Visceral function detection device, control method thereof, diet management method and system

Abstract

This application relates to a device for detecting organ function and its control method, as well as a dietary management method and system, belonging to the technical field of medical device testing. The control method compares and analyzes changes in the microscopic features of the same area of ​​the tongue before and after eating, and calculates the corresponding rate of change in metabolic index, constructing a mapping relationship between the tongue's microscopic morphology and organ function, providing a basis for assessing internal physiological function through external characterization. Simultaneously, the dietary management method uses physiological responses before and after eating as real-time biofeedback signals to generate dynamically adjusted dietary management plans, forming a personalized health management closed loop of "monitoring-assessment-intervention-remonitoring," achieving precise adaptive adjustment of intervention strategies.

Description

Technical Field

[0001] This application relates to the technical field of medical device testing, and in particular to a device for detecting organ function and its control method, as well as a dietary management method and system. Background Technology

[0002] Traditional Chinese medicine tongue diagnosis relies on the physician's subjective experience and lacks standardized data output, making remote diagnosis difficult. Existing digital tongue diagnosis equipment mostly uses ordinary macro lenses and ring light sources, primarily limited to identifying the macroscopic features of the entire tongue. It cannot clearly distinguish the three-dimensional microscopic morphology of filiform and fungiform papillae on the tongue surface, making it difficult to objectively differentiate between physiological wear and pathological atrophy. Meanwhile, routine nutritional assessments mainly rely on blood biochemistry tests or subjective scales; the former is invasive and lagging, while the latter lacks objective data support. More critically, current technologies have failed to establish a quantitative correlation model between the tongue's microstructure and organ function, and nutritional intervention plans are mostly static recommendations, lacking the dynamic closed-loop regulatory capacity based on immediate physiological feedback after eating, thus hindering the realization of precise nutritional support. Summary of the Invention The main purpose of this application is to provide a device for detecting organ function and its control method, as well as a dietary management method and system, in order to solve the technical problems in the prior art, such as inaccurate digestive function assessment and poor personalization and adaptability of nutritional support programs, caused by insufficient recognition ability of tongue micromorphology, lack of quantitative correlation model between tongue image and digestive function, and lack of dynamic closed-loop feedback mechanism in nutritional intervention programs.

[0003] To achieve the above objectives, this application proposes a control method for an organ function testing device, comprising: Microscopic images were acquired at a first time point and a second time point, respectively; wherein, the first time point is the time point of basal metabolic state, and the second time point is the time point after the user ingested a preset food load, and the microscopic images at the two time points are microscopic images of the same target area on the user's tongue; Image processing is performed on the microscopic images at two time points to obtain the first microscopic feature parameters corresponding to the first time point and the second microscopic feature parameters corresponding to the second time point; wherein, the microscopic feature parameters include at least a three-dimensional morphology factor for characterizing the three-dimensional morphology of the tongue papilla. The first microscopic feature parameter and the second microscopic feature parameter are respectively input into a preset efficiency index model to calculate the basal metabolic index corresponding to the first time point, the load metabolic index corresponding to the second time point, and the rate of change of the load metabolic index relative to the basal metabolic index. Based on the rate of change and the first and second microscopic feature parameters, the user's organ functions are determined, and detection results are generated.

[0004] In one embodiment, the specific steps for acquiring the microscopic images at the first time point and the second time point include: Use a macro imaging device to align with a preset area in the middle region of the tongue; A cross-polarized light source was turned on for supplemental lighting, and polarization filtering was used to eliminate specular reflections caused by liquid on the tongue surface to obtain a texture image of the basal tissue of the tongue papillae. The first specific wavelength light source and the second specific wavelength light source were alternately turned on to acquire hemoglobin absorption images reflecting the microvascular filling degree and autofluorescence images reflecting the degree of tongue coating keratinization, respectively.

[0005] In one embodiment, the first specific wavelength light source is a narrowband green light source, and the second specific wavelength light source is an ultraviolet light source; the specific steps of performing image processing on the microscopic images at the two time points to obtain the first microscopic feature parameter corresponding to the first time point and the second microscopic feature parameter corresponding to the second time point include: Based on the texture image, extract the height, density, or volume data of the lingual papillae, and calculate the density parameter of the filiform papillae and the stereomorphic factor. Microvascular filling parameters are extracted based on the hemoglobin absorption image, keratinization degree parameters are extracted based on the autofluorescence image, and aberrant / target region ratio parameters are generated based on the keratinization degree parameters. The microscopic feature parameters of the corresponding state are formed by combining the density of filamentous papillae, three-dimensional morphology factor, microvascular filling degree parameter, keratinization degree parameter, and atypical / target region ratio parameter extracted at the same time point.

[0006] In one embodiment, the specific steps of inputting the first microscopic feature parameter and the second microscopic feature parameter into a preset efficiency index model to calculate the basal metabolic index corresponding to the first time point, the load metabolic index corresponding to the second time point, and the rate of change of the load metabolic index relative to the basal metabolic index include: After normalizing the micro-feature parameters at the first and second time points respectively, the parameters are input into a preset efficiency index model to calculate and output the basal metabolic index at the first time point and the load metabolic index at the second time point. Based on the basal metabolic index, the metabolic load index, and the preset rate of change formula, the rate of change is calculated; wherein, the preset rate of change formula is: ΔE = (DMEI1- DMEI0) / DMEI0; Where ΔE is the rate of change, DMEI0 is the basal metabolic index, and DMEI1 is the load metabolic index.

[0007] In one embodiment, the preset efficiency index model is a nonlinear weighted detection model based on support vector regression, used to map multiple tongue surface microstructure parameters to a digestive metabolic efficiency index (DMEI) in the range of 0 to 100, expressed as: DMEI = f (w1) D fp w2 F 3D w3 I perf -w4 R atypical ); Wherein, DMEI is the metabolic index, D fp F is the density parameter of the lingual papillae. 3D For 3D morphology factor, I perf R is a parameter for microvascular filling. atypical The percentage of heterogeneous / target regions is a parameter, where w1, w2, w3, and w4 are preset weight parameters, and w4 > 0. The metabolic index maps the barrier function of the digestive system mucosa and the level of protease secretion.

[0008] In one embodiment, the specific steps for determining the user's organ functions based on the rate of change and the first and second microscopic feature parameters include: If the rate of change is >0, and the microvascular filling parameter at the second time point is higher than the microvascular filling parameter at the first time point, then the user's organ condition is determined to be good. If the rate of change is less than or equal to 0, or if the percentage of the irregular / target region at the second time point is higher than the percentage of the irregular / target region at the first time point, then the user's organ condition is determined to be poor.

[0009] Furthermore, to achieve the above objectives, this application proposes a dietary management method, comprising: The control method of the organ function detection device described above is used to detect the user's organ function and obtain the detection results; Based on the test results, a dietary management plan corresponding to the current dietary structure is generated.

[0010] In one embodiment of a diet management method, the specific steps for generating a corresponding diet management plan based on the detection results include: If the test results indicate that the internal organs are in good condition, the current diet is determined to be well tolerated, and a diet management plan is generated that includes increasing food concentration and / or increasing the molecular weight of the ingested nutrients. If the test results indicate poor organ function, the current diet will be classified as dietary damage, and a dietary management plan will be generated that includes reducing food concentration and / or reducing the molecular weight of ingested nutrients, or adjusting solid foods to a liquid / semi-liquid diet.

[0011] In addition, to achieve the above objectives, this application proposes an organ function detection device, comprising: a processor and a memory, wherein the memory is used to store a computer program, and the processor executes the computer program to implement the steps of the control method for the organ function detection device as described above.

[0012] Furthermore, to achieve the above objectives, this application proposes a diet management system, comprising: An optical acquisition device is used to acquire microscopic images of the same target area on the user's tongue at a first time point and a second time point. Such as the organ function testing equipment mentioned above; The controller is electrically connected to the organ function detection device and is used to implement the steps of the diet management method described above. The terminal is connected in communication with the controller and is used to output the detection results of the organ functions and / or dietary management plans.

[0013] One or more technical solutions proposed in this application have at least the following technical effects: This application constructs a mapping relationship between the microscopic features of the same area of ​​the tongue before and after a user eats, and calculates the corresponding metabolic index change rate, thus providing a basis for assessing internal physiological functions through external characterization. Simultaneously, using physiological responses before and after eating as real-time biofeedback signals, it generates dynamically adjusted dietary management plans, forming a personalized health management closed loop of "monitoring-assessment-intervention-remonitoring," achieving precise adaptive adjustment of intervention strategies. Attached Figure Description

[0014] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this application and, together with the description, serve to explain the principles of this application.

[0015] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, for those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0016] Figure 1 This is a flowchart illustrating an embodiment of a control method for a device for detecting organ function according to this application; Figure 2This is a schematic flowchart of step S100 provided in an embodiment of the control method for an organ function testing device according to this application; Figure 3 This is a schematic flowchart of step S200 provided in an embodiment of the control method for a device for detecting organ function according to this application; Figure 4 This is a schematic flowchart of step S300 provided in an embodiment of the control method for an organ function testing device according to this application; Figure 5 This is a schematic flowchart of step S400 provided in an embodiment of the control method for an organ function testing device according to this application; Figure 6 This is a flowchart illustrating an embodiment of a diet management method according to this application; Figure 7 This is a schematic diagram of step S500 provided in an embodiment of a diet management method according to this application.

[0017] The purpose, features, and advantages of this application will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0018] It should be understood that the specific embodiments described herein are merely illustrative of the technical solutions of this application and are not intended to limit this application.

[0019] To better understand the technical solution of this application, a detailed description will be provided below in conjunction with the accompanying drawings and specific implementation methods.

[0020] Traditional Chinese medicine tongue diagnosis relies on the physician's visual observation and accumulated personal experience, essentially a qualitative analysis. This model's judgment of key aspects of the tongue's appearance, such as the presence or absence of a "rooted" tongue coating, often results in significant individual differences due to the physician's academic school of thought, clinical experience, and even the condition at the time, making it difficult to establish unified, quantifiable diagnostic standards. This subjectivity and ambiguity not only limit its standardized application in modern medicine but also hinder its integration into digital scenarios such as telemedicine, thus restricting the transmission of knowledge and large-scale service provision.

[0021] To overcome the limitations of traditional methods, various digital tongue diagnosis devices have emerged on the market. However, most existing devices are only equipped with ordinary macro lenses and general ring light sources, focusing on the extraction of macroscopic features of the entire tongue, such as overall tongue color, coating color, and the presence or absence of teeth marks or cracks. While these solutions achieve preliminary image digitization, they are limited by hardware resolution and imaging principles, making it impossible to clearly capture the details of the tongue's microscopic structure. A key technical bottleneck lies in their inability to distinguish the specific three-dimensional morphology of individual filiform papillae and fungiform papillae. This lack of distinguishing ability makes it difficult for the system to differentiate between two seemingly similar appearances: one is the temporary "peeling" of the tongue coating caused by physical friction, which is a physiological change; the other is the "mirror tongue" caused by the atrophy, flattening, or even disappearance of the tongue papillae themselves, which is a pathological manifestation. In traditional Chinese medicine theory, this corresponds to the basis for judging whether the tongue coating is "rooted" or "unrooted." Because existing technologies cannot perform quantitative measurement of microscopic morphology, they cannot objectively distinguish between the two, significantly reducing the accuracy of diagnosis.

[0022] Meanwhile, in the field of clinical nutritional assessment, conventional methods mainly rely on blood biochemical marker tests or subjective overall assessment scales. While blood tests can provide specific protein data, they are invasive, and changes in these markers often lag behind real-time changes in physiological state, reflecting nutritional status over a past period rather than immediate functional status. Subjective assessment scales, on the other hand, heavily rely on the assessor's experience and lack objective, continuous physiological data support. Neither of these methods establishes a direct, dynamic link with the immediate functional state of the digestive system.

[0023] The core shortcomings of existing technologies are thus highlighted: First, at the image level, insufficient resolution leads to misjudgment of microscopic morphology, making it impossible to objectively quantify the essence of traditional Chinese medicine tongue diagnosis. A deeper deficiency lies in the lack of a mathematical model linking the tongue's microscopic structure to digestive function. Specifically, the density, keratinization, and microvascular fullness of the tongue papillae are theoretically closely related to deeper functions such as the integrity of the digestive system's mucosal barrier and the level of digestive enzyme secretion, but current technologies have failed to establish such a cross-level, quantifiable mapping model. Furthermore, existing nutritional intervention programs are mostly static templates, unable to dynamically adjust based on the immediate digestive and metabolic responses reflected in the tongue's appearance and other physiological indicators after a user ingests specific foods. In other words, the system cannot capture and utilize this biofeedback to optimize subsequent dietary recommendations in real time, thus failing to achieve truly personalized, adaptive metabolic regulation.

[0024] To address the above problems, this application proposes a control method for an organ function testing device, see reference. Figure 1 This includes steps S100 to S400: S100: Acquire microscopic images at the first time point and the second time point respectively; wherein, the microscopic images at the two time points are microscopic images of the same target area on the user's tongue surface; S200: Perform image processing on the microscopic images at the two time points to obtain the first microscopic feature parameter corresponding to the first time point and the second microscopic feature parameter corresponding to the second time point; wherein, the microscopic feature parameter includes at least a stereomorphic factor for characterizing the three-dimensional morphology of the tongue papilla. S300: Input the first micro-feature parameter and the second micro-feature parameter into the preset efficiency index model respectively, and calculate the basal metabolic index at the first time point, the load metabolic index at the second time point, and the rate of change of the load metabolic index relative to the basal metabolic index. S400: Based on the rate of change and the first and second microscopic characteristic parameters, the system determines the user's organ functions and generates test results.

[0025] This application can be understood as being based on holographic biology, considering the tongue mucosa and the digestive system mucosa as homologous structural units, and believing that the renewal state of the tongue epithelial cells and the level of local microcirculation perfusion can reflect the functional state of the corresponding digestive tract mucosa. This embodiment, starting from first principles, decomposes the functions of the internal organs into two types of observable and calculable basic physiological processes: Epithelial cell renewal and metabolism corresponds to the rhythm of proliferation, differentiation, and shedding of epithelial cells on the tongue surface, and is mainly indirectly characterized by the morphology of tongue papillae, the degree of keratinization, and the renewal of surface texture. Peripheral microcirculation perfusion level corresponds to the fullness, permeability, and perfusion dynamics of the tongue surface microvascular network, and is mainly characterized by indicators such as microvascular morphology, density, diameter, and color distribution. Based on these two dimensions, this embodiment uses a unified set of physical and statistical quantities to map the abstract concept of visceral function into a joint indicator of the epithelial metabolism and microcirculation dimensions, and further integrates them into the Digestive Metabolic Efficacy Index (DMEI) to describe the comprehensive functional state of the digestive system.

[0026] This application constructs a mapping relationship between the microscopic morphology of the tongue and the function of internal organs by comparing and analyzing the changes in the microscopic features of the same area of ​​the tongue before and after eating, and calculating the corresponding metabolic index change rate, thus providing a basis for evaluating internal physiological functions through external characterization.

[0027] This embodiment will be described in conjunction with the following steps S100-S400.

[0028] In step S100, the control method, through communication control with the imaging component, positions and acquires images of the user's tongue. The system first acquires a positioning image of the user's tongue, determining the spatial location of the target area using the tongue tip position, the tongue midline, or a preset reference marker. Based on this, the microscopic imaging module is driven to magnify and acquire the target area, obtaining a microscopic image at the first time point. After the user ingests a preset food load or completes a preset time interval, the user is guided to extend their tongue again in the same posture, and a new macroscopic positioning image is acquired and image registration is performed. Based on the registration result, the same target area is locked, and a microscopic image at the second time point is obtained. This method ensures that the microscopic images at the two time points correspond to the same target area on the user's tongue, reducing errors caused by shooting position deviations.

[0029] In step S200, image processing algorithms are applied to the microscopic images at the first and second time points, respectively. Image processing may include preprocessing operations such as noise suppression, brightness and color correction, and local contrast enhancement to improve the discriminability of the tongue papilla contours and surface textures. Subsequently, the tongue papilla regions are separated from the background using a segmentation algorithm, and morphological analysis, deep learning detection networks, or 3D reconstruction algorithms are further employed to identify the morphological features of each individual tongue papilla. Based on the above identification results, three-dimensional morphological factors related to the three-dimensional structure of the tongue papilla are extracted, such as tongue papilla height, base width, apical curvature, aspect ratio, and surface undulation. Optionally, tongue papilla density, surface roughness, keratinization degree, local color, and microvascular-related features are calculated. In this way, a first set of microscopic feature parameters corresponding to the first time point and a second set of microscopic feature parameters corresponding to the second time point are obtained, and the two sets are consistent in parameter dimensions.

[0030] In step S300, the first and second microscopic feature parameters are respectively input into a preset efficacy index model. The efficacy index model can be pre-trained based on multiple sets of calibration samples. The input is a vector composed of three-dimensional morphological factors and other tongue surface microscopic features, and the output is a metabolic efficacy index related to organ function. When receiving the first microscopic feature parameter, the model outputs the basal metabolic index corresponding to the first time point, used to characterize the user's organ function metabolic level under basal conditions; when receiving the second microscopic feature parameter, it outputs the load metabolic index corresponding to the second time point, used to characterize the user's organ function metabolic response under load conditions. Subsequently, the control method calculates the rate of change of the load metabolic index relative to the basal metabolic index based on the basal metabolic index and the load metabolic index, for example, by using a normalization method combining difference and ratio to generate the rate of change index. The rate of change is used to quantify the relative change in organ function efficacy before and after load.

[0031] In step S400, the user's organ function is comprehensively assessed based on the rate of change, the first microscopic feature parameter, and the second microscopic feature parameter. Specifically, the basal metabolic index, the load metabolic index, and the rate of change can be combined with a preset grading threshold or classification model to determine the level and type of the user's organ function. Simultaneously, by combining the changing trends of three-dimensional morphological factors and other microscopic features between two time points, a comprehensive analysis is conducted on whether there is atrophy of the lingual papillae structure, whether the tongue coating has a "rooted" characteristic, and whether there are abnormal reactions in local microvessels, thereby refining the assessment results of organ function status. Finally, the control method generates corresponding detection results, which may include quantitative indices, functional levels, and structural prompts, and are output through a display interface or data interface for clinical evaluation or individualized intervention strategy development.

[0032] Through the coordination of the above steps, the specific embodiment realizes the acquisition, quantitative analysis and mapping of the microstructural features of the same target area on the tongue under basic and load conditions to the efficacy of visceral functions, so that the assessment of visceral functions is transformed from qualitative experience judgment to a repeatable quantitative detection process.

[0033] In one embodiment, the specific steps of step S100, which involves acquiring the microscopic images at the first and second time points, are described in the following document. Figure 2 This includes steps S110 to S130: S110: Use a macro imaging device to align with a preset area in the middle of the tongue surface; S120: Turn on the orthogonal polarized light source for supplemental lighting, and use polarization filtering to eliminate specular reflections caused by liquid on the tongue surface, thereby obtaining a texture image of the basal tissue of the tongue papillae. S130: Alternately turn on the first specific wavelength light source and the second specific wavelength light source to acquire hemoglobin absorption images reflecting microvascular filling and autofluorescence images reflecting the degree of tongue coating keratinization, respectively.

[0034] This embodiment can be understood as achieving high-magnification imaging of the microstructure of the tongue surface through a specially designed optical module. This optical module may include: a high numerical aperture macro lens, a controllable multi-band light source array, a polarization filter assembly, and an optical path design for depth-extended imaging, etc.

[0035] The optical module's spatial resolution is configured to distinguish the contours, apical morphology, and fine surface textures of individual lingual papillae, allowing observation of subtle changes in the keratinized layer structure on the papillae, as well as the orientation and branching of some microvessels. High-precision magnified imaging of localized areas of the tongue surface is achieved by controlling the working distance and magnification. Multi-band light source configuration and polarizer combinations enhance the contrast of epithelial surface texture and deep microvessels, respectively. Multispectral imaging allows the system to acquire several microscopic images of different spectral channels at the same location, providing a basis for subsequent differentiation of keratinized layer structures and blood flow signals. By combining overall and local microscopic images of the tongue surface, along with macroscopic features such as the tongue tip contour, midline, and edges, the target detection area within the standardized tongue region is located. The optical module, in conjunction with a fixed support structure, reference scale, and guide interface, prompts the user to extend their tongue in a stable posture, ensuring regional consistency and angle repeatability during multiple imaging sessions.

[0036] In step S110, the macro imaging device is aimed at a preset area in the middle of the tongue. The system first guides the user to extend their tongue in a preset posture, keeping the tongue horizontal, straight, and under stable lighting conditions. Through the real-time image display and position assistance prompts of the imaging device, the focal length and angle of the imaging module are adjusted so that the imaging field of view covers the target area in the middle of the tongue. This area corresponds to a preset area in traditional Chinese medicine diagnosis, and its tissue characteristics are highly correlated with the functional state of the internal organs. After confirming that the imaging position and focal length are stable, the system locks the target area to ensure that the microscopic images acquired at the subsequent two time points have spatial correspondence.

[0037] In step S120, the control method activates an orthogonally polarized light source to supplement the target area. The polarized light source has two mutually perpendicular polarization directions: a transmitter and a receiver. The system significantly reduces specular reflections caused by saliva or mucosal surfaces on the tongue through polarization filtering. After polarization extinction processing, the imaging system can obtain a texture image that reflects the morphology of the tongue papillae and the true color of the underlying tissue.

[0038] In step S130, the control method alternately activates the first specific band light source and the second specific band light source to perform band-selective imaging of the same target area. The first specific band light source is narrow-band green light, located near the main absorption peak of hemoglobin, which can enhance the contrast between the microvascular network and surrounding tissues, thereby forming a hemoglobin absorption image to characterize the filling state and distribution characteristics of the microvessels on the tongue surface. The second specific band light source is ultraviolet light, which can excite the autofluorescence signal of the tongue coating and keratinized substances, and acquire fluorescence images reflecting the degree of keratinization and metabolic accumulation of the tongue coating.

[0039] At the implementation level, this embodiment performs time synchronization control on the imaging process of orthogonal polarization and specific band light sources to ensure that images in each band are acquired based on the same spatial location, and achieves multispectral image fusion through image registration algorithms. The entire acquisition process uses a custom macro lens with a magnification of 20x to 50x, with a spatial resolution better than 10μm, which can clearly distinguish individual lingual papillae and their surface microstructure.

[0040] In the image preprocessing stage, the system performs color calibration, brightness equalization, and geometric correction on the acquired raw image sequence. In the fusion processing stage, it combines an algorithm of orthogonal polarization dereflection and specific wavelength band (540nm / 365nm) enhancement display to improve the accuracy of subsequent image segmentation and feature extraction. After this combined processing, the boundary recognition rate of the lingual papillae is significantly improved, the signal-to-noise ratio of the microvascular image is enhanced, and the fluorescence signal in the keratinized region is more concentrated and clear.

[0041] Ultimately, the system obtained three types of image data of the mid-tongue region that can be used for microscopic analysis: Texture images under orthogonally polarized light are used for color analysis of the three-dimensional structure and basal tissue of the lingual papillae; Hemoglobin absorption images at 540nm wavelength are used for microvascular filling and blood flow status analysis. Autofluorescence images at 365nm were used for the analysis of keratinization degree and metabolite accumulation.

[0042] In one embodiment, step S200 involves image processing of the microscopic images at two time points to obtain the first microscopic feature parameter corresponding to the first time point and the second microscopic feature parameter corresponding to the second time point. For details, please refer to [link to relevant documentation]. Figure 3 This includes steps S210 to S230: S210: Extract the height, density, or volume data of the lingual papillae based on the texture image, and calculate the density parameters and three-dimensional morphology factors of the filiform papillae. S220: Extract microvascular filling parameters based on hemoglobin absorption images, extract keratinization parameters based on autofluorescence images, and generate aberrant / target region proportion parameters based on keratinization parameters; S230: Combine the filamentous papillary density, stereomorphology factor, microvascular filling parameter, keratinization degree parameter, and heteromorphic / target region ratio parameter extracted at the same time point to form the microscopic feature parameters of the corresponding state.

[0043] This embodiment can be understood as follows: it uses a deep learning model to automatically segment the tissue structures in a microscopic image of the tongue surface, distinguishing different types of microstructures and generating high-level semantic annotation results. The deep learning segmentation model identifies and segments at least one of the following structures in the microscopic image: filamentous papillae region, fungiform papillae region, epithelial keratinization enhancement region, microvascular imaging region, tongue coating coverage area, and exposed mucosa area.

[0044] The training data consists of manually labeled microscopic images of the tongue surface, including annotations of papilla boundaries, various papilla types, blood vessel orientation, and areas of varying keratinization thickness. Through supervised learning, the model automatically outputs pixel-level segmentation results for corresponding tissue categories on new images. The segmentation maps output by the model undergo connected component analysis, small-region denoising, and boundary smoothing to obtain clear contours and structural labels for each lingual papilla unit, each microvascular branch, and various epithelial regions. Through deep learning segmentation, this embodiment transforms the original pixel-level images into structural-level data, enabling subsequent calculation of cell viability and microcirculation perfusion parameters at the tissue unit level.

[0045] Based on the segmented lingual papillae and epithelial regions, quantitative parameters reflecting the epithelial cell renewal and metabolism capacity were extracted. These parameters were mainly constructed from three aspects: three-dimensional morphology, surface texture, and keratinization characteristics.

[0046] Three-dimensional morphological indices of the lingual papillae are calculated based on multispectral and multi-angle imaging, combined with shape reconstruction or illumination-shadow analysis. For each lingual papilla unit, the following parameters are calculated: papilla height to base diameter ratio, apical curvature and surface convexity, papilla density (number of papillae per unit area), and papilla morphological integrity (degree of defect, atrophy, and flattening). These indices reflect the proliferation and maintenance capacity of epithelial cells in the papillary region. When the papilla morphology is full, the density is stable, and the height and curvature are distributed within a specific range, it can be inferred that the epithelial cell renewal in that region is in a relatively good state. When the overall papilla height decreases, the density decreases, or a large number of papillae are flattened, it can be inferred that the epithelial renewal and metabolic capacity is weakening.

[0047] Texture analysis algorithms are used to statistically analyze the fineness, directionality, and consistency of epithelial surface texture. By combining data from consecutive time points, the update rate and stability of texture patterns can be analyzed, indirectly reflecting the rhythmic characteristics of epithelial cell shedding and regeneration.

[0048] Within the segmented keratinized enhanced region, optical density, mean brightness, and reflection characteristics related to keratin thickness are calculated and superimposed with papillary distribution to form a "keratinization-papillary joint feature". When papillary structures are present but the degree of keratinization is temporarily weakened, and the texture in the same area still maintains a certain continuity, it can be judged as temporary lichenification; when the papilla itself atrophies or disappears and is accompanied by long-term weakening of keratinization, it can be judged as long-term insufficient epithelial nutrition and renewal capacity.

[0049] The above parameters together constitute the feature vector of the epithelial cell renewal and metabolic capacity dimension.

[0050] Furthermore, this embodiment extracts quantitative parameters describing the perfusion status of peripheral microcirculation based on the segmented microvascular regions and color distribution characteristics. These parameters include microvascular structural indices, statistical analysis within the microvascular imaging area, microvascular branch density and frequency, trunk and branch vessel diameter distribution, vessel tortuosity and course consistency, and microvascular network connectivity. These structural features are related to microcirculation patency and vascular adaptability. When perfusion is sufficient and stable over a long period, the microvascular network structure maintains a certain regularity and connectivity. When predictable changes in the red component of the same region before and after a load occur within the normal range, it can be inferred that the microcirculation's response to digestive load stimulation remains within a relatively stable range; when the changes are abnormally amplified or weakened, it indicates that the microcirculation's regulatory capacity deviates from the normal range.

[0051] In step S210, height, density, or volume data of the lingual papillae are extracted based on the texture image to calculate the filiform papilla density parameters and stereomorphic factors. The system uses an enhanced image segmentation network to perform instance-level segmentation of the texture image, identifies and labels individual filiform papilla regions, and uses a connection domain algorithm to count the number of papillae per unit area, thereby obtaining the filiform papilla density D. fp This parameter reflects the neonatal and metabolic activity of the lingual papilla tissue. Subsequently, the system reconstructs the local height distribution of the papilla using shape from shading or multi-angle illumination, and calculates the aspect ratio H / W as the three-dimensional morphology factor F based on the papilla height (H) and base width (W). 3D F 3D This parameter quantifies the upright or collapsed state of the nipple, with different value ranges corresponding to different tissue support and nutritional status. It digitally represents the concept of "rooted / rootless tongue coating" in traditional Chinese medicine tongue diagnosis, providing a calculable objective indicator to distinguish between temporary nipple coverage or slight flattening (physiological change) and nipple atrophy or collapse (pathological change), thus solving the problem that traditional methods cannot quantitatively differentiate between the two.

[0052] In step S220, microvascular and keratinization-related parameters are extracted based on multispectral images. Based on hemoglobin absorption images, the system extracts component brightness distribution in color space, and calculates local vessel density and red signal intensity using a vascular texture detection operator to obtain the microcirculation perfusion index I. perf This parameter reflects the fullness of the microvascular network and the level of blood perfusion in the tongue tissue, and can objectively describe the state of blood flow and nutrient delivery within the internal organs.

[0053] Secondly, based on the autofluorescence image, the system performs threshold segmentation on the bright areas to extract the keratinized layer distribution area, and performs registration calculation with the stroma layer area to obtain the keratinization-stromal ratio R. km Parameter R km This is the ratio of the area or optical density of the keratinized layer to the area or optical density of the matrix layer, used to quantify the thickness, accumulation level, and degree of metabolic waste accumulation of the tongue coating. The aberrant / target area ratio parameter R is further calculated based on the distribution of keratinized areas. atypical This is used to measure the uniformity of keratinization distribution and the proportion of abnormal areas. Through the above processing, key parameters reflecting two dimensions of keratinization and blood supply on the tongue surface are obtained.

[0054] In step S230, D extracted at the same time point fp F 3D I perf R km and the parameter R representing the proportion of irregular / target areas. atypical These parameters are combined to form a set of microscopic feature parameters for the corresponding time points. This set of feature parameters includes structural dimensions (nipple density and three-dimensional morphology), hemodynamic dimensions (microcirculation perfusion), and metabolic keratinization dimensions (keratinization degree and distribution), forming a multi-channel feature matrix for subsequent input into the efficacy index model.

[0055] In this specific embodiment, the proposed stereomorphic factor F 3D As a quantitative indicator of the geometric attributes of the lingual papillae, this method transforms the three-dimensional morphology of the tongue surface structure into numerical values ​​by calculating the height-to-width ratio. This establishes a correlation between qualitative descriptions in traditional Chinese medicine tongue diagnosis, such as "rooted / rootless tongue coating," and modern image-based quantitative indicators. Compared to traditional methods that rely solely on color or texture, this method can distinguish between the erect, flat, or absent states of the papillae, thereby differentiating between temporary changes in the tongue coating caused by physical friction and persistent, smooth changes resulting from papillary structural degeneration. This improves the structural resolution and accuracy of organ function assessment.

[0056] In one embodiment, step S300 involves inputting the first microscopic feature parameter and the second microscopic feature parameter into a preset performance index model to calculate the basal metabolic index at the corresponding first time point, the load metabolic index at the corresponding second time point, and the rate of change of the load metabolic index relative to the basal metabolic index. For details, please refer to [link to relevant documentation]. Figure 4 This includes steps S310 to S320: S310: After normalizing the micro-feature parameters at the first time point and the second time point respectively, input them into the preset efficiency index model, calculate and output the basal metabolic index corresponding to the first time point and the load metabolic index corresponding to the second time point. S320: Calculate the rate of change based on the basal metabolic index, the load metabolic index, and the preset rate of change formula; where the preset rate of change formula is: ΔE = (DMEI1 - DMEI0) / DMEI0, where ΔE is the rate of change, DMEI0 is the basal metabolic index, and DMEI1 is the load metabolic index.

[0057] This can be understood as follows: In a specific embodiment, step S300 involves inputting the first microscopic feature parameter and the second microscopic feature parameter into a preset efficiency index model for calculation, such as... Figure 4 As shown, steps S310 to S320 may be included, as detailed below: In step S310, the control method normalizes the microscopic feature parameters at the first and second time points respectively, and then inputs them into a preset efficacy index model. Normalization can employ interval scaling or standardization methods to map feature parameters of different dimensions to a uniform numerical range, thereby reducing the impact of dimensional differences on model calculations. The preset efficacy index model is used to establish a nonlinear mapping relationship between tongue surface microstructural parameters and the metabolic state of visceral functions. This model is a weighted detection model based on support vector regression (SVR). The model input includes the tongue papilla density parameter D. fp 3D morphology factor F 3D Microvascular filling parameter I perf and the parameter R representing the proportion of irregular / target areas. atypical The model output is the Digestive and Metabolic Efficiency Index (DMEI).

[0058] The basic form of the model can be expressed as: DMEI = f (w1) D fp w2 F 3D w3 I perf -w4 R atypical ); Wherein, DMEI is the metabolic index, D fp F is the density parameter of the lingual papillae. 3D For 3D morphology factor, I perf R is a parameter for microvascular filling. atypical The parameter represents the proportion of the heterogeneous / target region. w1, w2, w3, and w4 are model weight parameters, with w4 > 0. The metabolic index maps the barrier function of the digestive system mucosa and the level of protease secretion. The DMEI is output as a continuous value from 0 to 100, corresponding to the comprehensive effectiveness of the tongue surface microstructure, digestive system mucosal barrier function, and protease secretion level.

[0059] The weight parameters can be obtained through joint training using a sample dataset and known organ function detection indicators. During the training phase, the model establishes a mapping between the input features and corresponding target values ​​from a large number of calibrated samples. During the detection phase, the same function form is used to calculate the parameters at the first and second time points, respectively, yielding the corresponding basal metabolic index (DMEI0) and metabolic load index (DMEI1). The basal metabolic index represents the stable level of organ function on the tongue surface under resting or fasting conditions, while the metabolic load index reflects the efficiency of the tongue surface's metabolic response after nutritional or food stimulation.

[0060] In step S320, the rate of change ΔE is calculated based on the basal metabolic index DMEI0 and the load metabolic index DMEI1, as well as a preset rate of change formula. The rate of change formula is defined as follows: ΔE = (DMEI1 - DMEI0) / DMEI0, where ΔE is used to quantify the magnitude of metabolic change in the load state compared to the basal state. When ΔE is positive and the magnitude of change is high, it indicates that the organ functions have a strong stress response and regulatory capacity; when ΔE is negative or the magnitude is low, it indicates that the organ metabolic response is weakened or that there is functional impairment.

[0061] In one specific embodiment, the efficacy index model establishes a nonlinear weighted relationship by fusing tissue structure and hemodynamic parameters, forming a DMEI nonlinear calculation model. This model breaks through the traditional single-image classification method based on color or texture judgment, simultaneously considering the structural state of the lingual papillae, microcirculation perfusion quality, and abnormal keratinization distribution, establishing a mathematical connection between the tongue's microstructure and the function of the digestive system mucosa and systemic nutritional metabolism. Through this model, the system can comprehensively evaluate lingual papilla activity, blood supply adequacy, and tissue metabolic levels, achieving a continuous mapping from image features to functional indices.

[0062] In one embodiment, step S400 determines the user's organ functions based on the rate of change and the first and second microscopic characteristic parameters, see [reference]. Figure 5 This includes steps S410 to S420: S410: If the rate of change is >0 and the microvascular filling parameter at the second time point is higher than the microvascular filling parameter at the first time point, then the user's organ condition is determined to be good. S420: If the rate of change is <= 0, or the percentage of the irregular / target area at the second time point is higher than the percentage of the irregular / target area at the first time point, then the user's organ condition is determined to be poor.

[0063] This can be understood as follows: In a specific embodiment, step S400, which determines the user's organ state based on the rate of change and the first and second microscopic feature parameters, is as follows: Figure 5 As shown, the process may include steps S410 to S420, as detailed below: In step S410, the control method uses the calculated rate of change ΔE and the microvascular filling parameter I corresponding to the two time points. perf Perform a positive judgment. When the rate of change ΔE > 0, and I at the second time point... perf I greater than the first time point perf This indicates that after eating or experiencing a nutritional load, the permeability of the microvessels on the tongue significantly increases, blood perfusion levels rise, and the color of the tongue papillae deepens and their structure becomes more clearly visible. This suggests that the user's digestive system responds normally to the food load, and that microcirculation and metabolism are in an improved state, indicating that the user's organ function is in good condition. This result represents a positive metabolic difference between the basal metabolic state and the load state, meaning that organ function has a high degree of regulatory capacity and tissue activity.

[0064] In step S420, the control method determines the ratio of the change ΔE and the proportion of the irregular / target area R. atypical Perform a reverse determination. When the rate of change ΔE ≤ 0, or R at the second time point... atypical R higher than the first time point atypical This indicates that the user did not show an increase in metabolism after ingesting a large amount of food, or even showed a decrease in the density of tongue papillae, a reduction in three-dimensional morphology factors, or an increase in tongue coating area and accumulation of surface keratin. In this case, it can be determined that the condition of the internal organs is poor. If R appears... atypical A significant increase accompanied by a decrease in ΔE indicates a state of dietary impairment or overload, i.e., a "relapse into food."

[0065] In practical applications, the system can combine ΔE and I. perf R atypical and F 3D Multidimensional judgment rules are constructed based on indicators such as: when ΔE>0 and I_perf improves, it is determined that the organs are well treated and the user can maintain or upgrade the diet formula; when ΔE≤0 or R_atypical increases, it is determined that the metabolic response is abnormal or the organs are overloaded, and the diet plan needs to be downgraded or the proportion of nutrients needs to be adjusted.

[0066] In addition, this application proposes a dietary management method, see reference. Figure 6 This includes steps S100 to S500: S100: Acquire microscopic images at the first time point and the second time point respectively; wherein, the microscopic images at the two time points are microscopic images of the same target area on the user's tongue surface; S200: Perform image processing on the microscopic images at the two time points to obtain the first microscopic feature parameter corresponding to the first time point and the second microscopic feature parameter corresponding to the second time point; wherein, the microscopic feature parameter includes at least a stereomorphic factor for characterizing the three-dimensional morphology of the tongue papilla. S300: Input the first micro-feature parameter and the second micro-feature parameter into the preset efficiency index model respectively, and calculate the basal metabolic index at the first time point, the load metabolic index at the second time point, and the rate of change of the load metabolic index relative to the basal metabolic index. S400: Based on the rate of change and the first and second microscopic characteristic parameters, the system determines the user's organ functions and generates test results; S500: Based on the detection results, generate a dietary management plan corresponding to the current dietary structure.

[0067] This can be understood as follows: In specific embodiments, the overall process of the diet management method is as follows: Figure 6 As shown, steps S100 to S500 are included. The specific execution process of steps S100 to S400 can be referred to the relevant descriptions in the foregoing embodiments, and will not be repeated here. This embodiment focuses on describing the logical framework and closed-loop control mechanism of step S500.

[0068] In step S500, the system dynamically generates a dietary management plan corresponding to the current dietary structure based on the organ function detection results obtained in step S400. The detection results include indicators such as the basal metabolic index DMEI0, the metabolic load index DMEI1, and the rate of change ΔE. The system uses ΔE as the core feedback signal to determine the user's metabolic adaptation level after eating, thereby realizing automatic adjustment of the dietary structure based on physiological response.

[0069] The specific implementation process is as follows: During the baseline assessment phase (T0), the system acquires tongue images in a fasting or resting state and calculates the initial digestive metabolic efficiency index (DMEI0). If the DMEI0 value is low (e.g., 25), the system determines that organ metabolic function is insufficient and generates a dietary instruction "Recommended intake of hydrolyzed protein liquid diet". The output of this phase is used to establish a reference model of the individual's basal metabolic level. During the load test phase (T1), 2–4 hours after the user eats according to the system's recommended plan, the system acquires tongue images again and calculates DMEI1. This value reflects the user's dynamic response capacity of organ function under food load.

[0070] The system calculates the rate of change ΔE=(DMEI1) DMEI0) / DMEI0, combined with the microvascular filling parameter I perf , percentage of irregular areas, R atypical Analyze the judgment conditions as follows: When ΔE > 0 and nipple congestion improves, the system interprets this as positive feedback, indicating good organ tolerance. The control logic automatically increases the intensity of the next stage of the diet plan, generating a diet management plan that includes increasing food concentration or nutrient molecular weight levels. For example, transitioning staple foods from hydrolyzed protein to whey protein formula.

[0071] When ΔE≤0 or the proportion of atypical areas increases (such as thickened tongue coating or blurred nipples), the system determines it as negative feedback, indicating that the current diet is causing functional overload or a "relapse" response. The control logic automatically downgrades the diet plan, generating plans that include reducing food concentration, reducing nutrient molecular weight, or adjusting solid foods to liquid / semi-liquid forms, such as converting whey protein to rice water or glucose electrolyte solution.

[0072] Through the above process, a specific embodiment establishes a control logic system based on the rate of change ΔE of the microscopic features of the tongue surface before and after eating, realizing closed-loop management of "measurement-intervention-feedback-correction". After each intake cycle is completed, the system automatically performs microscopic image acquisition and efficacy index calculation, using ΔE as a dynamic feedback signal to continuously guide the next stage of dietary decisions, transforming nutritional intervention from static template management to a real-time adaptive adjustment mode.

[0073] In one embodiment of the diet management method, the specific steps of step S500, which generates a corresponding diet management plan based on the detection results, are described in the following reference. Figure 7 This includes steps S510 to S520: S510: If the test results show that the organ condition is good, the current diet is judged to be well tolerated, and a diet management plan is generated that includes increasing food concentration and / or increasing the molecular weight of the ingested nutrients. S520: If the test results indicate poor organ function, the current diet will be classified as dietary damage, and a dietary management plan will be generated that includes reducing food concentration and / or reducing the molecular weight of ingested nutrients, or adjusting solid foods to liquid / semi-liquid diets.

[0074] This can be understood as, in a specific embodiment, step S500 being the process of generating a corresponding dietary management plan based on the detection results, such as... Figure 7 As shown, steps S510 to S520 may be included, as detailed below: In step S510, the control method first receives the organ function detection results output in step S400 and determines whether the user's organ status is good. When the detection results show a change rate ΔE greater than 0, and the microvascular filling at the second time point is higher than at the first time point, the papillary structure is stable, and there is no obvious accumulation of keratinized layer, the system determines that the user's current diet is well tolerated within the physiological range. At this time, the control method generates a corresponding diet management plan, marks the current diet as well tolerated, and, according to the model matching rules, appropriately increases the food concentration of the next meal and / or increases the molecular weight of the ingested nutrients within a safe threshold range. For example, the system can automatically adjust from a hydrolyzed protein diet to a whey protein diet, or increase the energy density and protein ratio while maintaining the main nutritional structure, to promote organ function training and improve nutrient absorption efficiency. The generated diet management plan may include nutritional categories, ratio structure, energy distribution, and monitoring cycle recommendations.

[0075] In step S520, if the detection results indicate poor organ function, the control method executes reverse adjustment logic. When ΔE is less than or equal to 0, or the percentage of aberrations / target regions at the second time point significantly increases (reflecting phenomena such as collapsed tongue papillae, thickened tongue coating, and decreased moisture), the system determines that the current dietary structure is causing excessive burden on the organs or an intolerance reaction. Based on this, the control method generates a "dietary damage" marker and outputs a downgraded dietary management plan. The downgrade logic includes: reducing food concentration, decreasing the proportion of high-molecular-weight nutrients, or adjusting solid foods to liquid or semi-liquid forms. For example, when the detection shows worsening aberrations, the system may suggest transitioning from a whey protein regimen to a hydrolyzed protein liquid regimen, or directly recommend the intake of light-load nutritional formulas such as rice water or glucose electrolyte solutions to reduce organ irritation and promote mucosal repair.

[0076] In a specific application scenario, the system can cyclically execute this positive and negative feedback mechanism, enabling the dietary adjustment process to form a closed-loop control based on the trend of ΔE change: When ΔE > 0 and nipple congestion improves, the system determines that the previous dietary intervention plan had an enhancing effect and automatically increases the dietary intensity in the next cycle. When ΔE ≤ 0 or atypical indicators rise, the system determines that a "relapse" phenomenon has occurred, automatically downgrades the plan and records intolerance markers, and then triggers a monitoring cycle shortening strategy to increase the frequency of the next assessment. Through the above steps S510 and S520, the specific embodiment realizes adaptive adjustment of the dietary plan, allowing the test results to directly drive the dynamic changes in nutritional structure and form, and constructing an integrated control logic from detection, judgment to intervention correction. The dietary management plan in this embodiment can be automatically generated by the controller and output through the terminal, allowing users or medical staff to adjust their actual dietary intake in real time, achieving precise nutrition management based on physiological feedback.

[0077] Furthermore, this application proposes an organ function detection device, comprising: a processor and a memory, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps of the control method for the organ function detection device as described above. This can be implemented using a main controller, such as a DSP (Digital Signal Processor), FPGA (Field Programmable Gate Array), MCU (Microcontroller Unit), or SOC (System on Chip).

[0078] It is worth noting that since the organ function detection device of the present invention is applied to the control method of the organ function detection device described above, the embodiments of the organ function detection device of the present invention include all the technical solutions of all embodiments of the control method of the organ function detection device described above, and the technical effects achieved are exactly the same, so they will not be repeated here.

[0079] Furthermore, this application proposes a diet management system, comprising: an optical acquisition device for acquiring microscopic images of the same target area on the user's tongue at a first time point and a second time point; an organ function detection device as described above; a controller electrically connected to the organ function detection device for implementing the steps of the diet management method as described above; and a terminal communicatively connected to the controller for outputting the detection results of organ function status and / or a diet management plan.

[0080] In a specific embodiment, the diet management system comprises an optical acquisition device, an organ function detection device, a controller, and a terminal. The system is used to achieve closed-loop control of digestive function assessment and diet management based on changes in the microstructure of the tongue surface. The specific structure and working process are as follows: The diet management system includes an optical acquisition device, an organ function detection device, a controller, and a terminal.

[0081] The optical acquisition device is used to acquire microscopic images of the same target area on the user's tongue at a first and a second time point. The device can be fixed or portable and includes a microscopic multispectral imaging module for positioning and imaging a preset area on the tongue and outputting image data to the controller. The organ function detection device is used to perform the aforementioned microscopic image analysis and metabolic index calculation of the tongue, generating the basal metabolic index, metabolic load index, and rate of change, thereby achieving quantitative detection of the user's organ function. The controller is electrically connected to the organ function detection device and is used to schedule steps such as optical acquisition, image analysis, index calculation, and diet plan generation. The controller contains a program module that can call the performance index model to perform numerical calculations. The terminal is communicatively connected to the controller and is used to receive test results and diet management plans, displaying organ function, trends, and personalized dietary prescriptions to the user or medical personnel. The terminal can be a mobile device, computer, or other display device.

[0082] Through the synergy of the above structure and functions, a specific embodiment forms a dietary management system integrating optical acquisition, intelligent analysis, and automated decision-making. This system enables non-invasive detection of organ function and dynamic optimization of nutritional plans based on microscopic changes on the tongue surface. Utilizing a closed-loop logic of "measurement-analysis-feedback-correction," the system transforms traditional experience-based dietary adjustments into a calculable and verifiable digital process, possessing real-time, adaptive, and scalable technical characteristics.

[0083] The above description is only a part of the embodiments of this application and does not limit the patent scope of this application. All equivalent structural transformations made under the technical concept of this application and using the contents of the specification and drawings of this application, or direct / indirect applications in other related technical fields, are included in the patent protection scope of this application.

Claims

1. A control method of a visceral function detection apparatus, characterized by, include: Microscopic images were acquired at a first time point and a second time point, respectively. The first time point was the time point when the user was in a fasting or resting state, and the second time point was the time point after the user ingested a preset food load. The microscopic images at the two time points were microscopic images of the same target area on the user's tongue. Image processing is performed on the microscopic images at two time points to obtain the first microscopic feature parameters corresponding to the first time point and the second microscopic feature parameters corresponding to the second time point; wherein, the microscopic feature parameters include at least a three-dimensional morphological factor for characterizing the erect, flat, or atrophic state of the lingual papillae. The first microscopic feature parameter and the second microscopic feature parameter are respectively input into a preset efficiency index model to calculate the basal metabolic index corresponding to the first time point, the load metabolic index corresponding to the second time point, and the rate of change of the load metabolic index relative to the basal metabolic index. Based on the rate of change and the first and second microscopic feature parameters, the user's organ functions are determined, and detection results are generated.

2. The control method of the zang-fu function detection apparatus according to claim 1, wherein The specific steps for obtaining the microscopic images at the first and second time points include: Use a macro imaging device to align with a preset area in the middle region of the tongue; A cross-polarized light source was turned on for supplemental lighting, and polarization filtering was used to eliminate specular reflections caused by liquid on the tongue surface to obtain a texture image of the basal tissue of the tongue papillae. The first specific wavelength light source and the second specific wavelength light source were alternately turned on to acquire hemoglobin absorption images reflecting the microvascular filling degree and autofluorescence images reflecting the degree of tongue coating keratinization, respectively.

3. The control method of the zang-fu function detection apparatus according to claim 2, wherein The first specific wavelength light source is a narrowband green light source, and the second specific wavelength light source is an ultraviolet light source; the specific steps of performing image processing on the microscopic images at the two time points to obtain the first microscopic feature parameter corresponding to the first time point and the second microscopic feature parameter corresponding to the second time point include: Based on the texture image, extract the height, density, or volume data of the lingual papillae, and calculate the density parameter of the filiform papillae and the stereomorphic factor. Microvascular filling parameters are extracted based on the hemoglobin absorption image, keratinization degree parameters are extracted based on the autofluorescence image, and aberrant / target region ratio parameters are generated based on the keratinization degree parameters. The microscopic feature parameters of the corresponding state are formed by combining the density of filamentous papillae, three-dimensional morphology factor, microvascular filling degree parameter, keratinization degree parameter, and atypical / target region ratio parameter extracted at the same time point.

4. The control method of the zang-fu function detection apparatus according to claim 3, wherein The specific steps of inputting the first microscopic feature parameter and the second microscopic feature parameter into a preset efficiency index model to calculate the basal metabolic index corresponding to the first time point, the load metabolic index corresponding to the second time point, and the rate of change of the load metabolic index relative to the basal metabolic index include: After normalizing the micro-feature parameters at the first and second time points respectively, the parameters are input into a preset efficiency index model to calculate and output the basal metabolic index at the first time point and the load metabolic index at the second time point. Based on the basal metabolic index, the metabolic load index, and the preset rate of change formula, the rate of change is calculated; wherein, the preset rate of change formula is: ΔE = (DMEI1 - DMEI0) / DMEI0, Where ΔE is the rate of change, DMEI0 is the basal metabolic index, and DMEI1 is the metabolic load index.

5. The control method of the zang-fu function detecting apparatus according to claim 4, wherein The preset efficiency index model is a nonlinear weighted detection model based on support vector regression, used to map multiple tongue surface microstructural parameters to a digestive and metabolic efficiency index (DMEI) in the range of 0–100, expressed as: DMEI = f (w1 D fp , w2 F 3D , w3 I perf , -w4 R atypical ); wherein DMEI is a metabolic index, D fp is a tongue papilla density parameter, F 3D is a stereomorphology factor, I perf is a microvessel filling degree parameter, R atypical is a heterotypic / target area ratio parameter, the w1, w2, w3, w4 are preset weight parameters, and w4>0, and the metabolic index maps the barrier function and protease secretion level of the digestive system mucosa.

6. The control method for the visceral function testing device as described in claim 5, characterized in that, The specific steps for determining the user's organ functions based on the rate of change and the first and second microscopic characteristic parameters include: If the rate of change is >0, and the microvascular filling parameter at the second time point is higher than the microvascular filling parameter at the first time point, then the user's organ condition is determined to be good. If the rate of change is less than or equal to 0, or if the percentage of the irregular / target region at the second time point is higher than the percentage of the irregular / target region at the first time point, then the user's organ condition is determined to be poor.

7. A dietary management method, characterized in that, include: The control method of the organ function testing device as described in any one of claims 1 to 6 is used to test the organ function of a user and obtain the test results; Based on the test results, a dietary management plan corresponding to the current dietary structure is generated.

8. The dietary management method as described in claim 7, characterized in that, The specific steps for generating a corresponding dietary management plan based on the detection results include: If the test results indicate that the internal organs are in good condition, the current diet is determined to be well tolerated, and a diet management plan is generated that includes increasing food concentration and / or increasing the molecular weight of the ingested nutrients. If the test results indicate poor organ function, the current diet will be classified as dietary damage, and a dietary management plan will be generated that includes reducing food concentration and / or reducing the molecular weight of ingested nutrients, or adjusting solid foods to a liquid / semi-liquid diet.

9. A device for detecting the function of internal organs, characterized in that, include: A processor and a memory, wherein the memory is used to store a computer program, and the processor executes the computer program to implement the steps of the control method for the visceral function detection device as described in any one of claims 1 to 6.

10. A food management system, characterized in that, include: An optical acquisition device is used to acquire microscopic images of the same target area on the user's tongue at a first time point and a second time point. The organ function testing device as described in claim 9; The controller is electrically connected to the organ function detection device and is used to implement the steps of the dietary management method as described in claim 7 or 8. The terminal is connected in communication with the controller and is used to output the detection results of the organ functions and / or dietary management plans.

Images