects unqualified image quality (such as blurry, too dark, or obstructed), or when the joint image is taken from the wrong location or the body position is not standard, the prompting module 1041 will generate clear and easy-to-understand graphic or voice guidance information according to the specific problem type. For example, if the system detects that the user has recently consumed coffee, it will prompt, "We have detected that you have recently ingested colored beverages, which may affect the tongue image judgment. Please drink water to rinse your mouth and wait 30 minutes before retaking the image"; if the tongue image is too dark, it will prompt, "Insufficient light. Please move to a bright place or turn on the fill light and retake the image"; if the joint shooting angle is incorrect, it will demonstrate the standard shooting posture through animation. The prompting module 1041 can be set in a mobile app or terminal device interface to guide users to operate in a standardized manner and improve the success rate and quality of image acquisition.

[0064] exist Figure 1 In this embodiment, the judgment module 102 is used to obtain the judgment result of cold and heat syndrome based on the cold and heat syndrome model 103; the cold and heat syndrome model 103 includes: tongue image feature extraction network 1031, joint image feature extraction network 1032 and cold and heat syndrome discrimination network 1033.

[0065] Specifically, as an example, the tongue image feature extraction network 1031 extracts tongue features closely related to the TCM differentiation of cold and heat syndromes from tongue images. In some embodiments, the tongue image feature extraction network 1031 adopts a convolutional neural network (CNN) architecture. During the forward propagation process, the network extracts local and global features of the tongue image step by step through multi-layer convolution and pooling operations: the shallow network captures basic information such as edges and textures; the middle network identifies the tongue outline and tongue coating area; and the deep network learns semantic-level features, such as "pale tongue color" indicating cold syndrome and "red tongue with yellow coating" indicating heat syndrome.

[0066] As an example, the joint image feature extraction network 1032 is used to extract joint features reflecting the inflammatory state of arthritis based on joint images. The joint image feature extraction network includes at least one of the following: a first joint detection network trained on infrared joint images, a second joint detection network trained on color joint images, and a third joint detection network trained on both infrared and color joint images.

[0067] In some embodiments, a target recognition network corresponding to the joint image and the tongue image can be set before the joint image feature extraction network 1032 and the tongue image feature extraction network 1031.

[0068] In some embodiments, joint temperature is detected using infrared thermal imaging technology. When arthritis patients experience joint inflammation, increased local blood flow and hypermetabolism lead to elevated skin temperature, while cold-related conditions such as blood stasis may manifest as lower temperatures. In some embodiments, an infrared thermal imager can be used to acquire thermal images of the joint area, presenting different temperature regions as color images. High-temperature areas (such as red-yellow) indicate potential "heat syndrome" or active inflammation. When patients acquire joint images via a mobile phone, if joint temperature cannot be captured, the joint image feature extraction network 1032 can extract other joint features from the joint image for analysis. Alternatively, the mobile phone can also acquire joint temperature by connecting a miniature, simple thermal imaging device via an interface.

[0069] As an example, the cold / heat syndrome discrimination network 1033 takes tongue features and joint features as input. The network 1033 obtains the corresponding cold / heat syndrome prediction result based on the combination analysis of the two features. The network 1033 can be trained to understand the relationship between tongue features, joint features, and cold / heat syndromes. Specifically, the network 1033 can be trained using training data consisting of combinations of tongue and joint images labeled with cold / heat syndromes. The prediction result can include the probability of different cold / heat syndrome types. For example, training data with the image combination of "pale tongue with white coating + joints that are neither red nor hot" can be labeled as cold syndrome; training data with the image combination of "red tongue with yellow coating + joints that are red, swollen, hot, and painful" can be labeled as heat syndrome; and training data with the image combination of "tongue leaning towards heat but joints are cold and painful" or "tongue leaning towards cold but localized heat" can be labeled as mixed cold / heat syndrome. The cold and heat syndrome discrimination network 1033 can be implemented as, for example, one of AlexNet, ResNet, EfficientNet, Transformer, ConVeXt, SAM, etc.

[0070] In other embodiments, considering that some types of cold and heat syndromes can change over time, such as cold-heat transition syndromes, accurate cold-heat syndrome prediction results can be obtained by combining the patient's cold-heat syndrome prediction results obtained at different times. For example, if the system contains historical prediction results for the patient, the discrimination network or post-processing module will also compare the current result with the historical state. For example, if the previous result was "cold syndrome" and the current result is "heat syndrome," and the change is clinically reasonable, it is determined to be a cold-heat transition syndrome, reflecting the process of the disease changing from cold to heat.

[0071] The so-called cold / heat transition syndrome is determined by comparing the current preliminary cold / heat syndrome prediction result obtained based on the tongue and joint features with the historical cold / heat syndrome prediction results when the patient has historical cold / heat syndrome prediction results in the diagnostic system 100, and discovering the change in cold / heat syndrome.

[0072] The results of the cold and heat syndrome assessment include at least one of the following syndrome types: cold syndrome, heat syndrome, cold and heat transition syndrome, and cold and heat alternation syndrome, thereby assessing the patient's corresponding cold and heat syndrome from various perspectives.

[0073] The cold syndrome is characterized by a pale white tongue with a greasy white coating, and localized coldness in the joints; the heat syndrome is characterized by a red or crimson tongue with a dry yellow coating, and red, swollen, hot, and painful joints; the alternating cold and heat syndrome refers to the simultaneous presence of contradictory cold and heat symptoms in the same patient, such as a hot tongue (red tongue with yellow coating) but cold, non-hot joints, or a cold tongue but localized hot joints. Specifically, the determination of the alternating cold and heat syndrome is based on the existence of historically verifiable alternating cold and heat syndrome prediction results in this diagnostic system 100; the system compares the current preliminary prediction result obtained through the fusion analysis of tongue and joint characteristics with the most recent valid historical result. If a clear change is found in the tongue or joint characteristics—for example, from "cold syndrome" to "heat syndrome," or from "heat syndrome" to "cold syndrome"—and this change conforms to the disease progression pattern, then it is determined to be an alternating cold and heat syndrome.

[0074] Since arthritis is not static, it can change depending on the patient's treatment and lifestyle. For example, the initial diagnosis may be "cold syndrome," such as cold pain in the joints and a pale tongue. However, as the condition progresses or is affected by the patient's lifestyle, the cold evil can stagnate and transform into heat, gradually turning into "heat syndrome" and showing signs of heat. If such a transformation can be detected in time, it indicates that the doctor should adjust the treatment principle accordingly, such as changing from warming the meridians and dispelling cold to clearing heat and unblocking the collaterals, to avoid misusing warm and drying medicines that aggravate the heat symptoms, or misusing cold medicines that damage the yang qi.

[0075] Optionally, in Figure 3 In this embodiment, the diagnostic system 100 further includes a storage module 105 for storing the patient's historical cold and heat syndrome prediction results and corresponding tongue and joint images; when the judgment module 102 generates the current cold and heat syndrome prediction result, it calls the historical cold and heat syndrome prediction results in the storage module 105 for comparison and analysis.

[0076] Specifically, the storage module 105 can store data in a local database or a cloud server. When a patient visits again and a new prediction result for current cold / heat syndrome is generated, the judgment module 102 will call the patient's most recent or multiple valid historical prediction results from the storage module 105 for time-series comparison analysis. By comparing the changing trends of the current and historical syndromes, such as a change from "cold syndrome" to "heat syndrome" or vice versa, it can be determined whether it belongs to a "cold / heat transition syndrome". At the same time, combined with historical image data, it can also help verify whether the changes in the current tongue appearance or joint status are continuous or sudden, and determine whether the tongue image or joint image or the corresponding features are reliable.

[0077] Optionally, the judgment module 102 calls the stored historical cold and heat syndrome prediction results, wherein the historical cold and heat syndrome prediction results are at least one cold or heat syndrome result generated by the system within a preset time interval for the patient. The preset time interval is set to avoid misjudgment due to excessively long intervals (such as old data from several years ago). For example, if the patient is currently judged to have a "heat syndrome", and the system retrieves a clear "cold syndrome" result from 2 months ago, and there are no major treatment intervention records during that period, then the judgment of "from cold to heat" can be supported; conversely, if the historical results exceed the preset time range or the records are missing, the conversion syndrome judgment logic will not be initiated.

[0078] The cold and heat syndrome discrimination network 1033 is used to extract the historical tongue features and historical joint features corresponding to the historical cold and heat syndrome prediction results, and to compare them with the current tongue features and current joint features to obtain the change in the current tongue features and the change in the joint features.

[0079] Specifically, when the system identifies a patient with valid historical records within a preset time range, the cold and heat syndrome discrimination network 1033 can access historical tongue and joint images. The historical features are compared item by item with the current tongue and joint features extracted from the current patient's visit, and the differences in features are calculated to obtain the changes in tongue features (e.g., tongue color changes from pale white to red, coating thickness changes from thick to thin) and joint features (e.g., redness and swelling area decreases, skin temperature decreases).

[0080] Specifically, when the current preliminary cold / heat syndrome prediction result obtained based on tongue and joint features differs from the historical cold / heat syndrome prediction results stored in the storage module 105 in terms of cold / heat attributes; for example, if the historical result was "cold syndrome" and the current result is "heat syndrome," or if the historical result was "heat syndrome" and the current result is "cold syndrome," it is determined to be a cold / heat transition syndrome. In some embodiments, the changes in tongue features (such as the degree of difference in feature vectors such as tongue color, coating color, and dryness / moisture) and joint features (such as the magnitude of changes in the proportion of redness and swelling area, skin temperature distribution, and degree of swelling) can also be calculated. When both exceed their respective preset judgment thresholds, it is determined to be a cold / heat transition syndrome. The judgment thresholds can be set based on statistical analysis of clinical data. Only when the syndrome type undergoes a clear change and the changes in both the tongue and joint features reach a clinically identifiable significant level, will the system ultimately determine it as a "cold / heat transition syndrome," and further determine the patient's condition as either "cold syndrome turning into heat syndrome" or "heat syndrome turning into cold syndrome" based on the current result and historical results.

[0081] In other embodiments, many arthritis patients do not present with simple cold or heat, but rather with a complex state of both, such as alternating cold and heat syndrome. For example, a patient may have a red tongue with a yellow, greasy coating (heat sign), but cold, painful joints in the limbs that prefer warmth (cold sign); or local joints may be red, swollen, and hot, while the overall constitution is weak and cold. This alternating cold and heat syndrome is particularly common in patients with chronic diseases, long-term illnesses, or those who have undergone treatment. Mechanically categorizing it as either a "cold syndrome" or a "heat syndrome" can easily lead to incorrect treatment approaches.

[0082] Optionally, the tongue features include, but are not limited to: tongue body color parameters, such as the RGB or HSV values ​​of the tongue color, used to distinguish between pale white tongue, red tongue, and dark purple tongue; tongue coating color parameters, such as the chroma and saturation of the coating color, used to identify white coating, yellow coating, and gray-black coating; and tongue coating moisture, wherein the degree of moisture or dryness of the tongue surface is assessed through image texture analysis to reflect the abundance or deficiency of body fluids and the thickness of the tongue coating, and the thickness of the coating is determined based on the pixel density or contrast of the tongue coating area. These features are identified and output by the tongue image feature extraction network 1031, providing objective evidence for the differentiation of cold and heat syndromes. For example, a pale white tongue with a white and greasy coating often indicates a cold syndrome, while a red tongue with a yellow and dry coating tends to indicate a heat syndrome.

[0083] Optionally, the joint features include, but are not limited to: the proportion of red and swollen area in the joint region; identifying the red and swollen area through image segmentation technology to obtain the extent of inflammation; skin temperature distribution characteristics; acquiring temperature field data based on infrared thermal imaging; and the degree of swelling; assessing the degree of joint bulging and abnormal joint structural parameters through contour analysis or bilateral comparison. These joint features comprehensively characterize the inflammatory activity and lesion state, helping to differentiate between cold and heat syndromes, especially providing a basis for judging syndromes such as "heat-induced arthralgia" or "cold-induced arthralgia."

[0084] Optionally, the cold and heat syndrome discrimination network 1033 is used to determine cold and heat syndrome based on at least one of the following conditions:

[0085] 1) Detect the presence of both cold and heat symptoms in the current tongue appearance.

[0086] When the system detects that the current tongue appearance simultaneously exhibits typical cold and heat symptoms, it determines that there is a tendency for alternating cold and heat. For example, the tongue body may be pale (a characteristic of cold syndrome), but the tongue coating may be yellow and greasy or dry (a characteristic of heat syndrome); or the tongue may be red (a heat syndrome) with a white and slippery coating (a cold syndrome).

[0087] 2) The proportion of redness and swelling area and the distribution of skin temperature in the joint features exceed the threshold of heat syndrome, and the abnormal joint morphology parameters meet the criteria of cold syndrome.

[0088] When the local inflammation indicators of the joint reach the criteria for heat syndrome, while the overall manifestations are consistent with the characteristics of cold syndrome, it indicates an alternating pattern of heat and cold.

[0089] 3) The coexistence ratio of cold symptoms and heat symptoms exceeds the preset overlap threshold.

[0090] Specifically, the system scores the extracted tongue and joint features for cold / heat attributes, calculating the proportion of components belonging to "cold symptoms" and "heat symptoms" respectively. When the coexistence of the two types of features (such as the ratio of heat symptoms score to cold symptoms score being close to 1:1, or both being higher than the set level) exceeds the preset "interlacing threshold", it is determined to be a cold-heat interlacing syndrome.

[0091] Optionally, the current medical information also includes the patient's clinical inflammatory markers, such as laboratory test data for C-reactive protein (CRP), erythrocyte sedimentation rate (ESR) (which may also include the ratio of CRP to ESR), rheumatoid factor (RF), and anti-CCP antibodies. Figure 4 In the illustrated embodiment, the diagnostic system 100 further includes a severity quantification module 106 for assessing the severity of a patient's symptoms to obtain a severity index. The severity quantification module 106 is based on extracted tongue features (such as the degree of tongue redness and the thickness of the yellow coating) and joint features (such as the percentage of red and swollen area, the range of increased skin temperature, and the degree of swelling), combined with clinical inflammatory indicators. For example, by weighted fusion of tongue color shift, joint hot zone area, and CRP level, a severity index reflecting the intensity of "heat symptoms" is calculated; or the severity of cold symptoms is assessed based on the degree of tongue paleness and the extent of decreased joint skin temperature. The severity indices for each symptom are output numerically and can be classified into mild, moderate, and severe levels.

[0092] Optionally, in Figure 4 In the illustrated embodiment, the diagnostic system 100 further includes a prescription generation module 107, used to match and generate corresponding traditional Chinese medicine treatment plans based on the cold and heat syndrome judgment results (such as cold syndrome, heat syndrome, cold and heat mixed syndrome, or cold and heat transition syndrome) output by the judgment module 102 and the symptom severity index generated by the cold and heat syndrome severity quantification module 106. In some embodiments, the prescription generation module 107 includes a traditional Chinese medicine syndrome differentiation and treatment knowledge base, which pre-stores classic prescriptions corresponding to various syndrome types (such as aconite decoction for cold syndrome, four wonders pill for heat syndrome, and cinnamon twig, peony and anemarrhena decoction for cold and heat mixed syndrome, etc.); in another embodiment, the traditional Chinese medicine syndrome differentiation and treatment knowledge base can also be set outside the prescription generation module 107, and the prescription generation module 107 calls the traditional Chinese medicine syndrome differentiation and treatment knowledge base and adjusts the drug composition and dosage according to the severity index. For example, for patients with "severe heat syndrome," the system increases the dosage or number of heat-clearing and detoxifying herbs (such as honeysuckle and phellodendron bark) in the basic prescription; for "mild cold syndrome," the dosage of warming herbs (such as aconite and cinnamon twig) is appropriately reduced to avoid excessive dryness and damage to yin. The prescription generation process can also perform safety checks based on the patient's age, weight, liver and kidney function, etc., to avoid overdosing or contraindications. The final output is a traditional Chinese medicine prescription, including the name of the herbs, dosage, usage, and decoction suggestions.

[0093] Optionally, in Figure 4 In the illustrated embodiment, the prescription generation module 107 can also directly determine the corresponding treatment prescription and drug dosage based on the diagnosis of cold and heat syndromes (such as cold syndrome, heat syndrome, alternating cold and heat syndrome, or cold and heat transition syndrome). For example, "cold syndrome" corresponds to Wutou Tang, "heat syndrome" corresponds to Simiao Wan, "alternating cold and heat syndrome" corresponds to Guizhi Shaoyao Zhimu Tang, and "cold and heat transition syndrome" is matched with a transitional prescription with harmonizing and transforming effects. After determining the prescription, the module sets the baseline dosage of each drug according to the standard prescription specifications and can make reasonable adjustments based on individual information such as the patient's age and weight to ensure the safety and personalization of medication.

[0094] The above embodiments are merely illustrative of the principles and effects of this disclosure and are not intended to limit this disclosure. Any person skilled in the art can modify or alter the above embodiments without departing from the spirit and scope of this disclosure. Therefore, all equivalent modifications or alterations made by those skilled in the art without departing from the spirit and technical concept disclosed in this disclosure should still be covered by the protection scope of this disclosure.

Claims

1. A diagnostic system for arthritis with chills and fever, characterized in that, The diagnostic system includes: The information acquisition module is used to acquire the patient's current medical information, which includes the patient's tongue image and joint images; the joint images include infrared joint images and color joint images. The judgment module is used to obtain the judgment result of cold and heat syndrome based on the cold and heat syndrome model; the cold and heat syndrome model includes: a tongue image feature extraction network, a joint image feature extraction network, and a cold and heat syndrome discrimination network; the tongue image feature extraction network is used to extract tongue image features from the tongue image, the joint image feature extraction network is used to extract joint features, and the cold and heat syndrome discrimination network is used to obtain the current cold and heat syndrome prediction result based on the combination of tongue image features and joint features; The tongue image feature extraction network is trained using tongue images; The joint image feature extraction network includes: a first joint detection network trained on infrared joint images and a second joint detection network trained on color joint images; Therefore, the cold and heat syndrome discrimination network is trained by using training data composed of tongue images and joint images with cold and heat syndrome labels; The predicted results of cold and heat syndromes include at least one of the following: cold syndrome, heat syndrome, cold-heat transition syndrome, and cold-heat alternation syndrome; the cold-heat transition syndrome is determined by comparing the current preliminary cold and heat syndrome prediction results obtained based on the tongue appearance and joint features with the historical cold and heat syndrome prediction results when the patient has historical cold and heat syndrome prediction results in the diagnostic system, and discovering changes in cold / heat syndrome. The cold and heat syndrome discrimination network is used to determine cold and heat syndrome based on at least one of the following conditions: 1) The current tongue appearance features both cold and heat symptoms; wherein, the cold symptoms of the tongue appearance include a pale white tongue body and a white and slippery tongue coating; the heat symptoms of the tongue appearance include a yellow and greasy or dry tongue coating and a red tongue body; 2) The proportion of redness and swelling area and the skin temperature distribution characteristics of the joint exceed the threshold for heat syndrome, and the abnormal joint morphology parameters meet the criteria for cold syndrome; wherein, the proportion of redness and swelling area and the skin temperature distribution characteristics exceeding the threshold for heat syndrome are heat syndrome characteristics of the joint; and the abnormal joint morphology parameters meeting the criteria for cold syndrome are cold syndrome characteristics of the joint. 3) The coexistence ratio of cold symptoms and heat symptoms exceeds the preset overlap threshold.

2. The diagnostic system according to claim 1, characterized in that, The tongue features include tongue color parameters, tongue coating color parameters, tongue coating moisture and thickness; and / or, the joint features include the percentage of red and swollen area in the joint region, skin temperature distribution characteristics, degree of swelling and abnormal joint structure parameters.

3. The diagnostic system according to claim 1, characterized in that, The information acquisition module is also used to acquire interference information that affects the characteristics of the tongue image. The interference information includes behaviors that cause changes in the tongue image within a preset time period before the tongue image acquisition time; the body state information represented by the changed tongue image deviates from the patient's actual body state information. The diagnostic system also includes an image confirmation module, which is used to confirm the availability of the tongue image based on the interference information, and outputs the confirmation to the judgment module when the tongue image is in a usable state.

4. The diagnostic system according to claim 3, characterized in that, It also includes a prompting module for generating acquisition guidance information when tongue images and / or joint images are unavailable.

5. The diagnostic system according to claim 1, characterized in that, It also includes a storage module for storing the patient's historical cold and heat syndrome prediction results and corresponding tongue and joint images; when the judgment module generates the current cold and heat syndrome prediction result, it calls the historical cold and heat syndrome prediction results in the storage module for comparison and analysis.

6. The diagnostic system according to claim 1, characterized in that, The judgment module calls the stored historical cold and heat syndrome prediction results, wherein the historical cold and heat syndrome prediction results are at least one cold or heat syndrome result generated by the system within a preset time interval for the patient; The cold and heat syndrome discrimination network is used to extract the historical tongue features and historical joint features corresponding to the historical cold and heat syndrome prediction results, and compare them with the current tongue features and current joint features to obtain the change in the current tongue features and the change in the joint features. Among them, when the current preliminary cold / heat syndrome prediction result and the historical cold / heat syndrome prediction result indicate different cold / heat syndromes, and the changes in tongue appearance characteristics and joint characteristics both exceed their respective preset thresholds, it is determined to be a cold / heat transition syndrome. The transition type of the cold / heat transition syndrome includes cold syndrome turning into heat syndrome or heat syndrome turning into cold syndrome.

7. The diagnostic system according to claim 1, characterized in that, The current medical information also includes clinical inflammatory marker information; the diagnostic system also includes a module for quantifying the severity of chills and fever, used to obtain a severity index of the corresponding symptoms based on the tongue appearance and joint features.

8. The diagnostic system according to claim 7, characterized in that, It also includes a prescription generation module, used to determine a prescription for treatment and the dosage of drugs in the prescription based on the results of the cold and heat syndrome assessment and the severity index.

9. The diagnostic system according to claim 1, characterized in that, It also includes a prescription generation module, which is used to determine the prescription for treatment and the dosage of the drugs in the prescription based on the diagnosis of cold and heat syndrome.

Images