Intelligent physical diagnosis system and method based on infrared thermal imaging double guidance

Abstract

The application discloses an intelligent physical diagnosis system and method based on infrared thermal imaging double guidance, relates to the technical field of diagnosis systems, and constructs a thermal field space topology network graph; maps metabolic heat flow characteristic parameters to the thermal field space topology network graph, combines a heat flow vector direction and a flow velocity to calculate metabolic driving weights of each heat diffusion path, generates a percolation type heat conduction path model with metabolic weights, and outputs a defect-thermal metabolism coupling graph; after analyzing the defect-thermal metabolism coupling graph, it is judged whether a physical imbalance early warning signal is triggered, abnormal source coordinates and related physical type tendencies are synchronously labeled, and intervention suggestion work orders containing positioning coordinates are automatically generated according to the early warning signal. The diagnosis method realizes multidimensional and high-precision collaborative analysis on a human body surface thermal field and metabolic heat flow, accurately positions an abnormal source, and outputs individualized intervention suggestions, thereby providing key information that can directly guide practice for clinical or health management.

Description

Technical Field

[0001] This invention relates to the field of diagnostic system technology, and specifically to an intelligent physical condition diagnostic system and method based on dual guidance of infrared thermal imaging. Background Technology

[0002] With the increasing awareness of health, the importance of accurate assessment and early warning of physical condition in disease prevention, personalized health maintenance, and sub-health intervention is becoming increasingly prominent. While existing infrared thermal imaging-based detection technologies can reflect the distribution of body surface temperature, they typically only focus on static hot spot identification, lacking in-depth analysis of the dynamic diffusion process of the heat field and its correlation with metabolic activities within the body. This fails to reveal the metabolic driving factors behind thermal anomalies, resulting in ambiguous biological significance of diagnostic results and hindering precise location of abnormalities and targeted intervention guidance. Therefore, there is an urgent need for a physical condition diagnostic method that integrates both thermal radiation tomography and metabolic heat flow information, reveals the heat-metabolism coupling mechanism through spatiotemporal synchronization, feature fusion, and topological modeling, and enables abnormal location and intelligent decision-making. This method would meet the demands of modern health management for precise, traceable, and interventionable diagnostic technologies. Summary of the Invention

[0003] The purpose of this invention is to provide an intelligent physical condition diagnosis system and method based on dual guidance of infrared thermal imaging, so as to solve the problems in the prior art.

[0004] To achieve the above objectives, the present invention provides the following technical solution: an intelligent physical constitution diagnosis method based on dual guidance of infrared thermal imaging, the diagnosis method comprising the following steps: S1: Divide the area to be detected on the human body surface into several unit areas, and obtain thermal radiation tomography images and metabolic heat flow displacement data of each unit area; S2: Unify the spatial coordinates of thermal radiation tomography images to the human anatomical coordinate system and synchronize the timestamps of thermal radiation tomography images and metabolic heat flow displacement data. S3: Extract the core point information of hot spots from thermal radiation tomography images, analyze metabolic heat flow displacement data, and form a set of thermal field characteristic parameters and a set of metabolic heat flow characteristic parameters; S4: Construct a spatial topology network diagram of the thermal field based on the spatial adjacency relationship of the hot spot core point and the topological connectivity of the thermal diffusion boundary; S5: Map the metabolic heat flow characteristic parameters to the thermal field space topology network diagram, combine the heat flow vector direction and flow velocity to calculate the metabolic driving weight of each heat diffusion path, generate a seepage heat conduction path model with metabolic weights, and output the defect-thermal metabolism coupling map. S6: After analyzing the defect-thermal metabolism coupling map, determine whether a constitution imbalance warning signal has been triggered, simultaneously mark the coordinates of the abnormal source and the associated constitution type tendency, and automatically generate an intervention suggestion work order containing the location coordinates based on the warning signal.

[0005] Preferably, in step S4, a thermal field spatial topology network diagram is constructed based on the spatial adjacency relationship of the hot spot core point and the topological connectivity of the thermal diffusion boundary, including the following steps: The thermal field characteristics are transformed into a network structure, and the nodes are defined based on the spatial adjacency relationship of the hot spot core points; Each hot spot core point is set as a node in the thermal field space topology network diagram. The connectivity trend between thermal diffusion boundaries is analyzed, and hot spot core points connected by continuous thermal diffusion paths are connected into edges. Calculation and comparison of fluctuation coefficients of thermal characteristic data from adjacent detection sites: Iterate through all adjacent detection positions in sequence, read the thermal feature data F_k and F_(k+1) of each adjacent detection position, measure the distance L_k between adjacent detection positions on the body surface space, calculate the fluctuation coefficient, compare the fluctuation coefficient K with the preset fluctuation threshold K_T, and determine whether the detection position density needs to be adjusted.

[0006] Preferably, the core points of the hot spots connected by a continuous thermal diffusion path are connected by an edge, wherein the initial weight of the edge is the absolute value of the temperature gradient of the regions where the two end nodes are located.

[0007] Preferably, in step S5, the metabolic heat flux characteristic parameters are mapped to the thermal field spatial topology network diagram, and the metabolic driving weights of each heat diffusion path are calculated by combining the heat flux vector direction and flow velocity to generate a percolation-type heat conduction path model with metabolic weights, including the following steps: The set of metabolic heat flow characteristic parameters is mapped to the constructed thermal field spatial topology network diagram according to the spatial location correspondence rule, and the direction of heat flow vector and velocity parameters are made to fall on the actual body surface area represented by the network nodes and edges, relying on human anatomical coordinates. After mapping is completed, metabolic driving weights are calculated for each heat diffusion path, taking into account the direction of the heat flow vector and the flow velocity. Using the nodes of the topological network as heat sources and the edges as flowable paths, the flowability of the edges is calculated by combining the absolute value of the temperature gradient and the metabolic driving weight, thus completing the construction of the seepage-type heat conduction path model.

[0008] Preferably, after mapping, for each heat diffusion path, metabolic driving weights are calculated by combining the heat flow vector direction and flow velocity, including the following steps: The heat flow vector direction corresponding to the search path is checked to confirm whether the heat flow vector direction is consistent with the direction of the known meridians or organ reflex zones. The flow velocity on the path is read, and the flow velocity is compared with the preset benchmark flow velocity. The weight is adjusted according to the difference ratio.

[0009] Preferably, in step S6, after analyzing the defect-thermometabolism coupling map, determining whether to trigger a physical imbalance warning signal includes the following steps: Lock all paths marked with a high-risk level in the map, continuously track the displacement rate of the metabolic heat flow and the expansion speed of the abnormal range of the thermal field, maintain a baseline rate for each high-risk level path. If it is found that the displacement rate of a certain high-risk level path exceeds the baseline rate within a continuous monitoring period and the abnormal range of the thermal field extends outward along this path, then trigger a physical imbalance warning signal.

[0010] Preferably, in step S3, extracting the information of the hot spot core points of the thermal radiation tomography image and analyzing the metabolic heat flow displacement data to form a thermal field characteristic parameter set and a metabolic heat flow characteristic parameter set includes the following steps: Perform hierarchical reconstruction on the thermal radiation tomography image, separate the temperature information at different depths into tomography planes, and execute a hot spot detection algorithm within each tomography plane: Perform vector analysis on the metabolic heat flow displacement data, determine the movement trend of the heat flow through time series difference, map the changes in the continuously collected heat signals to the spatial direction, and obtain the heat flow vector direction; After classifying and sorting out the two types of characteristic data, form a thermal field characteristic parameter set and a metabolic heat flow characteristic parameter set respectively.

[0011] Preferably, after calculating the fluctuation coefficient of the thermal characteristic data of adjacent detection positions, determining whether to adjust the detection position density includes the following steps: Compare the fluctuation coefficient K with the preset fluctuation threshold K_T; If K > K_T, feedback the interval and prompt to increase the detection position density; If K < K_T, feedback the interval and prompt to relax the detection position density.

[0012] Preferably, in step S5, the defect-thermometabolism coupling map includes the abnormal temperature gradient of the thermal field, the metabolic drive weight, and the heat flow displacement vector, and marks the abnormal direction, risk level, and corresponding metabolic heat flow displacement details of each heat conduction path.

[0013] This application also provides an intelligent physical diagnosis system based on dual infrared thermal imaging guidance, including: Data acquisition module: Divide the area to be detected on the human body surface into several unit areas, obtain the thermal radiation tomography images and metabolic heat flow displacement data of each unit area, unify the spatial coordinates of the thermal radiation tomography images to the human anatomical coordinate system, synchronize the timestamps of the thermal radiation tomography images and the metabolic heat flow displacement data, and send the thermal radiation tomography images and the metabolic heat flow displacement data to the map output module; The graph output module extracts core hotspot information from thermal radiation tomography images, analyzes metabolic heat flow displacement data, and forms a set of thermal field feature parameters and a set of metabolic heat flow feature parameters. Based on the spatial adjacency of the hotspot core points and the topological connectivity of the heat diffusion boundary, it constructs a thermal field spatial topology network diagram, maps the metabolic heat flow feature parameters to the thermal field spatial topology network diagram, calculates the metabolic driving weights of each heat diffusion path by combining the heat flow vector direction and flow velocity, generates a percolation-type heat conduction path model with metabolic weights, and outputs a defect-thermal metabolism coupling graph. The defect-thermal metabolism coupling graph is then sent to the early warning and suggestion module. Early warning and suggestion module: After analyzing the defect-thermal metabolism coupling map, it determines whether a constitution imbalance early warning signal is triggered, simultaneously marks the coordinates of the abnormal source and the associated constitution type tendency, and automatically generates an intervention suggestion work order containing the location coordinates based on the early warning signal.

[0014] The technical effects and advantages provided by the present invention in the above technical solution are as follows: This invention extracts core hot spot information and analyzes metabolic heat flow data to form a dual feature parameter set. It also constructs a thermal field spatial topology network diagram by combining hot spot adjacency relationships and thermal diffusion boundary topological connectivity. This allows for the extraction of physiologically significant structured features from complex thermal distributions, clearly presenting the spatial organization patterns of the thermal field and the correlation patterns of potential abnormal regions, thereby improving the interpretability of thermal field features and the accuracy of anomaly localization.

[0015] This invention maps metabolic heat flow characteristic parameters to a topological network diagram and calculates metabolic driving weights to generate a percolation-type heat conduction path model and a defect-thermal metabolism coupling map. It integrates the physical process of heat diffusion with the biological driving mechanism of metabolic activity, revealing the intrinsic coupling relationship between abnormal thermal field and metabolic imbalance. This deepens the diagnosis of physical constitution from observation of phenomena to tracing the mechanism, significantly enhancing the scientific nature and pertinence of diagnosis. Attached Figure Description

[0016] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in this invention. For those skilled in the art, other drawings can be obtained based on these drawings.

[0017] Figure 1 This is an overall flowchart of the diagnostic method of the present invention.

[0018] Figure 2 This is a schematic diagram of the hot spot core point and metabolic heat flow of the present invention.

[0019] Figure 3 This is a schematic diagram of the abnormal thermal metabolism region of the present invention.

[0020] Figure 4This is a schematic diagram of the percolation-type heat conduction path model structure of the present invention. Detailed Implementation

[0021] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0022] Example: This example provides an intelligent physical constitution diagnosis method based on dual guidance of infrared thermal imaging. Please refer to [link / reference]. Figure 1 As shown, the diagnostic method includes the following steps: S1: The area to be detected on the human body surface is divided into several unit areas by spatial gridding. The thermal radiation tomography images of each unit area are obtained by a high-resolution infrared thermal imager to analyze the temperature gradient and thermal spot morphology from the skin surface to the superficial fascia layer. Based on the flexible biosensors attached to key acupoints and visceral reflex zones, metabolic heat flow displacement data of deep tissues (such as the heat diffusion rate caused by local blood perfusion and the amount of tissue heat capacity change) are collected.

[0023] S2: Using the built-in bio-positioning chip, the spatial coordinates of thermal radiation tomography images are unified to the human anatomical coordinate system (with the manubrium of the sternum as the origin and the sagittal, coronal, and transverse planes as axes to establish a three-dimensional reference). At the same time, the timestamps of the two types of data are synchronized to the UTC standard time through the Beidou spatiotemporal positioning chip, eliminating the spatial misalignment problem caused by differences in acquisition timing or body position offset, and outputting thermal radiation tomography images and metabolic heat flow displacement data with consistent coordinates and synchronized timestamps.

[0024] S3: Perform feature extraction on the thermal radiation tomography image, such as... Figure 2 As shown, the three-dimensional coordinates of the core point of the hot spot, the thermal diffusion boundary and the temperature anomaly gradient (such as the temperature of a certain area being 1.5℃ lower than that of normal tissue in the same layer and the gradient change rate reaching 0.8℃ / mm) are extracted. At the same time, the direction of the heat flow vector (such as converging from the extremities to the trunk) and the flow velocity parameter (thermal diffusion distance per unit time) in the metabolic heat flow displacement data are analyzed to form a set of thermal field characteristic parameters and a set of metabolic heat flow characteristic parameters.

[0025] S4: Based on the spatial adjacency relationship of the hot spot core points and the topological connectivity of the heat diffusion boundary, construct a topological network diagram of the heat field space (the nodes are the hot spot core points, the edges are the heat diffusion paths, and the weights are the absolute values of the temperature gradients). Calculate the fluctuation coefficient KK of the thermal characteristic data of adjacent detection positions KK = |F_k - F_(k + 1)| / L_k, where F_k is the thermal characteristic data of the k-th detection position, F_(k + 1) is the thermal characteristic data of the (k + 1)-th detection position, and L_k is the distance between adjacent detection positions. Compare the fluctuation coefficient K with the preset fluctuation threshold K_T: If K > K_T, it indicates that the thermal distribution between adjacent detection positions is significantly different, and it is necessary to feedback this interval to the operator and prompt to increase the density of the detection positions (such as reducing the original distance by 50%); if K < K_T, it indicates that the thermal distribution has high uniformity, and this interval can be feedback and prompt to relax the density of the detection positions (such as expanding the original distance by 30%). Dynamically adjust the layout of the detection positions through real-time data feedback to avoid redundant acquisition or omission of key information.

[0026] S5: Map the metabolic heat flow characteristic parameters to the topological network diagram of the heat field space, and calculate the metabolic driving weight of each heat diffusion path in combination with the heat flow vector direction and velocity (for example, when the heat flow velocity along the lung meridian reaches 0.6 cm / s, the corresponding path weight is increased by 30%). Generate a percolation heat conduction path model with metabolic weights, as Figure 4 shown. Integrate the abnormal temperature gradient of the heat field, the metabolic driving weight, and the heat flow displacement vector, and mark the abnormal direction (such as the heat flow convergence from the hands and feet to the chest and abdomen), the risk level (classified into levels I - III according to the product of the temperature gradient and the velocity), and the corresponding details of the metabolic heat flow displacement of each heat conduction path, and output a defect - heat metabolism coupling map. The heat metabolism abnormal area of the defect - heat metabolism coupling map is as Figure 3 shown, visually presenting the "heat manifestation - metabolic factor" correlation characteristics of physical abnormalities, and this map can be dynamically iterated with the optimization of the detection position density and data update.

[0027] S6: Analyze the metabolic heat flow displacement rate of the heat conduction path and the expansion speed of the abnormal range of the heat field in the defect - heat metabolism coupling map, and judge whether to trigger the physical imbalance warning signal according to the corresponding warning threshold. For example, when the displacement rate of a high - risk heat conduction path exceeds the baseline rate (such as the average of the previous 3 cycles is 0.2 cm / s, and the current cycle reaches 0.5 cm / s) within 2 consecutive monitoring cycles (each cycle is 5 minutes) and the abnormal range spreads along the path (such as spreading from the wrist to the elbow), trigger the physical imbalance warning signal, and synchronously mark the abnormal source coordinates (such as "the area of the right Taiyuan acupoint") and the associated physical type tendency (such as "tendency to lung yang deficiency"). Automatically generate an intervention suggestion work order including the positioning coordinates according to the warning signal (such as "Focus on monitoring the heat flow path of the right lung meridian, and it is recommended to cooperate with warming yang conditioning").

[0028] This embodiment provides an intelligent physical diagnosis system based on dual - guidance of infrared thermography, including: Data acquisition module: Divide the human body surface to be detected area into several unit areas, acquire thermal radiation tomography images and metabolic heat flow displacement data of each unit area, unify the spatial coordinates of thermal radiation tomography images to the human anatomical coordinate system, synchronize the timestamps of thermal radiation tomography images and metabolic heat flow displacement data, and send thermal radiation tomography images and metabolic heat flow displacement data to the atlas output module. The graph output module extracts core hotspot information from thermal radiation tomography images, analyzes metabolic heat flow displacement data, and forms a set of thermal field feature parameters and a set of metabolic heat flow feature parameters. Based on the spatial adjacency of the hotspot core points and the topological connectivity of the heat diffusion boundary, it constructs a thermal field spatial topology network diagram, maps the metabolic heat flow feature parameters to the thermal field spatial topology network diagram, calculates the metabolic driving weights of each heat diffusion path by combining the heat flow vector direction and flow velocity, generates a percolation-type heat conduction path model with metabolic weights, and outputs a defect-thermal metabolism coupling graph. The defect-thermal metabolism coupling graph is then sent to the early warning and suggestion module. Early warning and suggestion module: After analyzing the defect-thermal metabolism coupling map, it determines whether a constitution imbalance early warning signal is triggered, simultaneously marks the coordinates of the abnormal source and the associated constitution type tendency, and automatically generates an intervention suggestion work order containing the location coordinates based on the early warning signal.

[0029] The following is a detailed explanation of each step in this application: The S1 stage is the process of spatially meshing and dividing the area to be detected on the human body surface into several unit regions, specifically including the following steps: Based on 3D scanning and body surface contour modeling, the human body surface is divided into uniform or non-uniform grids according to the preset sampling density and anatomical structure characteristics to ensure that different functional areas (such as chest, abdomen, waist, back, limbs, etc.) can be covered by unit areas of appropriate resolution.

[0030] A high-resolution infrared thermal imager was used to acquire each unit region frame by frame, obtaining thermal radiation tomography images with high spatial detail. During this process, noise suppression and radiometric correction were performed on the infrared images, and the temperature gradient distribution extending from the skin surface to the superficial fascia layer was extracted using a layered inversion algorithm. At the same time, the morphological characteristics of hot spots were identified, including the location of the core hot spot, the extent of boundary diffusion, the regularity of the shape, and the relative spatial relationship between hot spots.

[0031] In stratified inversion, the temperature gradient is calculated as the ratio of the temperature difference between adjacent sampling points along the normal direction to the distance between them. The gradient G is equal to the absolute value of the difference between the temperature T_n at point n and the temperature T_(n+1) at point n+1, divided by the distance d between the two points, i.e., G = |T_n - T_(n+1)| / d. In a sampling in the chest and abdomen region, the temperature at point n on the skin surface was measured to be 36.2℃, and the temperature at point n+1 in the superficial fascia layer was 34.7℃. The distance between the two points along the normal direction was 2.0 mm. The calculated value is G = |36.2 - 34.7| / 2.0 = 1.5 / 2.0 = 0.75℃ / mm. This value reflects the rate of temperature decrease from the surface to the interior of the region.

[0032] The core location of the hotspot was obtained through connected component analysis. High-temperature pixels were first extracted using a threshold 1.2 times higher than the global mean, and then aggregated into a high-temperature block using an eight-neighbor connectivity method. The spatial location of the core point was calculated by averaging the pixel coordinates within the block. In a back-side acquisition, the average pixel coordinates of the high-temperature block were (x=128, y=96, z=12), which is the core point of the hotspot. The boundary diffusion range was determined based on the length and width of the circumscribed rectangle of the high-temperature block. The rectangle was measured to be 18 mm long and 14 mm wide, meaning the diffusion range was 18 × 14 square millimeters. Shape regularity was evaluated by the ratio of the boundary perimeter P to the area A. P was measured to be 52 mm and A to be 196 square millimeters, with a ratio of 52 / 196 ≈ 0.265. A low ratio close to a circle indicates a relatively regular shape. The relative spatial relationship between hotspots was obtained through the Euclidean distance between the core points. The coordinates of the two hotspot core points were (128, 96, 12) and (158, 102, 13), with a distance of √[(158-128)]. 2 +(102-96) 2 +(13-12) 2 =√[900+36+1]=√937≈30.6 mm. This value is used to determine the degree of dispersion or aggregation of hot spot distribution.

[0033] The hotspot morphology analysis, combined with edge detection and region growing, is processed as follows: First, gradient thresholding is used to locate potential high-temperature connected regions. Then, neighboring regions are merged based on the temperature consistency within the region and the temperature difference constraint between adjacent pixels to form a complete set of hotspot objects. The working mechanism of the flexible biosensor is to continuously record the displacement data of metabolic heat flow by sensing the minute changes in the rate of heat diffusion caused by local blood perfusion and the amount of heat accumulation or loss caused by changes in tissue heat capacity with metabolic state. For example, when the blood flow velocity in a certain area increases, heat energy is carried away more quickly, and the rate of local temperature decrease measured by the sensor will be significantly accelerated, and vice versa. This change is encoded into quantifiable heat flow displacement parameters through signal amplification and analog-to-digital conversion by the sensor.

[0034] Phase S2 involves coordinate unification and temporal synchronization. The built-in bio-positioning chip is used to transform and calibrate the spatial position of each pixel or sampling point in the thermal imaging to a three-dimensional human anatomical coordinate system with the manubrium of the sternum as the origin. This coordinate system uses the sagittal, coronal, and transverse planes as orthogonal reference axes, ensuring consistency and comparability in spatial description of image data acquired from different individuals and in different positions. The specific processing is as follows: Coarse registration is performed using surface markers or pre-set anatomical feature points, followed by fine registration based on the spatial constraints of the image grayscale distribution and known anatomical structures. Iterative optimization minimizes the difference between the image coordinates and the standard anatomical coordinates.

[0035] For timestamp synchronization, the high-precision UTC standard time signal provided by the BeiDou spatiotemporal positioning chip is used to uniformly time-stamp both thermal radiation tomography images and metabolic heat flux displacement data. The synchronization processing logic includes detecting the acquisition trigger time of the two types of data, comparing their deviations relative to UTC time, and performing linear or piecewise correction based on the internal clock drift characteristics of the acquisition equipment. This eliminates spatiotemporal misalignment caused by differences in the start-up and shutdown timing of the acquisition equipment, data transmission delays, or slight changes in the subject's position. If the thermal imager captures an image frame at 1024 milliseconds, while the flexible sensor records heat flux in the same physiological event at 1032 milliseconds, the timestamps of both will be uniformly corrected to the same UTC time reference, ensuring that subsequent analysis can be conducted within a completely consistent temporal and spatial framework. The output result is a dual-modal dataset with consistent coordinates and synchronized timestamps.

[0036] In stage S3, feature extraction is performed on the coordinate-unified and time-synchronized thermal radiation tomography images. The processing flow involves layer-by-layer reconstruction of the image, separating temperature information at different depths into tomographic planes that can be analyzed independently, and then executing a hotspot detection algorithm within each layer. The algorithm's processing is as follows: Pixels with temperatures higher than the global average by a certain multiple are selected as candidate high-temperature regions. Continuous high-temperature blocks are then extracted through connected component analysis. The geometric center of each connected component is calculated and its coordinates in three-dimensional space are traced back to obtain the three-dimensional position of the hot spot core point. The boundary with significant temperature changes is traced along the periphery of the hot spot to form thermal diffusion boundary data. The temperature difference between the boundary and the adjacent normal tissue and the temperature change rate along the boundary normal are recorded.

[0037] Vector analysis is performed on metabolic heat flow displacement data from flexible biosensors. The heat flow trend is determined by time series difference. The changes in continuously collected heat signals are mapped to spatial directions to obtain the heat flow vector direction. For example, if the heat signal is detected to be stronger at the extremities and gradually increases in the trunk direction, it can be determined that the heat flow direction converges from the extremities to the trunk. At the same time, the heat signal propagation distance per unit time is measured and statistically analyzed to obtain the flow velocity parameter.

[0038] After being categorized and organized, the two types of feature data are respectively formed into a thermal field feature parameter set and a metabolic heat flow feature parameter set. The former covers hot spot location, boundary morphology, temperature anomaly amplitude and gradient change rate, while the latter includes heat flow direction vector and flow velocity value, providing structured input for subsequent cross-modal analysis.

[0039] In a chromatographic study of the lumbar region, the global average temperature was 33.5℃. A threshold of 43.55℃ was obtained by multiplying this by 1.3. Pixels with temperatures above this value were selected as candidate high-temperature regions. Eight-neighbor connectivity analysis yielded a high-temperature block with an average pixel coordinate of (x=142, y=88, z=9), representing the three-dimensional location of the hotspot's core. The heat diffusion boundary was extracted from areas of significant temperature gradient change along the outer edge of the high-temperature block. The measured temperature at the boundary was 35.0℃, while the temperature of adjacent normal tissue was 33.8℃, a difference of 1.2℃. With a distance of 1.5 mm between adjacent points along the boundary normal, the temperature decreased from 35.0℃ to 34.2℃, resulting in a temperature change rate of |35.0-34.2| / 1.5=0.8 / 1.5≈0.53℃ / mm.

[0040] Vector analysis was performed on metabolic heat flux displacement data from flexible biosensors. The heat flux movement trend was determined by time series difference. The changes in continuously acquired heat signals were mapped to spatial directions to obtain the heat flux vector direction. In the acquisition from the limbs to the trunk, the heat signal amplitude at the distal end of the time series increased from 0.82 mV to 0.95 mV, and at the proximal end from 0.78 mV to 0.93 mV. The increase at the distal end was greater than that at the proximal end, indicating that the heat flux direction converged from the extremities to the trunk.

[0041] The distance of thermal signal propagation per unit time is calculated based on the difference in peak position of thermal signal at adjacent time points. At time t1, the peak is located 5.2 cm from the wrist joint, and at time t2, it is located 5.8 cm away. The time interval is 10 seconds, and the flow velocity is |5.8-5.2| / 10=0.6 / 10=0.06 cm / s.

[0042] After being categorized and organized, the two types of characteristic data form a set of thermal field characteristic parameters and a set of metabolic heat flow characteristic parameters. The former includes the location of hot spots (142,88,9), boundary morphology, temperature anomaly amplitude of 1.2℃ and gradient change rate of 0.53℃ / mm, while the latter includes the heat flow direction vector converging from the extremities to the trunk and the flow velocity of 0.06 cm / s.

[0043] In the S4 stage, the thermal field characteristics are transformed into a network structure that can be used for topology analysis. Nodes are defined based on the spatial adjacency of the hot spot core points, that is, each hot spot core point is set as a node in the graph. The connectivity trend between the heat diffusion boundaries is examined, and core points that can be connected to each other through continuous heat diffusion paths are connected by edges. The initial weight of the edge is taken as the absolute value of the temperature gradient of the regions where the two end nodes are located, which is used to reflect the difference in the intensity of heat transfer.

[0044] A logic for calculating and comparing the fluctuation coefficient K of thermal feature data from adjacent detection sites is introduced. All adjacent detection sites are sequentially traversed, and their respective thermal feature data F_k and F_(k+1) are read. Their distance L_k on the body surface is measured, and the fluctuation coefficient K is obtained by dividing the absolute value of the difference between the two by the distance. The calculation is K = |F_k - F_(k+1)| / L_k. In the left chest region, the thermal feature data of the k-th site is 36.2℃, and that of the (k+1)-th site is 34.7℃. The distance L_k is 10 mm, and K is calculated as K = |36.2 - 34.7| / 10 = 1.5 / 10 = 0.15.

[0045] The system presets the fluctuation threshold K_T to 0.12. Since K is greater than K_T, it indicates that there is a significant difference in thermal distribution between adjacent detection positions, which may mask the details of rapid local changes. At this time, the system will provide feedback to the operator on this interval and suggest increasing the density of detection positions, reducing the original spacing from 10 mm to 5 mm by 50% to improve resolution and capture detailed changes.

[0046] In another set of adjacent detection positions in the right back region, F_k is 35.1℃, F_(k+1) is 35.3℃, and the spacing L_k is 12 mm. The calculated K = |35.1-35.3| / 12 = 0.2 / 12 ≈ 0.017, which is less than K_T, indicating high uniformity of heat distribution. Excessive sampling density would lead to data redundancy and increased acquisition burden. Feedback is provided to this interval, suggesting a wider detection position density. The original spacing is increased by 30% from 12 mm to 15.6 mm, thereby reducing the acquisition load while ensuring no loss of key information. By calculating K in real time and comparing it with K_T, dynamic control of the detection position layout is achieved, enabling the construction of the thermal field spatial topology network to achieve both detail resolution and acquisition efficiency.

[0047] In stage S5, the set of metabolic heat flux characteristic parameters is mapped to the constructed thermal field spatial topology network diagram according to the spatial location correspondence rule. Relying on the previously unified human anatomical coordinates, the direction of heat flux vectors and velocity parameters can be accurately located on the actual body surface areas represented by network nodes and edges.

[0048] After mapping, for each heat diffusion path, metabolic driving weights are calculated by combining the heat flux vector direction and flow velocity: The heat flow vector direction corresponding to the search path is checked to confirm whether it is consistent with the direction of known meridians or organ reflex zones. The flow velocity value on the path is read, and the flow velocity is compared with the preset benchmark flow velocity. The weight is adjusted according to the difference ratio.

[0049] like Figure 4As shown, the permeation-type heat conduction path model treats the thermal field as a medium that can permeate and flow along a specific path. Its flow capacity is influenced by both the abnormal temperature gradient of the thermal field and the metabolic driving weights. The construction logic is as follows: The nodes of the topological network are used as heat sources and the edges are used as flowable paths. The flowability of the edges is calculated by combining the absolute value of the temperature gradient and the metabolic driving weights. The directionality is preserved in the model to reflect the trend of heat flow convergence or divergence.

[0050] The model generation process integrates three types of elements, including thermal field temperature anomaly gradient, metabolic driving weights, and heat flow displacement vectors, and performs attribute labeling on each path.

[0051] The anomaly direction is determined by calculating the principal direction of the heat flow vector along the path; the risk level is classified based on the product of the temperature gradient and the flow velocity, specifically: Several product intervals are defined, with those falling into the highest interval classified as Level III risk, those in the middle interval as Level II risk, and those in the lower interval as Level I risk; the corresponding metabolic heat flow displacement details are directly quoted from the original direction, flow rate, and fluctuation data collected by the flexible sensor.

[0052] After annotation, the attributed paths, along with node information, are output as a unified defect-thermal metabolism coupling map. The map presents the "thermal manifestation-metabolic factor" correlation characteristics of physical abnormalities in a visual form.

[0053] Along the path of the right lung meridian, the baseline flow velocity is 0.46 cm / s, the measured flow velocity is 0.6 cm / s, the adjustment ratio is |0.6-0.46| / 0.46×100%≈30.43%, the original weight is 1.0, then the adjusted weight is 1.0×(1+0.3043)≈1.304.

[0054] The logic of constructing the permeation-type heat conduction path model is as follows: the nodes of the topological network are heat sources or heat sinks, and the edges are flowable paths. The flowability of the edges is calculated by combining the absolute value of the temperature gradient and the metabolic driving weights, and the directionality is preserved in the model to reflect the convergence or divergence trend of heat flow. During the model generation process, the abnormal temperature gradient of the thermal field, the metabolic driving weights, and the heat flow displacement vector are integrated to label the attributes of each path. The anomaly direction is obtained by calculating the main direction of the heat flow vector on the path. The risk level is classified according to the product of the temperature gradient and the flow velocity. Several product intervals are set. Those falling into the highest interval are classified as Level III risk, those in the medium interval are Level II, and those in the lower interval are Level I. The product is calculated by multiplying the absolute value of the temperature gradient by the flow velocity.

[0055] The absolute value of the temperature gradient measured along a certain path was 0.75℃ / mm, and the flow velocity was 0.06 cm / s (0.6 mm / s). The product was 0.75 × 0.6 = 0.45. The system classifies a product greater than or equal to 0.4 as Level III risk, and this path is classified as Level III. The details of the corresponding metabolic heat flow displacement are directly quoted from the original direction, flow velocity, and fluctuation data collected by the flexible sensor. The heat flow direction along this path converges from the extremities towards the torso, with a flow velocity of 0.06 cm / s and a fluctuation coefficient of 0.015.

[0056] In stage S6, after the defect-thermal metabolism coupling map is generated, dynamic monitoring and early warning determination are performed on the heat conduction paths within it. All pathways marked as high-risk (e.g., Level III) in the map are identified, and their metabolic heat flux displacement rates and the expansion rate of abnormal thermal fields are continuously tracked. A baseline rate is maintained for each high-risk pathway, which is the average displacement rate over several consecutive monitoring periods. If the displacement rate of a high-risk pathway exceeds the baseline for two consecutive monitoring periods, and the abnormal thermal field extends outward along the pathway, a body imbalance warning signal is triggered.

[0057] The above judgment, combining the conditions of rate increase and spatial diffusion, can reduce false alarms caused by occasional fluctuations. Simultaneously with triggering the warning, the location of the anomaly source is marked based on the node coordinates in the map, and the correlation of related constitution types is indicated by pattern matching and comparison with the existing constitution classification database.

[0058] The next step is to generate an intervention suggestion work order, and the processing logic is as follows: The system retrieves meridian and organ information associated with the abnormal source and pathway, combines it with risk level and heat flow characteristics, and automatically generates actionable health management recommendations. The location coordinates of the abnormal source are embedded in the work order for easy follow-up or clinical verification. This work order can be directly pushed to the operating terminal or health management system, realizing a closed loop from abnormal detection to intervention guidance, supporting the early identification and personalized treatment of constitution imbalances.

[0059] After the defect-thermal metabolism coupling map is generated, the heat conduction paths within it are dynamically monitored and an early warning system is established. All paths marked as high-risk are identified, and their metabolic heat flow displacement rates and the expansion rate of abnormal thermal field ranges are continuously tracked. A baseline rate is maintained for each high-risk path; this baseline rate is the average displacement rate over several consecutive monitoring periods. It is calculated by summing the displacement rates measured in the previous n periods and then averaging them. The baseline rate is equal to the sum of the displacement rates in the previous n periods divided by n.

[0060] In the monitoring of the high-risk pathway along the right lung, the first three cycles, each 5 minutes long, showed displacement velocities of 0.19, 0.21, and 0.20 cm / s, respectively. The baseline velocity was (0.19 + 0.21 + 0.20) / 3 = 0.60 / 3 = 0.20 cm / s. In the subsequent two consecutive monitoring cycles, the displacement velocities were 0.48 and 0.51 cm / s, both greater than 0.20 cm / s. Furthermore, the abnormal thermal field range expanded along this pathway from the wrist region coordinates (112, 64, 8) to the elbow coordinates (140, 70, 9). The expansion distance, calculated using Euclidean distance, was √[(140 - 112)]. 2 +(70-64) 2 +(9-8) 2 =√[784+36+1]=√821≈28.65 mm, indicating that the abnormal thermal field range extends outward along the path, satisfying both the rate increase and spatial diffusion conditions, triggering a physical imbalance warning signal.

[0061] While triggering the warning, the location of the abnormal source is marked as the coordinates of the right Taiyuan acupoint area (118, 66, 8) based on the node coordinate information in the atlas. The pattern is then compared with the existing constitution classification database to indicate that the associated constitution type tends to be lung yang deficiency.

[0062] Entering the intervention suggestion work order generation stage, the system retrieves meridian and organ information associated with the abnormal source and pathway, combines risk level and heat flow characteristics, and automatically generates executable health management suggestions. The location coordinates of the abnormal source are embedded in the work order. The suggestion content is to focus on monitoring the heat flow pathway of the right lung meridian and to combine it with warming and yang conditioning. The location coordinates of the abnormal source are marked as (118,66,8). This work order can be directly pushed to the operation terminal or health management system to realize a closed loop from abnormal detection to intervention guidance, supporting the early identification and personalized treatment of physical imbalance.

[0063] In the description of this specification, references to terms such as "an embodiment," "example," "specific example," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of the invention. In this specification, illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples.

[0064] The preferred embodiments of the present invention disclosed above are merely illustrative of the invention. These preferred embodiments do not exhaustively describe all details, nor do they limit the invention to any specific implementation. Clearly, many modifications and variations can be made based on the content of this specification. This specification selects and specifically describes these embodiments to better explain the principles and practical applications of the invention, thereby enabling those skilled in the art to better understand and utilize the invention. The invention is limited only by the claims and their full scope and equivalents.

Claims

1. An intelligent physical constitution diagnosis method based on dual guidance of infrared thermal imaging, characterized in that: The diagnostic method includes the following steps: S1: Divide the area to be detected on the human body surface into several unit areas, and obtain thermal radiation tomography images and metabolic heat flow displacement data of each unit area; S2: Unify the spatial coordinates of thermal radiation tomography images to the human anatomical coordinate system and synchronize the timestamps of thermal radiation tomography images and metabolic heat flow displacement data. S3: Extract the core point information of hot spots from thermal radiation tomography images, analyze metabolic heat flow displacement data, and form a set of thermal field characteristic parameters and a set of metabolic heat flow characteristic parameters; S4: Construct a spatial topology network diagram of the thermal field based on the spatial adjacency relationship of the hot spot core point and the topological connectivity of the thermal diffusion boundary; S5: Map the metabolic heat flow characteristic parameters to the thermal field space topology network diagram, combine the heat flow vector direction and flow velocity to calculate the metabolic driving weight of each heat diffusion path, generate a seepage heat conduction path model with metabolic weights, and output the defect-thermal metabolism coupling map. S6: After analyzing the defect-thermal metabolism coupling map, determine whether a constitution imbalance warning signal has been triggered, simultaneously mark the coordinates of the abnormal source and the associated constitution type tendency, and automatically generate an intervention suggestion work order containing the location coordinates based on the warning signal.

2. The intelligent physical constitution diagnosis method based on dual guidance of infrared thermal imaging according to claim 1, characterized in that: In step S4, based on the spatial adjacency relationship of the hot spot core points and the topological connectivity of the thermal diffusion boundary, a thermal field spatial topology network diagram is constructed, including the following steps: The thermal field characteristics are transformed into a network structure, and the nodes are defined based on the spatial adjacency relationship of the hot spot core points; Each hot spot core point is set as a node in the thermal field space topology network diagram. The connectivity trend between thermal diffusion boundaries is analyzed, and hot spot core points connected by continuous thermal diffusion paths are connected into edges. Calculation and comparison of fluctuation coefficients of thermal characteristic data from adjacent detection sites: Iterate through all adjacent detection positions in sequence, read the thermal feature data F_k and F_(k+1) of each adjacent detection position, measure the distance L_k between adjacent detection positions on the body surface space, calculate the fluctuation coefficient, compare the fluctuation coefficient K with the preset fluctuation threshold K_T, and determine whether the detection position density needs to be adjusted.

3. The intelligent physical constitution diagnosis method based on dual guidance of infrared thermal imaging according to claim 2, characterized in that: Edges are formed between the core points of hot spots connected by continuous thermal diffusion paths, where the initial weight of the edge is the absolute value of the temperature gradient of the regions where the two end nodes are located.

4. The intelligent physical constitution diagnosis method based on dual guidance of infrared thermal imaging according to claim 1, characterized in that: In step S5, the metabolic heat flux characteristic parameters are mapped to the thermal field spatial topology network diagram. The metabolic driving weights of each heat diffusion path are calculated by combining the heat flux vector direction and flow velocity, generating a percolation-type heat conduction path model with metabolic weights. This includes the following steps: The set of metabolic heat flow characteristic parameters is mapped to the constructed thermal field spatial topology network diagram according to the spatial location correspondence rule, and the direction of heat flow vector and velocity parameters are made to fall on the actual body surface area represented by the network nodes and edges, relying on human anatomical coordinates. After mapping is completed, metabolic driving weights are calculated for each heat diffusion path, taking into account the direction of the heat flow vector and the flow velocity. Using the nodes of the topological network as heat sources and the edges as flowable paths, the flowability of the edges is calculated by combining the absolute value of the temperature gradient and the metabolic driving weight, thus completing the construction of the seepage-type heat conduction path model.

5. The intelligent physical constitution diagnosis method based on dual guidance of infrared thermal imaging according to claim 4, characterized in that: After mapping, for each heat diffusion path, metabolic driving weights are calculated by combining the heat flux vector direction and velocity, including the following steps: Retrieve the direction of the heat flux vector corresponding to the retrieval path, confirm whether the direction of the heat flux vector is consistent with the known meridians or the direction of the visceral reflex areas, read the flow rate on the path, compare the flow rate with the preset reference flow rate, and adjust the weight according to the difference ratio.

6. The intelligent physical constitution diagnosis method based on dual guidance of infrared thermal imaging according to claim 5, characterized in that: In step S6, after analyzing the defect-heat metabolism coupling map, determine whether to trigger the physical imbalance warning signal, including the following steps: Lock all paths marked with a high-risk level in the map, continuously track the displacement rate of the metabolic heat flux and the expansion rate of the abnormal range of the heat field, maintain a baseline rate for each high-risk level path. If it is found that the displacement rate of a certain high-risk level path exceeds the baseline rate within consecutive monitoring periods and the abnormal range of the heat field extends outward along this path, then trigger the physical imbalance warning signal.

7. The intelligent physical constitution diagnosis method based on dual guidance of infrared thermal imaging according to claim 1, characterized in that: In step S3, extract the information of the core points of the heat spots in the thermoradiometric tomography image, analyze the metabolic heat flux displacement data, and form a set of heat field characteristic parameters and a set of metabolic heat flux characteristic parameters, including the following steps: Perform hierarchical reconstruction on the thermoradiometric tomography image, separate the temperature information at different depths into tomography planes, and execute the heat spot detection algorithm within each tomography plane: Perform vector analysis on the metabolic heat flux displacement data, determine the movement trend of the heat flux through time series difference, map the changes in the continuously collected heat signals to the spatial direction, and obtain the direction of the heat flux vector; After classifying and organizing the two types of characteristic data, form a set of heat field characteristic parameters and a set of metabolic heat flux characteristic parameters respectively.

8. The intelligent physical constitution diagnosis method based on dual guidance of infrared thermal imaging according to claim 2, characterized in that: After calculating the fluctuation coefficient of the heat characteristic data at adjacent detection positions, determine whether to adjust the detection position density, including the following steps: Compare the fluctuation coefficient K with the preset fluctuation threshold K_T; If K > K_T, feedback the interval and prompt to increase the detection position density; If K < K_T, feedback the interval and prompt to relax the detection position density.

9. The intelligent physical constitution diagnosis method based on dual guidance of infrared thermal imaging according to claim 1, characterized in that: In step S5, the defect-heat metabolism coupling map includes the abnormal temperature gradient of the heat field, the metabolic drive weight, and the heat flux displacement vector, and marks the abnormal direction, risk level, and corresponding metabolic heat flux displacement details of each heat conduction path.

10. An intelligent constitution diagnosis system based on dual guidance of infrared thermal imaging, used to implement the diagnostic method according to any one of claims 1-9, characterized in that: Include: Data acquisition module: Divide the area to be detected on the human body surface into several unit areas, acquire the thermoradiometric tomography images and metabolic heat flux displacement data of each unit area, unify the spatial coordinates of the thermoradiometric tomography images to the human anatomical coordinate system, synchronize the timestamps of the thermoradiometric tomography images and the metabolic heat flux displacement data, and send the thermoradiometric tomography images and the metabolic heat flux displacement data to the map output module; Map output module: Extract the information of the core points of the heat spots in the thermoradiometric tomography image, analyze the metabolic heat flux displacement data, form a set of heat field characteristic parameters and a set of metabolic heat flux characteristic parameters, construct a heat field spatial topology network diagram based on the spatial adjacency relationship of the core points of the heat spots and the topological connectivity of the heat diffusion boundary, map the metabolic heat flux characteristic parameters to the heat field spatial topology network diagram, calculate the metabolic drive weight of each heat diffusion path in combination with the heat flux vector direction and the flow rate, generate a percolation-type heat conduction path model with metabolic weights, output the defect-heat metabolism coupling map, and send the defect-heat metabolism coupling map to the warning and suggestion module. Early warning and suggestion module: After analyzing the defect-thermal metabolism coupling map, it determines whether a constitution imbalance early warning signal is triggered, simultaneously marks the coordinates of the abnormal source and the associated constitution type tendency, and automatically generates an intervention suggestion work order containing the location coordinates based on the early warning signal.

Images