Health status detection system based on embedded infrared thermal imaging in a mouse
By embedding an infrared thermal imaging module, a contact sensing array, and an environmental compensation module into the mouse, combined with an AI analysis unit, the problem of continuous, non-invasive, high-precision palm temperature monitoring, which is currently impossible in existing technologies, has been solved. This enables accurate assessment of palm temperature distribution and vascular activity, providing early warnings of health abnormalities.
Patent Information
- Application Number
- CN202511617370.4
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2025-11-06
- Publication Date
- 2026-11-10
- Estimated Expiration
- 2045-11-06
AI Technical Summary
Existing health monitoring equipment cannot achieve continuous, non-intrusive, high-precision monitoring of palm temperature distribution, nor can it acquire rich temperature information of vascular areas, resulting in inaccurate health status assessments.
An infrared thermal imaging module, a contact sensing array, and an environmental compensation module are embedded in the mouse. Combined with an AI analysis unit, it can accurately collect the temperature field of the palm and locate and calibrate anatomical areas, and assess health status through AI algorithms.
It achieves seamless, continuous, and high-precision health monitoring, accurately assessing the temperature distribution and vascular activity of the palm, providing early warnings of health abnormalities, and improving the accuracy and convenience of monitoring.
Smart Images

Figure CN121122740B_ABST
Abstract
Description
Technical Field
[0001] This application relates to the field of health monitoring technology, specifically to a detection system based on embedded infrared thermal imaging in a mouse to sense health status. Background Technology
[0002] With increasing health awareness, the demand for routine health monitoring is growing. Currently, mainstream solutions include: Traditional medical devices, such as forehead thermometers and thermal imagers, offer high accuracy but cannot provide continuous, unobtrusive daily monitoring, limiting their application scenarios.
[0003] Smart wearable devices, such as smart bracelets and watches, can achieve continuous monitoring, but their temperature measurement method based on single or a few points on the wrist is greatly affected by the environment and the tightness of the fit, resulting in limited accuracy. Furthermore, they cannot obtain temperature distribution information in the area of the palm with abundant blood vessels. Summary of the Invention
[0004] In view of this, the embodiments of this application are committed to providing a detection system for health status based on infrared thermal imaging embedded in a mouse. This system can continuously and accurately monitor the temperature distribution of the user's palm during normal use without the user's awareness, and assess the health status through AI algorithms, thereby achieving daily health screening and early warning.
[0005] This application provides a health status detection system based on embedded infrared thermal imaging in a mouse, comprising: a mouse body, and a monitoring system disposed within the mouse body, the monitoring system comprising: The infrared thermal imaging module is used to collect the temperature field of the area covered by the user's palm and generate thermal image data; A contact sensor array is disposed on the palm contact surface of the mouse body to sense the contact state of the user's palm on the mouse surface. An environmental compensation module is used to measure the environmental parameters of the environment in which the mouse is located and generate environmental compensation data; the environmental parameters include ambient temperature and / or ambient light intensity. The AI analysis unit is electrically connected to the infrared thermal imaging module, the contact sensing array, and the environmental compensation module, and is configured as follows: The heat map data is calibrated based on the environmental compensation data; Based on the contact state, determine the spatial position information of the palm on the mouse; Based on the spatial location information, the calibrated thermal image data is registered with a preset hand anatomical feature model to calibrate the temperature value corresponding to a specific anatomical region in the thermal image. The user's health status is assessed using a health analysis model based on the temperature values of one or more of the specific anatomical regions.
[0006] Optionally, the contact sensing array is a pressure sensing array, and the contact state is a contact pressure distribution.
[0007] Optionally, the contact sensing array is a capacitive touch sensing array, and the contact state is a capacitance change distribution.
[0008] Optionally, the AI analysis unit is configured to register the calibrated heatmap data with the preset hand anatomical feature model, specifically including: The temperature distribution in the heat map data is correlated with the thenar eminence, hypothenar eminence, or Laogong acupoint in the palm anatomical feature model.
[0009] Optionally, the AI analysis unit is configured to analyze the temperature values of the specific anatomical region, specifically including: Time-series analysis of temperature data from the specific anatomical region is performed to assess the user's health status; and / or, The microcirculation resistance index is calculated based on the temperature gradient between the specific anatomical feature points to assess the user's vascular activity status; the user's health status is determined based on the vascular activity status.
[0010] Optionally, the AI analysis unit performs the time-series analysis specifically including: When a temperature rise of more than 0.5°C in a preset area is detected within a unit of time, a stress response is identified. When a user's daytime basal body temperature fluctuation exceeds 1.2°C, an immune system abnormality is identified. A personal health baseline model is established based on the user's historical temperature data, and the real-time monitored temperature data is compared with the personal health baseline model to identify changes in health status.
[0011] Optionally, the AI analysis unit is further configured to: Obtain user mouse operation behavior data; The user's health status is assessed by combining temperature values from one or more of the specific anatomical regions with mouse operation data.
[0012] Optionally, the health analysis model includes an early warning unit; The AI analysis unit is configured to control the early warning unit to issue an alarm signal when an abnormal health status is assessed.
[0013] Optionally, the monitoring system further includes: The data transmission module is used to send the user's health status assessment results or the heat map data to an external device via wired or wireless means.
[0014] Optionally, the AI analysis unit is further configured to: The environmental compensation data, the thermal map data, and the contact status are sent to a preset cloud server via the data transmission module. The cloud server predicts the user's health status based on the environmental compensation data, the heat map data, and the contact status, combined with meteorological and epidemiological data.
[0015] Compared with existing technologies, this application has the following beneficial effects: It significantly improves monitoring accuracy and reliability. By introducing a contact sensing array for positioning, it solves the fundamental problem that existing technologies cannot accurately correlate temperature readings with specific anatomical locations on the palm. This ensures that each measurement targets the same anatomical area, achieving precise and repeatable temperature monitoring of a specific region, thereby significantly improving the accuracy and reliability of health status assessment. It enriches data dimensions and enables in-depth analysis. By using an infrared thermal imaging module instead of a single-point infrared sensor, it can acquire a complete palm temperature field distribution map, providing a rich data foundation for deeper health analysis such as vasoactivity analysis and microcirculation status assessment. It can detect early abnormal signs that traditional single-point temperature measurement cannot detect. It achieves non-invasive, high-frequency continuous monitoring. By seamlessly integrating the monitoring system into the mouse that users frequently use, it achieves non-invasive and continuous screening of the user's health status. Compared with wearable devices that require active wearing and inconvenient dedicated medical devices, it has significant advantages in terms of ease of use and continuous data collection. Attached Figure Description
[0016] The above and other objects, features, and advantages of this application will become more apparent from the more detailed description of the embodiments of this application in conjunction with the accompanying drawings. The drawings are provided to further illustrate the embodiments of this application and form part of the specification. They are used together with the embodiments of this application to explain this application and do not constitute a limitation thereof. In the drawings, the same reference numerals generally represent the same components or steps.
[0017] Figure 1 This is a schematic diagram of the structure of a health status detection system based on embedded infrared thermal imaging in a mouse, provided in one embodiment of this application.
[0018] Figure 2 This is a partial structural schematic diagram of a health status detection system based on embedded infrared thermal imaging in a mouse, provided in one embodiment of this application.
[0019] Figure 3 This is a partial flowchart of a health status detection system based on embedded infrared thermal imaging in a mouse, provided in one embodiment of this application. Detailed Implementation
[0020] 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, and 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.
[0021] This application provides a health status detection system based on embedded infrared thermal imaging within a mouse. To better understand the technical solution of this application, the embodiment will be described below with reference to the accompanying drawings. This health status detection system based on embedded infrared thermal imaging within a mouse creatively integrates infrared thermal imaging technology, contact sensing technology, and environmental compensation technology, and utilizes artificial intelligence analysis methods to achieve seamless, continuous, and reliable monitoring of the user's health status.
[0022] Please see Figures 1 to 3 This is a schematic diagram of a health monitoring system based on embedded infrared thermal imaging in a mouse, provided in an embodiment of this application. The system is visually similar to a conventional computer mouse, with its main body being a mouse body 1 for the user to hold. In one embodiment, a complete monitoring system 2 is integrated inside the mouse body 1. This system includes an infrared thermal imaging module 10, a contact sensor array 20, an AI analysis unit 30, and an environmental compensation module 40. It should be noted that all of these modules can be integrated onto a printed circuit board inside the mouse body 1. Through a reasonable structural layout, the added health monitoring function can be achieved without significantly altering the existing shape and grip of the mouse.
[0023] Specifically, the infrared thermal imaging module 10 can be installed inside the mouse body 1, and its lens field of view is designed to cover the area where the user's palm is in contact with or suspended above the mouse shell when holding the mouse normally. This module is used to non-contactly collect the temperature field distribution of this area and generate thermal image data containing temperature information of multiple pixels.
[0024] In some embodiments, the mouse shell can be transparent, so that the various modules located inside the mouse body 1 can also collect light signals from outside the mouse. Furthermore, the placement of the infrared thermal imaging module 10 is one of the key designs of this invention. Its core design principle is to minimize interference with the original structure, feel, and functionality of the mouse while ensuring that the thermal images of the main blood vessel areas of the palm can be collected completely and clearly.
[0025] Therefore, the placement of the infrared thermal imaging module 10 is quite flexible.
[0026] For example, the measurement location could be the bottom front area of the mouse body 1: this location directly faces the thenar and hypothenar eminences of the palm when the user holds the mouse. These areas are rich in subcutaneous blood vessels, making them ideal measurement points for reflecting the body's core temperature and microcirculation status, far superior to the back of the wrist. The palm area has abundant muscle and blood vessel tissue, avoiding the obstruction of thermal signals by the wrist bones, resulting in more accurate and stable temperature distribution data. When the user holds the mouse naturally, this area fits closely to the palm with relatively uniform pressure, helping to maintain a stable measurement distance and reducing measurement errors caused by suspension or excessive pressing.
[0027] Furthermore, an infrared-transmitting window is made on the mouse shell at the location corresponding to the sensor's photosensitive element. The window is preferably made of germanium (Ge) wafer or a polymer infrared-transmitting material (such as PE or TPX), which has high transmittance for infrared radiation with a wavelength of 8-14μm, while also being hard enough to be wear-resistant and can be treated with an oleophobic and hydrophobic coating to prevent stains from affecting measurement accuracy.
[0028] Embedded Installation: The infrared thermal imaging module 10 should be fully embedded inside the mouse, with its photosensitive element facing the infrared-transmitting window. The sensor surface should maintain an appropriate distance from the inner wall of the window and be light-shielded and sealed to completely isolate internal interference such as LED light source and circuit heat generation inside the mouse, ensuring that the collected thermal signal comes entirely from the user's palm.
[0029] It should be noted that although the bottom front area is preferred, depending on the different mouse shapes (such as vertical mice and trackball mice), the infrared thermal imaging module 10 can also be set in other positions that can maintain stable contact with the blood vessel areas of the palm. For example, for vertical mice, it can be set on the inner side wall of the mouse, directly opposite the thenar eminence area at the base of the thumb.
[0030] For high-end gaming mice: consider using multiple infrared sensors positioned on the bottom and sidewalls respectively to achieve more comprehensive data collection and multi-angle verification.
[0031] The contact sensor array 20 is deployed under the palm contact surface of the upper shell of the mouse body 1 to sense the contact state between the user's palm and the mouse surface in real time, such as the contact area, contour and pressure distribution, and generate contact state data that can characterize the spatial position of the palm.
[0032] The environmental compensation module 40 can also be located inside the mouse body 1, with its sensor probe positioned to exchange heat with the external environment. It measures environmental parameters (such as ambient temperature and ambient light intensity) and generates compensation data for calibrating the thermal image data. As the core of the entire monitoring system, the AI analysis unit 30 is electrically connected and communicates with the aforementioned infrared thermal imaging module 10, contact sensing array 20, and environmental compensation module 40, and is responsible for executing a series of complex data processing and analysis tasks.
[0033] The core design task of the environmental compensation module 40 is to accurately and in real-time measure the actual parameters of the mouse's external environment, rather than the parameters of the mouse's internal environment or the local microenvironment affected by the user's body temperature. Its placement must adhere to two main principles: exposure to ambient air and isolation from internal heat sources / interference.
[0034] The preferred location and rationale for the environmental compensation module 40 are as follows: placing the module on the outer surface of the mouse shell (such as the top rear end, the non-grip area of the side skirt) or behind a specially designed ventilation grille can ensure that its sensor is in direct contact with the ambient air, and that the perceived ambient temperature and ambient light intensity are real and undistorted, providing a reliable benchmark for infrared temperature measurement calibration.
[0035] The AI analysis unit (main control chip) and infrared sensors generate heat during operation. Physically isolating them from these internal heat sources prevents the internal heat from heating the surrounding air, which could cause the measured ambient temperature to be higher than the true value. If the measured T_env is too high, the system will incorrectly interpret the ambient heat radiation as strong, thus overcompensating (underestimating the body surface temperature) and introducing systematic errors.
[0036] The mouse should be positioned away from the main area where the hand grips the mouse (such as the sides of the mouse or the left side skirt) to prevent heat from the user's hand from being conducted to the module through the mouse shell after holding it for a long time, which could affect its stability.
[0037] Typical layout scheme: Scheme A (independent external): The ambient temperature sensor and ambient light sensor are integrated into a separate sub-board and installed in the non-contact area at the top or bottom edge of the mouse tail. This location has good airflow and is away from the internal motherboard and the heat source of the palm.
[0038] Option B (integrated into the motherboard but physically isolated): The sensor chip of the environmental compensation module 40 is placed on the edge area of the main circuit board of the monitoring system, ensuring that its light-sensing / temperature-sensing components are directly facing the specially designed micro-vents on the casing. Simultaneously, a thermal isolation groove can be designed between the chip and the motherboard heat source to prevent heat conduction from the motherboard through the copper foil.
[0039] Ambient light sensor: Its photosensitive window must be transparent to visible light, so the window material is usually glass or transparent plastic. The window needs to be frosted to homogenize the incident light, avoid directional interference, and achieve accurate measurement of ambient light intensity.
[0040] Ambient temperature sensor: A fast-response, high-accuracy digital temperature sensor (e.g., ±0.5℃ accuracy) should be selected. Its installation should ensure good thermal contact with the mouse casing (usually plastic) to quickly respond to changes in ambient temperature, while simultaneously isolating internal heat sources from heat conduction through structural design.
[0041] Furthermore, the AI analysis unit is electrically connected to the infrared thermal imaging module, the contact sensing array, and the environmental compensation module, and is configured as follows: The heat map data is calibrated based on the environmental compensation data; Based on the contact state, determine the spatial position information of the palm on the mouse; Based on the spatial location information, the calibrated thermal image data is registered with a preset hand anatomical feature model to calibrate the temperature value corresponding to a specific anatomical region in the thermal image. The user's health status is assessed using a health analysis model based on the temperature values of one or more of the specific anatomical regions.
[0042] The AI analysis unit 30, acting as the "intelligent brain" of the entire monitoring system, establishes an electrical connection with the infrared thermal imaging module 10, the contact sensing array 20, and the environmental compensation module 40 through its internally embedded algorithm program, and performs collaborative control and data fusion processing. Its data processing flow constitutes the core innovative method of this invention: The thermal image data is calibrated based on the environmental compensation data; the spatial position information of the hand on the mouse is determined based on the contact state; the calibrated thermal image data is registered with a preset hand anatomical feature model based on the spatial position information to calibrate the temperature value corresponding to a specific anatomical region in the thermal image; and the user's health status is assessed through a health analysis model based on the temperature values of one or more of the specific anatomical regions.
[0043] In some embodiments, the contact sensing array is a pressure sensing array, and the contact state is a contact pressure distribution. Alternatively, the contact sensing array may be a capacitive touch sensing array, and the contact state may be a capacitance change distribution.
[0044] The contact sensing array 20 is a key sensor for achieving precise spatial positioning. Its core function is to sense the contact state between the user's palm and the mouse surface, providing the AI analysis unit 30 with raw data for calculating spatial position information. This invention provides two preferred, interchangeable technical solutions, both of which can efficiently and reliably accomplish this task: Option 1: The contact sensing array 20 can be specifically implemented as a pressure sensing array.
[0045] The array consists of multiple miniature pressure sensor elements (such as piezoresistive, piezoelectric, or capacitive force sensors) arranged in a matrix. When a user's palm contacts different parts of the mouse with varying force, each sensor element senses the corresponding pressure and outputs an electrical signal proportional to the pressure. The signals from all elements together form a contact pressure distribution map. The contact state is this pressure distribution map. Peak pressure areas in the map typically correspond to bony support points such as metacarpal prominences, while areas with more uniform pressure correspond to muscular areas such as the thenar and hypothenar eminences. In one specific embodiment, a flexible thin-film pressure sensor (such as an FSR) with a thickness of less than 0.2 mm can be used, perfectly fitting under the complex curved surface of the mouse shell without affecting the mouse's appearance and feel. Its sensitivity is preferably 0.1 N, sufficient to detect minute pressure changes when the palm is lightly placed.
[0046] The pressure distribution is strongly correlated with the anatomical structure of the palm, the data is stable, it is less affected by environmental humidity and oil, and the cost is relatively low.
[0047] Option 2: The contact sensing array 20 can be alternatively implemented as a capacitive touch sensing array.
[0048] This array consists of a series of capacitive sensing electrodes (such as ITO patterns) etched onto a PCB or flexible circuit. When a human hand (a conductor) approaches or touches these electrodes, it causes a change in capacitance between the electrodes and ground. An integrated circuit (such as a dedicated touch chip) continuously monitors the capacitance value of each electrode, generating a capacitance change distribution map. The contact state is this capacitance change distribution map. The intensity of the capacitance change reflects the proximity or contact area between the hand and the mouse surface, and its distribution pattern can clearly outline the contour and center of gravity of the hand. The electrode pattern can be designed as a diamond grid to maximize coverage area and detection sensitivity. Its advantage lies in its ability to detect "near-field" or "light touch" states, effectively sensing the presence and location of the hand even when the hand is not fully pressing down.
[0049] Regardless of which approach is adopted, the AI analysis unit 30 is configured to receive the corresponding distribution map data (pressure or capacitance). By analyzing the shape of the distribution map, calculating its centroid, and identifying local extrema, the AI analysis unit 30 can accurately determine the spatial position information of the hand on the mouse, providing an indispensable spatial transformation basis for subsequent registration of heat maps and anatomical models.
[0050] In some embodiments, the AI analysis unit is configured to register the calibrated heat map data with the preset palm anatomical feature model, specifically including: matching the temperature distribution in the heat map data with the thenar eminence, hypothenar eminence, or Laogong acupoint in the palm anatomical feature model.
[0051] In a monitoring system based on infrared thermal imaging embedded in a mouse to detect health conditions, the AI analysis unit registers calibrated thermal image data with a preset hand anatomical feature model to achieve a precise correlation between "temperature data" and "anatomical region," providing core support for subsequent targeted health analysis. This registration process, based on the spatial position of the hand determined by the contact sensing array, is implemented in four steps: "model foundation construction - anchor point positioning - dynamic adaptation - region calibration," as detailed below: The AI analysis unit's built-in "Preset Hand Anatomical Feature Model" is a standardized digital model built based on hand anatomy data (including CT angiography data and body surface anatomical measurement data) of 1,000 healthy adults aged 18-60. Its core parameters include three key anatomical regions: the thenar eminence, hypothenar eminence, and Laogong acupoint, ensuring the model's universality and accuracy. Specific parameters are as follows: The thenar eminence: Oval-shaped, located on the radial side (thumb side) of the palm, composed of the abductor pollicis brevis and flexor pollicis brevis muscles, etc., with a long diameter of 35-45mm, a short diameter of 20-30mm, and an area of 8-12cm². 2 This corresponds to the area of the superficial palmar branch of the radial artery, where blood vessels are densely packed, and temperature changes can reflect the blood supply status of the upper limb arteries.
[0052] The hypothenar eminence is elongated and located on the ulnar side (little finger side) of the palm. It is formed by the abductor digiti minimi and flexor digiti minimi brevis muscles, with a long diameter of 40-50 mm, a short diameter of 15-25 mm, and an area of 6-10 cm². 2 This corresponds to the area of the superficial palmar branch of the ulnar artery, and together with the thenar eminence, it forms the main blood supply network of the palm. Temperature differences can reflect vascular symmetry. Laogong acupoint: It is round and located between the 2nd and 3rd metacarpal bones in the palm, near the radial edge of the 3rd metacarpal bone. It has a diameter of 10-15mm and its center is about 30-40mm away from the base of the palm (wrist crease). It corresponds to the dense microcirculation area of the palm. The skin there is relatively thin and is sensitive to sympathetic nerve excitation and inflammatory response to temperature changes.
[0053] Meanwhile, the model also stores the relative positional relationship data of three types of regions, such as the straight-line distance between the center of the thenar eminence and the center of the Laogong acupoint being 25-30mm, and the straight-line distance between the center of the hypothenar eminence and the center of the Laogong acupoint being 30-35mm, providing a positional reference for subsequent registration.
[0054] It should be noted that, in practical applications, the core of a specific standardized digital model can be adjusted based on the user's actual needs.
[0055] The core prerequisite for registration is to determine the "common anchor point" between the heatmap data and the model—that is, the actual anatomical feature points of the palm on the mouse. This anchor point is determined by the contact state data collected by the contact sensing array (pressure sensing array or capacitive touch sensing array). The specific positioning logic is divided into two categories depending on the type of contact sensing array: The pressure distribution matrix identifies "pressure peak points" as anchor points. Since the thenar eminence and Laogong acupoint areas are in the closest contact with the mouse surface during normal grip, resulting in the highest pressure values (thenar eminence pressure peak 1.2-1.8N, Laogong acupoint pressure peak 0.8-1.2N), the AI analysis unit extracts the two points with the highest values from the pressure distribution matrix, labeling them as "thenar eminence anchor point" and "Laogong acupoint anchor point," respectively, and records their coordinates in the mouse coordinate system (e.g., thenar eminence anchor point coordinates (X15, Y10), Laogong acupoint anchor point coordinates (X25, Y20)); alternatively, the capacitance change distribution matrix identifies "capacitance change peak points" as anchor points. Because the skin in the Laogong acupoint area is thin and has a high water content, the capacitance value formed with the touch electrode is the largest (capacitance change 12-16pF), followed by the thenar eminence area (10-14pF). The AI analysis unit extracts the two points with the largest capacitance changes and marks them as "Laogong acupoint anchor point" and "thenar eminence anchor point" respectively, and records their coordinates simultaneously.
[0056] In addition, to ensure the accuracy of the anchor points, the AI analysis unit will verify whether the distance between the anchor points conforms to the relative positional relationship of the anatomical regions in the model (e.g., the anchor point spacing should be within the range of 25-30mm). If the deviation exceeds 5mm, it will be judged as "abnormal grip posture", and the contact state data will be re-collected until the anchor points meet the standard.
[0057] Size scaling: Measure the actual distance between two anchor points (such as the thenar eminence anchor point and the Laogong acupoint anchor point) in the mouse coordinate system (e.g., actual distance 28mm), compare it with the standard distance of the corresponding anchor points in the model (e.g., standard distance 27mm), calculate the scaling ratio (28 / 27≈1.037), and enlarge or reduce the overall size of the model according to this ratio to ensure that the size of the anatomical area in the model is consistent with the user's actual palm. Angle rotation: Adjust the rotation angle of the model according to the direction of the line connecting the anchor points in the mouse coordinate system (e.g., the line is tilted at a 30° angle) so that the direction of the line connecting the anchor points in the model is consistent with the direction of the line connecting the actual anchor points, thus avoiding registration deviation caused by grip tilt. Position translation: The scaled and rotated model is translated as a whole so that the anchor points in the model completely coincide with the actual anchor points determined by the contact state data in the heat map, thus completing the initial position alignment between the model and the heat map.
[0058] After aligning the model with the heatmap, the AI analysis unit uses the U-Net image segmentation algorithm to precisely delineate the pixel regions in the calibrated heatmap that correspond to the thenar eminence, hypothenar eminence, and Laogong acupoint in the model, achieving the final correspondence between "temperature distribution" and "anatomical region." The specific process is as follows: Region contour extraction: The U-Net algorithm learns the morphological features of anatomical regions in the model (such as the elliptical contour of the thenar eminence and the circular contour of the Laogong acupoint), and identifies sets of pixels with the same morphological features in the heat map to form region contours. Regional temperature data extraction: The temperature values of all pixels within the defined region are statistically analyzed to calculate the "regional average temperature" (e.g., the average temperature of the thenar eminence region is 35.8℃), the "regional temperature standard deviation" (e.g., 0.3℃), and the "regional maximum temperature" (e.g., 36.1℃). These data are then associated with and stored with the corresponding anatomical region names (e.g., "thenar eminence - average temperature 35.8℃"). Registration result verification: Verify whether the temperature distribution in the calibration area conforms to physiological laws (e.g., the temperature of the thenar eminence should be 0.5-1.5℃ higher than that of the hypothenar eminence). If not, readjust the model adaptation parameters until the temperature distribution pattern is consistent with physiological characteristics to ensure the registration result is valid.
[0059] Through the above registration, the AI analysis unit can transform thermal image temperature data that originally had no clear anatomical significance into targeted data with physiological correlations, such as "thenar eminence temperature" and "Laogong acupoint temperature". This provides an accurate data foundation for subsequent analysis of health status (such as vasoactivity and stress response) through temperature in specific areas, avoiding misjudgments of health caused by "mismatch between temperature data and anatomical regions".
[0060] In some embodiments, the AI analysis unit is configured to analyze the temperature values of the specific anatomical region, specifically including: performing time-series analysis on the temperature data of the specific anatomical region to assess the user's health status; and / or calculating the microcirculation resistance index based on the temperature gradient between the specific anatomical feature points to assess the user's vascular activity status; and determining the user's health status based on the vascular activity status.
[0061] Specifically, the time-series analysis performed by the AI analysis unit includes: When a temperature rise of more than 0.5°C in a preset area is detected within a unit of time, a stress response is identified; when a user's daytime basal body temperature fluctuation exceeds 1.2°C, an immune system abnormality is identified; a personal health baseline model is established based on the user's historical temperature data, and the real-time monitored temperature data is compared with the personal health baseline model to identify changes in health status.
[0062] In some embodiments, after completing anatomical registration and obtaining the real-time temperatures of the thenar eminence (TA1), hypothenar eminence (TA2), and Laogong acupoint (TA3), the AI analysis unit initiates a dual-channel analysis process: a temporal anomaly detection channel and a microcirculation resistance index (MRI) channel. The two channels can operate independently or be weighted and fused to ultimately output a health status label.
[0063] Temporal analysis includes: The AI analysis unit 30 is configured to perform time-series analysis on temperature data of a specific anatomical region, that is, to continuously track its changes over a period of time, thereby assessing the user's health status trend. This analysis specifically includes the following aspects: Short-term abnormality monitoring (acute change monitoring): Algorithm: The system calculates the rate of temperature change (i.e., slope, dT / dt) in real time for the target area (preferably the Laogong acupoint area). This calculation is based on a continuously acquired sequence of thermal image frames.
[0064] Judgment and output: When the temperature of the preset area is detected to rise by more than 0.5°C within a unit of time (1 minute), the AI analysis unit 30 determines that the user has a short-term stress response.
[0065] Physiological significance: This rapid temperature spike is usually related to physiological responses such as sympathetic nerve excitation and adrenaline secretion, and may be triggered by factors such as acute infection, emotional stress, or severe external stimuli. This function enables early, imperceptible warning of acute physiological changes.
[0066] Long-term trend analysis (monitoring of chronic changes): Algorithm: The system filters and analyzes users' basal body temperature data on different dates and during the same time period (e.g., the working hours from 10:00 to 11:00 AM every day), calculates the daily average or the stable value within that period, and observes the fluctuations over multiple days.
[0067] Judgment and output: When the fluctuation range of the user's daytime basal body temperature (the difference between the maximum and minimum values) is detected to exceed 1.2℃, the AI analysis unit 30 determines that the user has signs of long-term immune system abnormality.
[0068] Physiological significance: A normal person's basal body temperature is relatively stable after excluding factors such as the menstrual cycle. Sustained and excessive diurnal fluctuations often indicate that the immune system is being continuously activated or in a disordered state, which may be related to chronic infection, autoimmune diseases, or low immune function.
[0069] Personalized baseline comparison (personalized health model): Algorithm: The health analysis model possesses self-learning capabilities. During the initial 7-day calibration period of the system, it continuously collects various temperature data of the user in a healthy state and establishes a personal health baseline model for the user through machine learning algorithms (such as Gaussian distribution modeling). This baseline includes not only the average temperature value but also its normal fluctuation range (such as ±0.3℃).
[0070] Judgment and output: In subsequent daily monitoring, the system will compare the real-time temperature data with this personal health baseline model.
[0071] Physiological significance: This eliminates individual differences in basal body temperature (some people are naturally warmer, some are naturally cooler), making monitoring completely personalized. The system only issues an alert when real-time data deviates significantly from its baseline (e.g., more than 2 standard deviations), thereby greatly improving the accuracy and specificity of the assessment and effectively reducing false alarms.
[0072] Spatial gradient analysis and vasoactivity assessment include: The AI analysis unit 30 is also configured to perform spatial gradient analysis to assess the local blood circulation status, specifically including: Algorithm: The system first acquires the temperature values of two or more specific anatomical feature points after registration (e.g., the temperature T_pc8 of the Laogong acupoint region and the temperature T_thenar of the thenar eminence region). Then, it calculates the temperature gradient between them (ΔT = T_pc8 - T_thenar). This ΔT value is highly correlated with the blood perfusion rate of subcutaneous tissue. The AI analysis unit 30 further substitutes ΔT into a pre-set physiological model (such as a mapping relationship established based on hemodynamics) to calculate a microcirculation resistance index (MRI).
[0073] Judgment and Output: This index is a core indicator for assessing vasculature. If the index consistently deviates from the normal range (for example, an elevated index indicates increased microcirculatory resistance and insufficient blood perfusion), it suggests that the user may have peripheral circulatory disorders, local inflammation, or autonomic nervous system dysfunction.
[0074] Physiological significance: Ultimately, the AI analysis unit 30, based on this vasoactivity state and combined with other indicators, comprehensively determines the user's health status. For example, even if body temperature is normal, a persistently abnormal microcirculation resistance index may indicate cardiovascular health risks or early diabetic peripheral vascular disease, achieving a deeper level of health insight.
[0075] In some embodiments, the AI analysis unit is further configured to: acquire user mouse operation behavior data; and analyze the data by combining temperature values of one or more specific anatomical regions with the mouse operation behavior data to assess the user's health status.
[0076] To further improve the accuracy of health status assessment, reduce false alarms, and uncover deeper health information, the AI analysis unit 30 is configured to perform multimodal data fusion analysis.
[0077] This process goes beyond a single body temperature parameter. By combining user behavioral data for cross-validation and comprehensive analysis, the specific implementation method is as follows: Acquisition of mouse operation behavior data The AI analysis unit 30 communicates with the mouse's main control chip and is configured to acquire real-time user mouse operation behavior data. This data consists of natural interaction signals generated by the user while operating the computer, and mainly includes, but is not limited to: Click frequency (CPM): The number of left and right clicks per unit of time.
[0078] Velocity: The speed at which the cursor moves across the screen; average speed and peak speed can be calculated.
[0079] Movement trajectory characteristics include the smoothness of movement (whether there is jitter), movement acceleration, and movement accuracy (such as the degree of deviation from the target).
[0080] Roller rolling speed and frequency.
[0081] Macro-level behavioral patterns: such as the percentage of idle time when the keyboard and mouse are not used per unit of time.
[0082] Fusion and analysis of multimodal data The health analysis model of the AI analysis unit 30 is designed to combine temperature values (physiological data) from one or more specific anatomical regions with the aforementioned mouse operation behavior data (behavioral data) for fusion analysis. Its core logic is that abnormalities in physiological state often subconsciously manifest as changes in behavioral patterns. This fusion analysis is mainly reflected on two levels: Cross-validation (increases confidence and reduces false positives): Scenario: The system detected an increase in temperature (ΔT>0.8℃) in the user's Laogong acupoint area, suspected of being a fever.
[0083] Judging by a single data point: relying solely on temperature data may not be enough to distinguish between physiological fever and the effects of a temporary environmental heat source.
[0084] Fusion Analysis: The AI analysis unit 30 simultaneously examines mouse operation behavior data during this time period. If the data simultaneously shows a significant decrease in the user's click frequency, a slower movement speed, and a longer idle time, this behavioral pattern is highly consistent with the state of "physical discomfort and mental lethargy."
[0085] Output: If behavioral data provides strong evidence of elevated body temperature, the system will increase the confidence level of "the user is in a state of illness" and generate a higher-level warning. Conversely, if the user's behavior remains active and precise when body temperature is elevated, the system may classify it as environmental interference or random fluctuations to avoid false alarms.
[0086] Uncovering Potential Health Information (Assessment Dimensions Beyond Body Temperature): Scenario: The user's temperature values across all anatomical regions are within the normal range. Single Data Judgment: Based solely on temperature data, the system would consider the user "healthy." Fusion Analysis: However, when analyzing behavioral data, AI Analysis Unit 30 discovered subtle but persistent tremors in the user's recent mouse movement trajectory (manifested as decreased smoothness of cursor movement and increased minor oscillations), along with decreased click accuracy.
[0087] Output: This change in behavioral pattern may indicate accumulated fatigue, insufficient sleep, abnormal neurological state (such as early Parkinson's symptoms), or excessive stress. Therefore, the system can generate health prompts related to "physical performance" or "neuromuscular state," such as "decreased operational precision detected, rest recommended," enabling the monitoring of non-fever-related health issues.
[0088] By introducing multimodal data fusion analysis, this invention achieves the following gains: High reliability: Through the mutual corroboration of physiological and behavioral signals, the false alarm rate caused by environmental interference or measurement noise is greatly reduced, making early warnings more reliable. Changes in behavioral patterns sometimes appear earlier than changes in body temperature. Fusion analysis can capture earlier, subclinical trends in declining health status. It expands health assessment from "whether body temperature is abnormal" to richer dimensions such as "physical fitness level" and "whether neural control is stable," providing a more comprehensive health profile.
[0089] In some embodiments, the health analysis model includes an early warning unit; the AI analysis unit is configured to control the early warning unit to issue an alarm signal when an abnormal health status is assessed.
[0090] The present invention provides a health status detection system based on embedded infrared thermal imaging in a mouse, integrating a complete closed-loop early warning system. The health analysis model includes an early warning unit, which is a software module integrating early warning logic, triggering conditions, and output control. It is deeply integrated with the AI analysis unit 30, ensuring that the evaluation results can be effectively translated into prompts for the user.
[0091] The AI analysis unit 30 is configured to continuously monitor the health status assessment results. When an abnormal health status is detected, an early warning process is triggered. The "abnormal" status includes, but is not limited to: Body temperature data that exceeds the range of an individual’s health baseline model (e.g., three consecutive measurements exceeding the baseline ± 2 standard deviations).
[0092] A sudden increase in temperature of more than 0.5°C per unit time in a specific anatomical region (short-term stress response).
[0093] A daytime basal body temperature fluctuation exceeding 1.2°C is a sign of long-term immune system abnormality.
[0094] The microcirculation resistance index continues to deviate from the normal threshold.
[0095] Multimodal fusion analysis (combining operational behavior data) determines that the health confidence level is below a preset threshold.
[0096] 2. Implementation method of early warning unit The "early warning unit" is not a single piece of hardware, but an integrated warning solution that sends out warning signals by controlling existing or newly added hardware units on the mouse. The AI analysis unit 30 is configured to control one or more of the following methods to provide the user with intuitive, multi-sensory warnings: Visual cue: Control the RGB light strip on the mouse (usually found in gaming mice) to emit specific colors and patterns of light.
[0097] For example: solid green indicates normal; slow yellow breathing indicates a minor abnormality or warrants attention; rapid red flashing indicates a serious abnormality warning.
[0098] Haptic cues: Control the tiny vibration motor inside the mouse (similar to a mobile phone vibrator) to produce vibrations of different intensities and patterns.
[0099] For example, a short vibration indicates a prompt; continuous rapid vibration indicates an emergency warning. This method offers good privacy and can be detected even in noisy environments.
[0100] The software prompts: The warning information will be sent to the connected computer or mobile device via the data transmission module 50.
[0101] For example, displaying a non-intrusive Toast notification on the computer screen, or showing detailed warning information and health advice in the accompanying health management software (such as "Elevated body temperature detected, it is recommended to replenish fluids").
[0102] Auditory cues: Preset cues are played through the computer operating system (software required).
[0103] 3. Early Warning Levels and Strategies The early warning unit supports configurable hierarchical early warning strategies to match different levels of anomaly severity: Level 1 alert (minor anomaly): This may only trigger the light strip to turn yellow, or record a status log in the software, without providing strong interference or reminders.
[0104] Level 2 Warning (Moderate Abnormality): May trigger a yellow breathing light and a short vibration to attract the user's attention.
[0105] Level 3 Alert (Severe Anomaly): Triggers a flashing red light, continuous rapid vibration, and a pop-up window on the screen. Users are strongly advised to pay attention immediately.
[0106] This tiered strategy ensures the effective delivery of important alerts while avoiding disruption to users' normal work and life due to frequent interruptions, reflecting the human-centered design.
[0107] 4. Technical Effects This invention, by integrating an early warning unit, enables immediate hardware-based feedback of assessment results to the user without requiring the user to actively check the device or app. The early warning system seamlessly integrates with mouse operation scenarios, allowing users to receive health feedback without changing any habits. Multiple methods, including light, vibration, and sound, cover different usage environments to ensure effective reception of early warning information. This forms a closed-loop monitoring system, ultimately completing the entire health monitoring cycle from "data acquisition" to "intelligent analysis" to "result output," transforming this invention from a simple measuring device into an effective health management tool.
[0108] In some embodiments, the monitoring system further includes: The data transmission module is used to send the user's health status assessment results or the heat map data to an external device via wired or wireless means.
[0109] The core function of the data transmission module 50 is to establish a communication link between the mouse-embedded infrared thermal imaging health monitoring system and the external world, for transmitting data to external devices via wired or wireless means. Wireless method (preferred): This module can integrate wireless communication chips such as Bluetooth Low Energy (BLE 5.2), Wi-Fi, or Zigbee. The wireless method provides maximum flexibility and avoids the constraints of cables.
[0110] The data transmission module 50 is configured to transmit the following two types of data, and can select to send one or a combination of them according to the application scenario: Health Status Assessment Results: This is lightweight, high-value information processed by the AI analysis unit 30. It features small data volume, high transmission efficiency, and is suitable for real-time display. Content includes: a final health status score or level (e.g., "Good," "Attention," "Warning"); specific abnormality type indicators (e.g., "Elevated Body Temperature," "Microcirculation Abnormality"); key quantitative indicators (e.g., current body temperature, heart rate variability index); and brief health recommendations generated by the system (e.g., "Rest is recommended").
[0111] Thermal image data: This is raw or calibrated infrared image data. It is large in volume and primarily transmitted for more in-depth purposes: Archiving and retrospection: Sending complete temperature distribution data to a computer or cloud for storage, facilitating users to review historical health trends or providing detailed, continuous data support for doctors' diagnoses. Cloud AI analysis: Providing data raw materials for running more complex and computationally intensive AI models in the cloud.
[0112] 3. Typical application scenarios and external devices Data is transmitted to various external devices to enable a wide range of application functions: Personal Computer (PC) / Laptop: This is the most direct external device. The mouse connects to the PC's accompanying health management software via Bluetooth or USB. After receiving data, the software can: Visual Dashboard: Displaying body temperature trend lines, health score history, etc. in chart form. Detailed Report Viewing: Providing more detailed daily / weekly health reports than mouse hardware prompts. Settings and Configuration: Allowing users to configure warning thresholds, view personal information, etc. Mobile Smart Devices (Phone / Tablet): Connecting to a mobile app via Bluetooth, allowing users to view health information anytime, anywhere, freeing them from the limitations of a computer. Cloud Server 3: Uploading data to the cloud via Wi-Fi or through a PC relay. This is the foundation for realizing macro-level health analysis and remote care functions.
[0113] By adding a data transmission module 50, this invention achieves the following functional expansions: Data persistence: Overcoming the limitation of limited local storage capacity of the mouse, enabling long-term, continuous health data recording. Functional expansion: Upgrading the mouse from an independent sensing device into an intelligent node that interacts with more powerful computing platforms (PC, mobile phone, cloud), greatly enriching its functional boundaries. Cloud-based services: Laying the technical foundation for providing value-added services such as big data-based epidemiological analysis and remote family doctor consultations. Interactive graphical interface: Providing richer and more detailed information display and interaction methods on external devices than hardware lighting effects, greatly improving the user experience. Wired connection: In scenarios involving the transmission of large amounts of data or ensuring connection stability, this module can also transmit data via the mouse's USB cable (when used as a wired mouse), utilizing the USB-HID protocol or virtual serial port protocol.
[0114] In some embodiments, the AI analysis unit is further configured to: send the environmental compensation data, the heat map data, and the contact status to a preset cloud server via the data transmission module; and predict the user's health status based on the environmental compensation data, the heat map data, and the contact status, combined with meteorological and epidemiological data.
[0115] The functionality of the health status detection system based on embedded infrared thermal imaging in a mouse, as described in this invention, is further extended, achieving a leap from "monitoring" to "prediction" through cloud collaboration. Specifically, the AI analysis unit 30 is configured to perform the following extended functions: 1. Cloud upload of multimodal data The AI analysis unit 30, through the data transmission module 50 (which must support Wi-Fi or cellular network functionality), packages and encrypts the locally processed multi-dimensional raw and derived data before uploading it to a preset cloud server. This data forms the cornerstone of cloud-based deep analysis and mainly includes: Environmental compensation data: This refers to the temperature and ambient light data of the microenvironment in which the mouse is located, which is used for data quality verification and more complex compensation algorithm verification in the cloud.
[0116] (Calibrated) thermal image data: data containing complete palm temperature distribution information, which is the core raw material for in-depth analysis.
[0117] Contact state data: namely pressure or capacitance distribution maps, used to reconstruct the spatial positioning process of the palm in the cloud, or for more advanced grip posture analysis.
[0118] 2. Macro-micro fusion analysis of cloud servers Cloud servers possess virtually unlimited computing resources and storage space, enabling them to run extremely complex and resource-intensive machine learning models. After receiving data uploaded by the mouse, it performs the following analysis: Data preprocessing and fusion: First, the cloud, with its powerful computing capabilities, may employ more complex algorithms than those used on the mouse to perform secondary calibration and enhancement of the heatmap data, and accurately register anatomical regions.
[0119] Integrating with macro-level data sources: This is the core advantage of cloud analytics. Cloud servers are configured to access and integrate external macro-level data sources, such as: Real-time meteorological data: Acquires ambient temperature, humidity, air pressure, etc., of the user's location. This data is crucial for interpreting fluctuations in the user's body temperature (e.g., distinguishing between elevated body temperature due to hot weather and internal physiological reasons).
[0120] Epidemiological data: Access the database of public health departments to obtain information on the activity, circulating strains, and incidence rates of infectious diseases such as influenza in the region.
[0121] Perform predictive modeling: The cloud server uses the fused dataset (user multimodal physiological data + macro-environmental and epidemiological data) to train and run disease prediction models.
[0122] For example, the model might find that if a user's "body temperature in the Laogong acupoint area" shows a slight upward trend (+0.3℃) for 12 consecutive hours, and the "influenza incidence rate" in their city is rising rapidly, and meteorological data shows a "sudden drop in temperature," then the model will calculate that the user's risk of developing flu symptoms within the next 48 hours is significantly increased.
[0123] 3. Technical Effects and Value Through this cloud-based collaborative architecture, the present invention achieves the following breakthroughs: From monitoring to prediction: The system is no longer limited to informing users whether there are any abnormalities "currently", but can also provide early warnings of potential health risks "in the future", thus achieving truly proactive health management.
[0124] Combining macro and micro perspectives: Interpreting individual physiological data within the context of the macro environment and epidemiology greatly improves the accuracy and contextual relevance of health status assessment and prediction, and reduces misjudgments caused by environmental factors.
[0125] Unlimited algorithm upgrades: Complex algorithm models are deployed in the cloud and can be continuously iterated and updated without requiring users to replace their mouse hardware, ensuring the cutting edge of the technology.
[0126] Achieving insights into population health: By aggregating data from all users under the premise of anonymization, regional "health heat maps" can be generated, providing valuable big data resources for public health research.
[0127] The technical solution of the present invention will now be described in detail and clearly with reference to the accompanying drawings and embodiments. It should be noted that the following description is intended to explain the present invention, not to limit its scope of protection.
[0128] This embodiment demonstrates the hardware configuration of a health status detection system based on embedded infrared thermal imaging in a mouse, including: Infrared thermal imaging module 10: Employs a FLIR Lepton 3.5 miniature infrared sensor, embedded in the front area of the mouse base, with a germanium infrared-transmitting window at the front. Its resolution is 160x120, frame rate is 5Hz, temperature range is 28-42℃, and accuracy is ±0.3℃.
[0129] Contact sensing array 20: It adopts a 64-point ITO coated capacitive touch sensing array, which is attached to the inner surface of the mouse top cover and can sense the capacitance changes caused by the palm covering.
[0130] Environmental compensation module 40: It includes a TI TMP117 high-precision digital temperature sensor and an AMS AS7341 eleven-channel multispectral sensor, which are integrated on a single board and mounted on the top of the mouse tail, directly exposed to the ambient air through a micro ventilation hole.
[0131] AI Analysis Unit 30: The main controller uses STMicroelectronics' STM32U5 series MCU, which has a built-in Cortex-M33 core and AI accelerator (NPU) to run lightweight neural network algorithms.
[0132] Data transmission module 50: adopts Nordic nRF5340 dual-core Bluetooth 5.2 chip, which supports low-power data transmission.
[0133] Workflow: When the user holds the mouse, the capacitive sensing array 20 detects valid contact and wakes up the system.
[0134] The environmental compensation module 40 collects ambient temperature T_env and ambient light spectrum data in real time.
[0135] Infrared sensor 10 acquires a frame of palm thermal image, AI analysis unit 30 simultaneously reads T_env, and uses pre-stored compensation curves to perform pixel-by-pixel calibration of the thermal image.
[0136] AI analysis unit 30 analyzes the capacitance distribution map and calculates the center coordinates (i.e., spatial location information) of the palm area.
[0137] Using this coordinate system, the calibrated heatmap is translated and rotated, and then registered with a pre-stored anatomical model to accurately select the Laogong acupoint area.
[0138] The average temperature for the area is calculated and input into the health analysis model. The model first compares this temperature to the user's personal health baseline established over the past 7 days; if the deviation exceeds 0.8°C, it triggers a warning.
[0139] Meanwhile, the model calculates the slope between the current temperature value and the value 1 minute ago. If the slope is greater than 0.5℃ / min, it is determined to be a short-term stress response.
[0140] Based on the above results, the warning message "Rapidly elevated body temperature detected, please relax and rest" is sent to the computer via Bluetooth for display.
[0141] Further advanced implementations of cloud-based collaboration include: The mouse uploads encrypted and anonymized data (including heatmap data, environmental data, and contact status data) to the cloud server via the data transmission module 50 (Wi-Fi functionality is added in this example).
[0142] Cloud servers possess powerful computing capabilities, enabling them to run more complex disease prediction models. These models correlate received individual data with macro-level data (such as local real-time weather data, regional influenza virus strains, and incidence rates).
[0143] For example, the cloud-based model detected an increase in the incidence of influenza in the user's city, and the user's average daily body temperature over the past three days showed a slight upward trend of 0.4°C from the baseline. Simultaneously, mouse activity data (such as a decrease in click frequency) indicated reduced activity. Based on this comprehensive analysis, the model determined the user's risk level for catching a cold to be "medium to high."
[0144] The cloud sends the risk level command to the mouse, which then controls its RGB light strip to emit an orange breathing light as a warning. Simultaneously, a more detailed health suggestion is generated on the computer's health assistant software: "A slight increase in health risk has been detected. It is recommended to increase water intake and pay attention to observation." Furthermore, this application's solution is based on a multimodal data acquisition and analysis process of a health status detection system embedded in a mouse, which uses infrared thermal imaging to sense health conditions. Through multi-stage processing of data such as temperature, it achieves accurate monitoring and tiered early warning of the user's health status, as detailed below: I. Process Initiation and Data Acquisition The process begins with "Start" and proceeds to the "Multimodal Data Synchronous Acquisition" stage. The detection system, based on an embedded infrared thermal imaging sensor in the mouse, simultaneously collects various data, including palm thermal image data acquired by the infrared thermal imaging module and palm contact state data collected by the contact sensing array. This multi-dimensional data provides the foundation for subsequent analysis.
[0145] II. Basic Verification and Data Assurance After data collection, a "basic verification" is performed to determine data completeness. If "data is missing," resampling is triggered to ensure sufficient and effective data support for subsequent analysis. If "data is complete," the process proceeds to the "thermograph-vascular mapping analysis" stage, where the calibrated thermograph data is registered with a pre-set hand anatomical feature model. This maps the temperature distribution in the thermograph to the thenar eminence, hypothenar eminence, and Laogong acupoint in the model, clarifying the anatomical location corresponding to the temperature data.
[0146] III. Dynamic Baseline Comparison and Anomaly Detection Next, "dynamic baseline comparison" is performed, which establishes a personal health baseline model based on the user's historical temperature data and compares the real-time monitored temperature data with this model. Then, "short-term anomaly detection" is performed. If the temperature of a preset area (such as the Laogong acupoint) rises sharply by more than 0.5℃ per unit time (1 minute) (i.e., ΔT > 0.5℃ / min), it indicates that a stress response may have occurred, and the "stress response protocol will be activated"; if ΔT is normal, "long-term trend analysis" is performed.
[0147] IV. Long-term trend analysis and microcirculation assessment In the "Long-Term Trend Analysis," if a user's daily basal body temperature fluctuation exceeds 1.2°C, it indicates a possible immune system abnormality, and a "Microcirculation Assessment" will be performed. If the daily fluctuation is normal, the "Personal Baseline" will be updated to more accurately reflect the user's health status. After entering the "Microcirculation Assessment," the microcirculation resistance index will be calculated based on the temperature gradient between specific anatomical feature points to assess the user's vascular activity.
[0148] V. Vascular Activity Assessment and Grading Early Warning The judgment is based on the "vasoactivity index," namely the microcirculation resistance index: If the microcirculation resistance is greater than 2.5, a "Level 3 warning" will be generated and sent to the user or relevant medical personnel via "emergency cloud push" so that timely measures can be taken. If 1.5 < microcirculation resistance ≤ 2.5, a "Level 2 warning" will be generated, and a "local audible and visual alarm" will be triggered to remind the user to pay attention. If the microcirculation resistance is ≤1.5, a "Level 1 prompt" will be generated, and a "mild reminder" will only be given on the interface.
[0149] VI. Record keeping and process completion Finally, regardless of the type of warning or alert, the decision-making basis will be "recorded" for easy review and analysis later, after which the process "ends".
[0150] In summary, this solution, through multi-step and multi-dimensional data analysis and processing, can accurately monitor users' health status and provide tiered early warnings based on different situations, thus providing effective support for users' health management.
[0151] The above description has been given for purposes of illustration and description. Furthermore, this description is not intended to limit the embodiments of this application to the forms disclosed herein. Although numerous exemplary aspects and embodiments have been discussed above, those skilled in the art will recognize certain variations, modifications, alterations, additions, and sub-combinations thereof.
Claims
1. A health status detection system based on embedded infrared thermal imaging in a mouse, comprising a mouse body and a monitoring system disposed within the mouse body, characterized in that, The monitoring system includes: The infrared thermal imaging module is used to collect the temperature field of the area covered by the user's palm and generate thermal image data; A contact sensing array is disposed on the palm contact surface of the mouse body. The contact sensing array is a capacitive touch sensing array, which is used to sense the contact state of the user's palm on the mouse surface and generate capacitance change distribution. An environmental compensation module is used to measure the environmental parameters of the environment in which the mouse is located and generate environmental compensation data; the environmental parameters include ambient temperature and / or ambient light intensity. The AI analysis unit is electrically connected to the infrared thermal imaging module, the contact sensing array, and the environmental compensation module, and is configured as follows: The heat map data is calibrated based on the environmental compensation data; The spatial position information of the palm on the mouse is determined based on the capacitance change distribution. Specifically, this includes: marking the point with the highest capacitance change as the Laogong acupoint anchor point, marking the point with the second highest capacitance change as the thenar eminence anchor point, and recording the coordinates of the two anchor points in the mouse coordinate system; comparing the actual distance between the two anchor points with the standard distance of the corresponding anchor points in the preset palm anatomical feature model; when the deviation of the actual distance from the standard distance exceeds 5mm, it is determined that the grip posture is abnormal and the capacitance change distribution is re-acquired until the deviation does not exceed 5mm. Based on the spatial location information, the calibrated heat map data is registered with the palm anatomical feature model. Specifically, this includes: determining the scaling ratio according to the actual distance and the standard distance and scaling the palm anatomical feature model; rotating the palm anatomical feature model according to the direction of the line connecting the two anchor points; and translating the palm anatomical feature model until the corresponding anchor point in the model coincides with the two anchor points in the mouse coordinate system, so as to calibrate the temperature values corresponding to the thenar eminence, hypothenar eminence and Laogong acupoint in the heat map data. Verify whether the temperature distribution after registration conforms to a preset physiological law; the preset physiological law includes the thenar eminence temperature being 0.5-1.5℃ higher than the hypothenar eminence temperature; when the temperature difference is not within the range, readjust at least one of the scaling ratio, rotation angle, or translation amount and perform registration again until the temperature difference is within the range. A time-series analysis was performed on the temperature data of the thenar eminence, hypothenar eminence, and Laogong acupoint, and the microcirculation resistance index was calculated based on the temperature gradient between Laogong acupoint and the thenar eminence to assess the user's vascular activity status. The time-series analysis included: identifying a stress response when the temperature in a preset area suddenly rises by more than 0.5°C per unit time; identifying an immune system abnormality when the user's daytime basal body temperature fluctuates by more than 1.2°C; and establishing a personal health baseline model based on the user's historical temperature data, comparing the real-time monitored temperature data with the personal health baseline model to identify changes in health status. The system acquires user mouse operation behavior data, and based on the results of the time-series analysis and the vasoactivity status, combined with the temperature values of one or more areas of the thenar eminence, hypothenar eminence, and Laogong acupoint, and the mouse operation behavior data, assesses the user's health status through a health analysis model.
2. The detection system for sensing health status based on embedded infrared thermal imaging in a mouse according to claim 1, wherein the monitoring system further includes an early warning unit, and the health analysis model includes an early warning unit; The AI analysis unit is configured to control the early warning unit to issue an alarm signal when an abnormal health status is assessed.
3. The health status detection system based on embedded infrared thermal imaging in a mouse according to claim 1, characterized in that, The monitoring system also includes: The data transmission module is used to send the user's health status assessment results or the heat map data to an external device via wired or wireless means.
4. The health status detection system based on embedded infrared thermal imaging in a mouse according to claim 3, characterized in that, The AI analysis unit is also configured to send the environmental compensation data, the heat map data, and the contact status to a preset cloud server via the data transmission module. The cloud server predicts the user's health status based on the environmental compensation data, the heat map data, and the contact status, combined with meteorological and epidemiological data.
Citation Information
Patent Citations
Multichannel health mouse system and health condition monitoring method
CN107153470A
Novel peripheral device combining traditional Chinese medicine diagnosis and health detection and application method thereof
CN113377218A
Palm temperature monitor for traditional Chinese medicine diagnosis
CN119833125A
