A high-resolution intelligent non-inductive multi-mode detection and human health monitoring device

Abstract

The application discloses a high-resolution intelligent non-inductive multi-mode detection and human health monitoring device, which comprises a high-resolution variable-multiple double-focal-plane imaging optical detection subsystem, which obtains high-resolution detection images of different sensitive areas of a human body, and realizes non-inductive rapid measurement of multi-mode auscultation information of the human body; a human body sensitive area automatic tracking clear magnification imaging and positioning segmentation acquisition subsystem, which is used for identifying a region of interest or a target, and sequentially performing automatic tracking positioning, automatic focusing clear imaging of the region of interest or the target, and segmentation of the region of interest or the target, and realizing high-resolution clear imaging of the region of interest or the target; a wearable and / or external sensing subsystem, which is used for acquiring physiological parameter indexes of the human body; and an AI-based human multi-physiological-index health analysis and risk prediction and early warning subsystem, which performs large model, large data and AI analysis based on multi-mode auscultation information and physiological parameter indexes of the human body to realize prediction and early warning of a human health risk and disease diagnosis.

Description

Technical Field

[0001] This invention relates to the fields of biomedical research and clinical diagnostic technology, specifically to a high-resolution intelligent non-invasive multi-modal detection and human health monitoring device. Background Technology

[0002] Human visual diagnosis plays a vital role in the diagnosis of human diseases and the prediction and early warning of health risks in both traditional Chinese and Western medicine. However, current methods of human visual diagnosis suffer from several common key technical problems, including: 1) Most rely primarily on human eye observation, which suffers from insufficient resolution in obtaining information from sensory areas, limited color differentiation, inability to store, manage, and quantify information, and significant influence from human factors. Therefore, developing precise and intelligent human visual diagnosis technologies and instruments that can replace human eye observation is crucial. 2) Some visual diagnosis instruments also exist, but most are limited to detecting only one type of visual information, such as instruments for facial diagnosis, tongue diagnosis, eye diagnosis, and fundus examination; a few instruments can detect two types of visual information, such as instruments for facial and tongue diagnosis. Obtaining complete human diagnostic information requires the serial use of multiple instruments, which introduces complexity to user operation, necessitates a large experimental space, and these instruments often have relatively low detection performance. Image acquisition cannot guarantee that the sensory area is centered in the field of view, and the acquired / stored image information contains much redundant and invalid information. Some key physiological and pathological characteristics cannot be accurately acquired, often requiring physical contact with the measurement site or close proximity, causing discomfort such as a feeling of pressure or visual oppression. Therefore, developing high-resolution, intelligent, non-invasive, and precise human diagnostic detection technology and multi-mode intelligent diagnostic instruments capable of simultaneously detecting multiple types of human diagnostic information is crucial.

[0003] Furthermore, while wearable sensing technology has played a positive role in measuring other physiological and biochemical indicators of the human body, the detection functions of existing wearable sensors are relatively limited. Typically, a single wearable sensor can only detect one or two physiological parameters. For example, a PPG sensor measures blood oxygen and pulse; a pressure sensor measures weight and gait; and a temperature sensor measures body temperature. To obtain information on multiple physiological parameters, various types of wearable sensors are required. Therefore, developing wearable sensors capable of simultaneously measuring multiple physiological parameters has significant scientific research value and practical application implications.

[0004] Furthermore, because current diagnostic instruments and wearable sensors are manufactured by different companies, the acquired diagnostic information and physiological parameter measurements are isolated information silos, rarely and difficult to integrate for comprehensive analysis. This leads to biases and omissions in human health analysis, often resulting in inaccurate or serious oversights in complex human health risk assessments. Therefore, it is essential to develop an intelligent analysis device that combines multi-modal diagnostics with wearable sensing to achieve integrated analysis of diagnostic information and physiological parameter measurements, allowing them to complement and corroborate each other, thereby improving the accuracy of complex human health risk assessments and avoiding omissions. Summary of the Invention

[0005] This invention aims to at least solve one of the technical problems existing in the prior art. Therefore, in response to the above-mentioned problems, the object of this invention is to provide a high-efficiency, fast, and low-cost high-resolution intelligent non-contact multi-modal detection and human health monitoring device, which can improve the accuracy of diagnosis of complex human diseases, health risk assessment, and predictive early warning.

[0006] To achieve the above-mentioned objectives, the present invention adopts the following technical solution: In a first aspect, the present invention provides a high-resolution intelligent non-intrusive multi-mode detection and human health monitoring device, comprising a high-resolution variable-magnification dual-focal-plane imaging optical detection subsystem, a human sensory area automatic tracking clear magnification imaging and positioning segmentation acquisition subsystem, a wearable and / or external sensing subsystem, and an AI-based human multi-physiological indicator health analysis and risk prediction and early warning subsystem, wherein: The high-resolution variable-magnification dual-focal-plane imaging optical detection subsystem is used to obtain high-resolution detection images of different human sensory areas, enabling non-invasive, rapid, and comprehensive measurement of multi-modal diagnostic information in traditional Chinese medicine, such as height, weight, facial features, tongue image, eye image, ear image, hand image, and skin. The automatic tracking, clear magnification imaging, positioning, and segmentation acquisition subsystem for human regions of interest is used to identify regions of interest or targets, and then perform automatic tracking and positioning, automatic focusing and clear imaging, and segmentation of regions of interest or targets in sequence, so as to achieve high-resolution clear imaging, acquisition, and storage of regions of interest or targets. The wearable and / or external sensing subsystem is worn on or in contact with important parts of the body and is used to acquire physiological parameters of the human body. The AI-based human multi-physiological indicator health analysis and risk prediction and early warning subsystem uses large-scale models, big data and AI analysis based on human multi-modal diagnostic information and human physiological parameter indicators to achieve prediction and early warning of human health risks and disease diagnosis.

[0007] In some possible implementations, the high-resolution variable-magnification dual-focal-plane imaging optical detection subsystem includes a front lens group, an intermediate lens group, a rear lens group, an automatic zoom control system, and a CCD detector. The front lens group is positioned near the focal plane for human detection. The intermediate lens group moves back and forth under the control of the automatic zoom control system to adjust the magnification. The CCD detector's sensing surface is located at the focal plane of the rear lens group to acquire human imaging signals. Different detection areas of the human body share a single optical detection subsystem, obtaining high-resolution detection images of different sensory areas of the human body. This enables non-invasive, rapid, and comprehensive measurement of various aspects of traditional Chinese medicine diagnosis, including height, weight, facial features, tongue image, eye image, ear image, hand image, and skin texture.

[0008] In some possible implementations, the high-resolution variable-magnification dual-focal-plane imaging optical detection subsystem further includes a position sensor and a microprocessor AI analysis system. The position sensor is used to measure the distance between the human body and the CCD detector, and transmits the distance value to the microprocessor AI analysis system. The microprocessor AI analysis system controls the imaging lens set in front of the CCD detector to complete the corresponding adjustment control based on the distance value, so that the human body can finally obtain a clear image on the CCD detector.

[0009] In some possible implementations, a light source is placed between the front lens and the human body. By designing the shape, size, and number of the light source, it can be used to achieve planar shadowless illumination, three-dimensional shadowless illumination, or four-directional independent controllable backlight shadowless illumination for the human body.

[0010] In some possible implementations, the identification of the region of interest or target is achieved through the YOLOv5 small target identification and analysis algorithm, and the optimal sharpness adjustment is automatically performed by searching for the maximum grayscale variance based on the hill-climbing algorithm. This enables high-resolution clear imaging, image acquisition, and image storage processing only for the region of interest or target, eliminating the influence of non-interested targets and background images, and saving storage space.

[0011] In some possible implementations, the wearable and / or external sensing subsystem is a wearable and / or sensor positioned within the range of human activity, which is used to perform rapid dynamic measurements of physiological parameters in the human sensory areas in a wearable contact manner or by human footsteps, without being noticed or interfering with daily life and work.

[0012] In some possible implementations, the sensor is worn on a vital part of the body to acquire physiological indicators of interest to the human body, including one or more of pressure sensors, photoelectric sensors, gas sensors, temperature and humidity sensors, inertial attitude sensors, and GPS position sensors, for collecting signals such as human physiological indicators, body posture, and location; or / and, the sensor is installed in the human activity area to acquire physiological indicators of the human body in the area of ​​interest, including one or more of pressure sensors, sound sensors, ultrasonic or laser rangefinders, infrared sensors, and infrared CCDs. When pressure, sound, distance, or temperature of the human body is detected, the sensor automatically triggers the acquisition of signals of human physiological indicators, body posture, and location, and calculates human height, weight, body temperature, blood circulation status, rheumatism, or performs a medical consultation.

[0013] In some possible implementations, when the sensor is a pressure sensor or a photoelectric sensor, using only one of the pressure sensor or photoelectric sensor, by establishing a matching theoretical model and AI algorithm analysis method, it is possible to simultaneously measure and provide multiple physiological indicators of the human body: pulse, blood pressure, heart rate, electrocardiogram related information, respiration and blood oxygen; it can also be combined with the gas sensor, temperature and humidity sensor, inertial attitude sensor and GPS position sensor to provide monitoring of disabled persons' turning over in bed, falls, urinary and fecal incontinence, and to locate their current location information.

[0014] Some possible implementations involve using large-scale models, big data, and AI analysis based on multi-modal visual diagnosis information and physiological parameters of the human body to predict and warn of health risks and diagnose diseases, including: Deep learning algorithms are used to extract features from multimodal visual diagnosis information and physiological parameters of the human body, such as pulse wave, blood pressure, heart rate, ECG related information, respiration, blood oxygen, height, weight, body shape, facial / tongue / ear / hand blemishes, blood vessels, texture, color, sclera / iris / fundus of the eyes, spots, blood vessels, color, shape, skin scars and other physiological and pathological features. The extracted features are compared using a knowledge graph of a big data expert system that pre-defines human physiological and pathological characteristics. Based on the similarity of features, AI analysis provides diagnostic and early warning results, enabling the prediction and early warning of human health risks and disease diagnosis, and providing suggestions on medication, diet therapy, and physical therapy.

[0015] In some possible implementations, the AI-based human multi-physiological indicator health analysis and risk prediction and early warning subsystem is deployed on cloud servers, local servers, single computers or mobile terminals, etc.

[0016] Because the present invention adopts the above technical solution, it has the following characteristics: 1. This invention employs a high-resolution variable-magnification dual-focal-plane imaging optical detection subsystem structure that maintains sufficient working distance to achieve visually comfortable, non-contact image acquisition. The sensing detection is wearable and other methods, enabling multi-parameter non-intrusive measurement of human visual diagnosis and sensing. This achieves pressure-free perception of multi-modal visual diagnosis information and various physiological indicators, and rapid measurement without unintentional or interference with daily life and work.

[0017] 2. The high-resolution variable-magnification dual-focal-plane imaging optical detection subsystem of the present invention optimizes the use of numerical aperture of the front / rear lens groups, sharing a single optical detection subsystem to obtain high-resolution detection images of different human sensory areas, such as eye image measurement resolution reaching the μm level, realizing non-invasive, rapid and comprehensive measurement of multimodal TCM diagnostic information such as human height, weight, facial image, tongue image, eye image, ear image, hand image, and skin.

[0018] 3. The automatic tracking, clear magnification imaging, and positioning segmentation acquisition subsystem for the human body's area of ​​interest of this invention can improve the detection performance of the instrument, ensure that the area of ​​interest is in the center of the field of view during image acquisition, significantly reduce redundant and invalid information in the acquired / stored image information, and ensure that all key information of human physiological and pathological characteristics is accurately acquired. This image acquisition only saves the information of the target of interest, while automatically removing all other non-target information, which greatly saves storage space, effectively saves subsequent analysis and processing computing resources, and improves the anti-interference ability of AI analysis.

[0019] 4. The wearable and / or external sensing subsystem of the present invention can simultaneously measure multiple physiological indicators such as human pulse, blood pressure, heart rate, electrocardiogram related information, and respiration using only one sensor and matching theoretical model algorithm analysis method; it can also use gas sensors, temperature and humidity sensors, inertial attitude sensors, and GPS position sensors for joint measurement to provide monitoring of disabled persons such as turning over in bed, falling, and urinary and fecal incontinence, and provide current location information.

[0020] 5. The AI-based human multi-physiological indicator health analysis and risk prediction and early warning subsystem of this invention integrates multi-modal diagnostic information with various human physiological indicators, such as pulse, blood pressure, heart rate, ECG-related information, respiration, blood oxygen, height, weight, body shape, facial / tongue / ear / hand blemishes, blood vessels, texture, and color, and blemishes, spots, blood vessels, color, shape, and skin scars on the sclera / iris / fundus of the eyes. It performs big data and large-scale AI model analysis to achieve prediction and early warning of human health risks and disease diagnosis, and can provide suggestions on medication, diet therapy, and physical therapy. For example, based on the infiltration state of gray haze and yellow halo at the edge of the iris and changes in iris microstructure, it can predict and warn of the risk of Alzheimer's disease; based on the distribution and color characteristics of spots on the iris, and the distribution, color characteristics, and shape characteristics of spots and blood vessels on the sclera, it can predict and warn of the risk of cardiovascular and cerebrovascular diseases, diabetes, tumors or cancer, and polycystic ovary syndrome; and based on the characteristics of skin scars, it can predict and warn of the risk of skin diseases.

[0021] In summary, this invention can be widely applied in comprehensive human body diagnosis using both traditional Chinese and Western medicine, measurement of various physiological indicators, precision medicine disease diagnosis, and health risk prediction and early warning. Attached Figure Description

[0022] Various other advantages and benefits will become apparent to those skilled in the art upon reading the following detailed description of preferred embodiments. The accompanying drawings are for illustrative purposes only and are not intended to limit the invention. Throughout the drawings, the same reference numerals denote the same parts. In the drawings: Figure 1 This is a schematic diagram illustrating the basic structural principle of the high-resolution intelligent non-contact multi-mode detection technology and human health monitoring system according to an embodiment of the present invention. Figure 2a This is a schematic diagram of the principle of the high-resolution variable magnification dual-focal-plane imaging optical detection subsystem according to an embodiment of the present invention; Figure 2b This is a schematic diagram illustrating how a sensor installed in the human activity area automatically triggers the acquisition of human physiological indicator signals, according to an embodiment of the present invention. Figure 3a This is a schematic diagram of the workflow of the automatic tracking, clear magnification imaging, positioning, segmentation, and acquisition subsystem for human sensory areas according to an embodiment of the present invention; Figure 3b This is a schematic diagram illustrating the automatic tracking and clear imaging positioning of the human sensory area according to an embodiment of the present invention; Figure 4a A schematic diagram illustrating the simultaneous measurement of multiple physiological indicators using a single sensor in an embodiment of the present invention is provided. Figure 4b This is a schematic diagram of a human pulse wave signal measured by a single sensor according to an embodiment of the present invention; Figure 4c This is a schematic diagram of respiratory signals measured simultaneously by a single sensor according to an embodiment of the present invention. Figure 4d This is a schematic diagram of the heartbeat signal measured simultaneously by a single sensor according to an embodiment of the present invention; Figure 4e This is a schematic diagram of the BCG signal given by a single sensor simultaneous measurement according to an embodiment of the present invention; Figure 4f This is a schematic diagram of the heart sound signal measured simultaneously by a single sensor according to an embodiment of the present invention; Figure 5a This is a schematic diagram illustrating AI-based multi-physiological indicator health analysis and Alzheimer's disease risk prediction and early warning in accordance with an embodiment of the present invention. Figure 5b This is a schematic diagram illustrating AI-based multi-physiological indicator health analysis and cardiovascular disease risk prediction and early warning in accordance with an embodiment of the present invention. Figure 5c This is a schematic diagram illustrating AI-based multi-physiological indicator health analysis and diabetes risk prediction and early warning in accordance with an embodiment of the present invention. Figure 5d This is a schematic diagram illustrating AI-based multi-physiological indicator health analysis and tumor or cancer risk prediction and early warning in accordance with an embodiment of the present invention. Detailed Implementation

[0023] It should be understood that the terminology used herein is for the purpose of describing particular exemplary embodiments only and is not intended to be limiting. Unless the context clearly indicates otherwise, the singular forms “a,” “an,” and “described” as used herein may also include the plural forms. The terms “comprising,” “including,” “containing,” and “having” are inclusive and therefore indicate the presence of the stated features, steps, operations, elements, and / or components, but do not exclude the presence or addition of one or more other features, steps, operations, elements, components, and / or combinations thereof. The method steps, processes, and operations described herein are not construed as requiring them to be performed in a particular order described or illustrated unless the order of performance is explicitly indicated. It should also be understood that additional or alternative steps may be used.

[0024] Although terms such as first, second, third, etc., may be used in this document to describe multiple elements, components, regions, layers, and / or segments, these elements, components, regions, layers, and / or segments should not be limited by these terms. These terms may be used only to distinguish one element, component, region, layer, or segment from another. Unless the context clearly indicates otherwise, terms such as "first," "second," and other numerical terms used herein do not imply order or sequence. Therefore, the first element, component, region, layer, or segment discussed below may be referred to as the second element, component, region, layer, or segment without departing from the teachings of the exemplary embodiments.

[0025] For ease of description, spatial relative terms may be used in the text to describe the relationship of one element or feature relative to another element or feature as shown in the figure. These relative terms include, for example, "inside," "outside," "middle," "outer," "below," "above," etc. Such spatial relative terms are intended to include different orientations of the device in use or operation, other than those depicted in the figure.

[0026] This invention provides a high-resolution intelligent non-contact multi-modal detection and human health monitoring device, comprising a high-resolution variable-magnification dual-focal-plane imaging optical detection subsystem, a human sensory area automatic tracking clear magnification imaging and positioning segmentation acquisition subsystem, a wearable and / or external sensing subsystem, and an AI-based human multi-physiological indicator health analysis and risk prediction and early warning subsystem. The combination of these subsystems enables non-contact measurement of multiple parameters through a combination of visual inspection and sensing. By employing visually comfortable, non-contact, or wearable methods, it achieves rapid measurement of various physiological indicators through a combination of multi-modal visual inspection and wearable sensing without pressure perception, unintentionally, or without interfering with daily life and work. Therefore, this invention enables rapid, high-resolution detection of multiple human physiological parameters and human health prediction and early warning.

[0027] Exemplary embodiments of the invention will now be described in more detail with reference to the accompanying drawings. While exemplary embodiments of the invention are shown in the drawings, it should be understood that the invention can be implemented in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided to enable a more thorough understanding of the invention and to fully convey the scope of the invention to those skilled in the art.

[0028] Example 1: The high-resolution intelligent non-intrusive multi-modal detection and human health monitoring device provided in this example includes a high-resolution variable-magnification dual-focal-plane imaging optical detection subsystem, an automatic tracking, clear magnification imaging and positioning segmentation acquisition subsystem for human sensory areas (areas or targets of interest, such as face, tongue, eyes, ears, hands, etc.), a wearable and / or external sensing subsystem, and an AI-based human multi-physiological indicator health analysis and risk prediction and early warning subsystem. By combining the above subsystems, multi-parameter non-intrusive measurement combining human visual examination and sensing is performed, achieving non-pressure perception and rapid measurement of multiple physiological indicators without unintentional or interference with daily life and work. The high-resolution variable-magnification dual-focal-plane imaging optical detection subsystem is used to obtain high-resolution detection images of different human sensory areas, enabling non-invasive, rapid, and comprehensive measurement of multi-modal diagnostic information in traditional Chinese medicine, such as height, weight, facial features, tongue image, eye image, ear image, hand image, and skin. The automatic tracking, clear magnification imaging, positioning, and segmentation acquisition subsystem for human body regions of interest is used to identify regions of interest or targets, and then perform automatic tracking and positioning of regions of interest or targets, automatic focusing and clear imaging, and segmentation of regions of interest or targets in sequence, so as to achieve high-resolution clear imaging, image acquisition, and image storage and processing of regions of interest or targets. Wearable and / or external sensing subsystems are used to acquire physiological parameters of the human body by wearing them on or in contact with important parts of the body. The AI-based human multi-physiological indicator health analysis and risk prediction and early warning subsystem is used to store human multi-modal diagnostic information and human physiological parameter information obtained by high-resolution variable magnification dual-focal plane imaging optical detection subsystem and wearable and / or external sensing subsystem. It performs large-scale model, big data and AI analysis to predict and warn of human health risks and diagnose diseases, and can further provide suggestions on medication, diet therapy, physiotherapy and other treatments.

[0029] In a preferred embodiment of the present invention, the high-resolution variable-magnification dual-focal-plane imaging optical detection subsystem adopts a sandwich optical lens combination structure of front-group, middle-group, and rear-group lenses. Human body detection is performed at the focal plane of the front group lens, the middle lens is adjusted for magnification to suit different object sizes, and the focal plane of the rear group lens acquires the object imaging signal. This optical detection subsystem design optimizes the use of the numerical aperture of the front / rear group lenses, achieving high-resolution object imaging quality. Different detection areas of the human body share a single optical detection system, obtaining high-resolution detection images of different human sensory areas, achieving non-invasive and rapid measurement of multi-modal diagnostic information at μm-level optimal resolution. Furthermore, the human body can also be located outside the focal plane of the front group lens; by adjusting the middle lens, the object image can be made clear, achieving non-invasive and rapid measurement of other multi-modal diagnostic information below optimal resolution. Therefore, by using a common optical detection subsystem for different detection areas of the human body, comprehensive measurement of multi-modal TCM diagnostic information such as height, weight, facial features, tongue image, eye image, ear image, hand image, and skin image can be achieved, with a resolution reaching the μm level.

[0030] In a preferred embodiment of the present invention, the human body interest area automatic tracking clear magnification imaging and positioning segmentation acquisition subsystem identifies the target of interest by using YOLOv5 small target recognition and machine learning training of the target of interest feature image intelligent analysis algorithm, and sequentially performs automatic tracking and positioning, magnification adjustment, automatic focusing clear imaging, image segmentation acquisition of interest area or target for storage processing, etc., so as to facilitate high-resolution clear imaging and signal acquisition and storage processing only for the target of interest, eliminate the influence of non-interested targets and background images, and save storage space. Among them, YOLOv5 small target recognition and machine learning are existing algorithms in the prior art, and the specific principles are not described in detail.

[0031] In a preferred embodiment of the present invention, the wearable and / or external sensing subsystem uses sensors to collect human physiological parameters.

[0032] In this embodiment, the sensor can be worn on important parts of the body to acquire physiological indicators of the region of interest. This includes one or more of a pressure sensor, photoelectric sensor, gas sensor, temperature and humidity sensor, inertial attitude sensor, and GPS location sensor, used to collect signals related to human physiological indicators, body posture, and location. Specifically, the sensor can use only a pressure sensor or photoelectric sensor. By establishing a supporting theoretical model and AI algorithm analysis method, it can simultaneously measure and provide information on multiple physiological indicators such as pulse, blood pressure, heart rate, ECG-related information, respiration, and blood oxygenation. Alternatively, a gas sensor, temperature and humidity sensor, inertial attitude sensor, and GPS location sensor can be used for combined measurements to provide monitoring of body temperature, sweating, bed turning, falls, and urinary and fecal incontinence for disabled individuals, and to provide current location information.

[0033] In this embodiment, the sensor may also employ one or more of the following: a pressure sensor, a sound sensor, an ultrasonic or laser rangefinder, an infrared sensor, or an infrared CCD, installed in the external human activity area. When pressure, sound, position, or temperature of the human body is detected, it automatically triggers the acquisition of human physiological indicator signals. For example, a pressure sensor installed on the ground in the activity area can measure human weight and the force exerted on the left and right legs, analyzing human movement stability, limb diseases, etc.; a sound sensor can measure the volume and speed of human speech and / or conduct medical consultations; an ultrasonic or laser rangefinder can measure the position and distance of the human body, and combined with a high-resolution variable-magnification dual-focal-plane imaging optical detection subsystem to measure the size of the human body image, waist circumference, etc., to calculate human height, weight, etc.; an infrared sensor or infrared CCD can non-contactly measure body temperature, human blood circulation status, rheumatism, etc.

[0034] In a preferred embodiment of the present invention, the AI-based human multi-physiological indicator health analysis and risk prediction and early warning subsystem performs large-scale model, big data and AI analysis on the obtained multi-modal diagnostic information and various human physiological indicator information, such as pulse, blood pressure, heart rate, ECG related information, respiration, blood oxygen, height, weight, body fat, facial / tongue / ear / hand blemishes, blood vessels, texture, color, and other features, as well as spots, dots, blood vessels, color, and shape of the sclera / iris / fundus of the eyes, and skin scars, etc., to achieve prediction and early warning of human health risks and disease diagnosis, and to propose suggestions for medication, diet therapy, physiotherapy and other treatments.

[0035] Furthermore, the AI-based subsystem for health analysis and risk prediction and early warning of multiple physiological indicators of the human body can be deployed on processors of different specifications, such as cloud servers, local servers, single computers, and mobile terminals, to perform storage management and AI analysis of human physiological indicators and health information. For example, it can predict and warn of the risk of Alzheimer's disease based on the infiltration state of gray haze and yellow halo at the edge of the iris and changes in the microstructure of the iris; it can predict and warn of the risk of cardiovascular and cerebrovascular diseases, diabetes, tumors or cancer based on the distribution and color characteristics of spots on the iris, and the distribution, color characteristics, and morphological characteristics of spots and blood vessels on the sclera; and it can predict and warn of the risk of skin diseases based on the characteristics of skin scars.

[0036] Because current diagnostic instruments and wearable sensors are manufactured by different companies, the acquired information, such as visual inspection data and physiological parameter measurements, are isolated information silos. They are rarely, if ever, integrated for comprehensive analysis, leading to biases and omissions in human health analysis. This often results in inaccurate or serious oversights in complex health risk assessments. This invention provides a high-efficiency, rapid, and low-cost high-resolution intelligent non-contact multi-modal detection and human health monitoring device that integrates multi-modal visual inspection with wearable sensing, improving the accuracy of complex disease diagnosis, health risk assessment, and predictive warning. The specific applications of this high-resolution intelligent non-contact multi-modal detection and human health monitoring device are detailed below through specific embodiments.

[0037] like Figure 1 As shown, the high-resolution intelligent non-intrusive multi-modal detection and human health monitoring device provided in this embodiment includes: a high-resolution variable-magnification dual-focal-plane imaging optical detection subsystem 1, a human sensory area automatic tracking clear magnification imaging and positioning segmentation acquisition subsystem 2, a wearable and / or external sensing subsystem 3, and an AI-based human multi-physiological indicator health analysis and risk prediction and early warning subsystem 4. The above subsystems are used in combination to perform non-intrusive multi-parameter measurement combining human visual diagnosis and sensing, enabling pressure-free perception and rapid measurement of human multi-modal visual diagnosis information and intelligent sensing of human multi-parameter information without unintentional interference with daily life and work. The high-resolution variable-magnification dual-focal-plane imaging optical detection subsystem 1 is used to obtain high-resolution detection images of different human sensory areas, and realize non-invasive and rapid measurement of multi-modal human diagnostic information. The automatic tracking, clear magnification imaging, positioning, segmentation, and acquisition subsystem 2 for human body receptive areas is used to dynamically process high-resolution detection images, select receptive areas or targets, and perform automatic tracking, clear magnification imaging, and positioning. Ultimately, it achieves high-resolution segmentation, acquisition, and storage of receptive areas or targets. This image acquisition method only saves the information of the receptive targets, while automatically removing all other non-target information, which greatly saves storage space, effectively saves subsequent analysis and processing computing resources, and improves the anti-interference capability of intelligent analysis. Wearable and / or external sensing subsystem 3, used to be worn on important parts of the body and / or placed on the outside of the human body to acquire physiological parameters of interest to the human body; The AI-based human multi-physiological indicator health analysis and risk prediction and early warning subsystem 4 uses human multi-model diagnostic information and human physiological parameter information to perform large-scale model, big data and AI analysis to achieve prediction and early warning of human health risks and disease diagnosis, and to propose suggestions such as medication, diet therapy and physical therapy.

[0038] In this embodiment, as Figure 2a As shown, the high-resolution variable-magnification dual-focal-plane imaging optical detection subsystem includes an illumination source LS, a front lens group L1, an intermediate lens group L2, a rear lens group L3, a forward and backward automatic zoom control system K1, a CCD detector D1, and a microprocessor AI analysis system. Object 01 is detected at the focal plane f1 position of the front lens L1. The middle lens L2 moves back and forth under the control of the automatic zoom control system K1, achieving zoom adjustment from 1× to 20×, suitable for detecting objects of different sizes. The sensing surface of the CCD detector D1 is located at the focal plane f3 position of the rear lens L3. The object imaging signal is acquired under the control of the automatic tracking clear magnification imaging and positioning segmentation acquisition subsystem 2 of the human body interest area. The automatic zoom control system K1 and the automatic tracking clear magnification imaging and positioning segmentation acquisition subsystem 2 of the human body interest area are controlled by receiving instructions from the microprocessor AI analysis system. A light source LS is set between the front lens L1 and the object being measured. The light source LS can adopt shadowless illumination or four-directional independent controllable backlight shadowless illumination in the form of ring, surface, three-dimensional sphere or spherical crown, or other multi-angle methods. The purpose of shadowless illumination is to avoid producing light source shadows, enabling comprehensive measurement of multimodal TCM diagnostic information such as human face image, tongue image, eye image, ear image, hand image, and skin.

[0039] Furthermore, the human body can also be located outside the focal plane of the front lens L1, and the image of the object can be made clear by adjusting the middle lens L2, so as to achieve non-intrusive and rapid measurement of other human body multi-mode diagnostic information below the optimal resolution.

[0040] Furthermore, such as Figure 2bAs shown, it also includes position sensors S1-S2 or a specific-position ultrasonic or laser rangefinder S3 and a microprocessor AI analysis system. Position sensors S1-S2 or the specific-position ultrasonic or laser rangefinder S3 are used to measure the distance from the human body to the CCD detector D1. This distance value is transmitted to the microprocessor AI analysis system, which issues instructions to adjust magnification, image clarity, etc., based on the distance value. The imaging lens L (including a front lens L1, a middle lens L2, a rear lens L3, and a forward / backward automatic zoom control system K1) positioned in front of the CCD detector D1 completes the corresponding adjustment control, ultimately obtaining a clear image on the CCD detector D1. In this system, a light source LS is placed between the imaging lens L and the object being measured. The light source LS can be a ring-shaped, surface-shaped, three-dimensional spherical or crown-shaped, or other multi-angle methods for shadowless illumination or four-directional independent controllable backlight shadowless illumination. The purpose of shadowless illumination is to avoid producing shadows from the light source, enabling comprehensive measurement of multimodal TCM diagnostic information such as human height, weight, posture, facial features, tongue images, eye images, ear images, hand images, and skin. For example, human height, weight, and posture can be measured at a position of approximately Z=5 meters; human facial, tongue, ear, and hand images can be measured at a resolution of 100 micrometers at a position of approximately Z=300 millimeters; and human eye images (including blood vessels, spots, shapes, and texture structures of the sclera and iris) and skin can be measured at a high resolution of 10 micrometers at a position of approximately f1=150 millimeters at the focal plane of the front lens L1. These are just examples, and the methods are not limited to these examples.

[0041] Furthermore, to maintain a sufficient working distance in the optical inspection system structure, the working distance or focal length f1 of the front lens group L1 can be increased to 100mm or more in the design of the high-resolution variable magnification dual-focal-plane imaging optical inspection subsystem 1, such as f1=150mm, 250mm, or 300mm, etc. Figure 2a As shown, this allows for non-contact measurement without visual discomfort or pressure. This is just one example; it is not the only option, and you should choose according to your actual measurement needs.

[0042] In this embodiment, as Figure 3aAs shown, the human body region of interest automatic tracking, clear magnification imaging, and positioning segmentation acquisition subsystem 2 uses YOLOv5 small target recognition and machine learning training of the target of interest feature image intelligent analysis algorithm to identify the region of interest or target, and sequentially performs automatic tracking and positioning, magnification adjustment, automatic focusing and clear imaging, high-resolution segmentation of the region of interest or target, and storage processing. This facilitates high-resolution clear imaging and signal acquisition and storage processing only for the target of interest, eliminating the influence of non-targets of interest and background images, and saving storage space. For example, in the Z=500~200 mm range, it automatically tracks and positions the human head or palm, adjusts the magnification, and automatically focuses and images clearly, extracting the contour areas of the human face, tongue, ears, and palm; in the Z=300~100 mm range, it automatically tracks and positions the human eyes or skin, adjusts the magnification, and automatically focuses and images clearly, extracting the contour areas of the human eyes and skin, etc. Figure 3b As shown in (1) to (5), Figure 3b Image (1) is a diagram showing the result of automatic tracking, positioning, and autofocusing to achieve clear imaging of the human face and eyes. Figure 3b (2) is a diagram showing the result of automatic tracking, positioning, and autofocusing of the human ear to achieve clear imaging. Figure 3b The middle (3) diagram shows the result of automatic tracking and positioning of the human tongue and clear imaging with automatic focusing. Figure 3b The middle (4) diagram shows the result of automatic tracking and positioning of the human hand, and automatic focusing to achieve clear imaging. Figure 3b Image (5) is a diagram illustrating the result of the human eye's automatic tracking, positioning, and autofocus for clear imaging. Figure 3b In steps (1) to (5), the image acquisition only saves the information of the segmented target corresponding to the target in the target recognition box, while automatically removing all other non-target information, such as... Figure 3b The middle (6) shows the Figure 3b The diagram shows the segmentation and storage results of the human eye in (5). This can save storage space significantly, effectively save subsequent analysis and processing computing resources, improve the anti-interference ability of intelligent analysis, and of course, it can also automatically track and locate the whole human body, adjust the magnification, automatically focus and clearly image, extract the whole human body contour area, etc. in the range of Z=10~2 meters, which is convenient for measuring human height, weight, posture, etc.

[0043] In this embodiment, as Figures 4a-4f As shown, the wearable and / or external sensing subsystem 3 can simultaneously measure and provide multiple physiological indicators such as human pulse rate, blood pressure, heart rate, ECG-related information, respiration, and blood oxygen by collecting pulse wave signals based on AI algorithm analysis methods using pressure sensors or photoelectric sensors and supporting theoretical models. It can also perform intelligent pulse diagnosis based on pulse waveform combined with traditional Chinese medicine theory and experience.

[0044] exist Figure 4a In this context, based on the algorithm analysis method of pressure sensor or photoelectric sensor and supporting theoretical model, the pulse rate M, heart rate X, respiratory rate H, etc. can be calculated from the time difference Δt between the peak values ​​of adjacent pulse waves. M=60 / Δt (beats / minute), X=M (beats / minute), H=M / 4 (or 5) (beats / minute).

[0045] like Figure 4a As shown, the algorithmic analysis method based on pressure sensors or photoelectric sensors and supporting theoretical models can establish a theoretical formula model of the relationship between pulse waves and blood pressure from the collected pulse wave feature points, and calculate the blood pressure value. The formula for calculating blood pressure (BP) is as follows: (1); Where RWTT is the time difference between t2 and t1 (RWTT=t2-t1), t1 is the time difference from the start of the pulse wave to the highest point of the pulse wave P1 peak within one pulse wave cycle, and t2 is the time difference from the start of the pulse wave to the highest point of the pulse wave P2 peak within one pulse wave cycle. AI r The ratio between R2 and R1 (AI) r =R2 / R1), R1 is the signal value corresponding to the highest point of the pulse wave P1 peak, R2 is the signal value corresponding to the highest point of the pulse wave P2 peak; HR is the reciprocal of the pulse wave period T (HR=1 / T), a, b, c, d are parameters in the theoretical formula model. The AI ​​large model is established through pre-training with hundreds of millions of open source data and tens of thousands of clinically collected pulse wave and blood pressure data, and is calculated by a deep learning model based on the actual collected human pulse wave signals and human physiological characteristics.

[0046] like Figure 4a As shown, an algorithmic analytical method based on pressure sensors or photoelectric sensors and supporting theoretical models can establish a theoretical formula model of the relationship between pulse wave and electrocardiogram (ECG) related information from the collected pulse wave signals, and calculate ECG related information signals. For continuous pulse wave signals... ECG-related information signals obtained through bandpass filtering The calculation formula is: (2); in, and These are the low and high cutoff frequencies of the filter, respectively. This indicates a bandpass filtering operation, which can be implemented using a digital filter. For discrete pulse wave signals... ECG-related information signals obtained by sampling and bandpass filtering It can be represented in convolution form as: (3); in, For the filter impulse response, This represents the filter order. By adjusting different cutoff frequencies, simultaneous extraction of respiration, heartbeat, BCG, and heart sounds can be achieved. The specific frequency bands are shown in Table 1 below.

[0047] Table 1

[0048] Figure 4b The image shown is an example illustration of the obtained human pulse wave signal; Figure 4c The corresponding respiratory signal showed that the subject's average respiratory rate within a 15-second time window was 22.7 breaths per minute; Figure 4d This corresponds to the heart rate signal, showing an average heart rate of 76.1 beats per minute; Figure 4e The corresponding BCG signal shows a typical W-shaped systolic wave group with a relatively weak diastolic component, which is highly consistent with the cardiac cycle characteristics of healthy resting subjects. Figure 4f These are the corresponding heart sound signals. The first heart sound S1 and the second heart sound S2 in the figure are clearly distinguishable, indicating that the present invention can effectively analyze the fine features of electrocardiogram-related multimodal physiological signals through pulse waves.

[0049] like Figure 4a As shown, based on the photoelectric sensor and its supporting theoretical model, the algorithmic analytical method can establish a theoretical formula model for the relationship between pulse wave and blood oxygen from the pulse wave signal collected by the red / infrared dual-wavelength PPG (photoplethysmography, PPG) sensor, and calculate the blood oxygen content. The formula for calculating the blood oxygen content S is as follows: (4); Where R is the "relative pulsation absorption ratio" between the 660nm red light channel and the 940nm infrared light channel. and These represent the wavelengths of oxyhemoglobin and deoxyhemoglobin, respectively. =Extinction coefficient at 660nm or 940nm It is the ratio of the average effective optical path length of 660nm red light to 940nm infrared light in the volume of a pulsating artery.

[0050] Furthermore, the wearable and / or external sensing subsystem 3 may also employ one or more of the following: pressure sensor, photoelectric sensor, gas sensor, temperature and humidity sensor, inertial attitude sensor, and GPS position sensor, worn on important parts of the body to acquire physiological indicators of interest, such as pulse, blood oxygen, body temperature, sweating, and other important physiological indicators, as well as information such as bed turning over, falls, urinary and fecal incontinence, and geographical location.

[0051] Furthermore, the wearable and / or external sensing subsystem 3 can also employ one or more of the following: pressure sensor, sound sensor, ultrasonic or laser rangefinder, infrared sensor, and infrared CCD. These can be installed in the human activity area and automatically trigger the acquisition of human physiological indicator signals when pressure, sound, or temperature is detected. For example, a pressure sensor can be installed on the ground in the activity area. Figure 2b As shown, the weight of a human body during walking motion can be measured as G = (G S1 +G S2 ) / 2, Weight at rest G=G S1 +G S2 And the force distribution on the left and right legs of the human body G S1 G S2 Among them, G S1 G S2 These correspond to the pressure values ​​measured by pressure sensors S1 and S2, respectively; ultrasonic or laser rangefinders can measure the distance to the human body, and then combine this with a high-resolution variable-magnification dual-focal-plane imaging optical detection subsystem to measure the size of the human body image, waist circumference, etc., to calculate the human height h1 = h2 × K, where K is the current actual magnification of the high-resolution variable-magnification dual-focal-plane imaging optical detection subsystem, h2 is the size of the full-body image obtained by CCD detector D1, and combined with the pressure sensor to measure the human body weight, the human body fat G / h1 is calculated; sound sensors can measure the volume and speed of human speech and / or conduct medical consultations; infrared sensors or infrared CCDs can non-contactly measure body temperature, human blood circulation status, rheumatism, etc.

[0052] In summary, the wearable and / or external sensing subsystem 3 is wearable and / or placed within the range of human activity, including but not limited to being made into a wristwatch, or being able to be attached to body parts, or being able to be bound to underwear, or being placed on the floor, etc., to perform rapid dynamic measurements of human sensory areas in an unintentional or non-interfering manner through contact or stepping.

[0053] In this embodiment, the AI-based human multi-physiological indicator health analysis and risk prediction and early warning subsystem 4 performs large-scale model, big data and AI analysis on the obtained multi-modal diagnostic information and various human physiological indicator information, such as pulse, blood pressure, heart rate, ECG-related information, respiration, blood oxygen, height, weight, body shape, facial / tongue / ear / hand blemishes, blood vessels, texture, color, and other features, as well as blemishes, spots, blood vessels, color, and shape of the sclera / iris / fundus of the eyes, and skin scars. Deep learning algorithms such as YOLOv5 and Unet are used to extract features from multimodal visual diagnosis information and physiological parameters of the human body, such as pulse wave, blood pressure, heart rate, ECG related information, respiration, blood oxygen, height, weight, body fat, facial / tongue / ear / hand blemishes, blood vessels, texture, color, sclera / iris / fundus of the eyes, spots, blood vessels, color, shape, skin scars and other physiological and pathological features. The extracted features are compared using a knowledge graph of a big data expert system that pre-defines human physiological and pathological characteristics. Based on the similarity of feature comparison results, AI analysis provides diagnostic and early warning results, enabling the prediction and early warning of human health risks and disease diagnosis, and providing suggestions on medication, diet therapy, and physical therapy.

[0054] Furthermore, the AI-based human multi-physiological indicator health analysis and risk prediction and early warning subsystem 4 can be deployed on processors of different specifications, such as cloud servers, local servers, single computers, and mobile terminals, to perform storage management and AI analysis of human physiological indicators and health information, for example... Figure 5a As shown, based on the infiltration state of gray haze and yellow halo at the iris edge, as well as changes in iris microstructure, risk prediction and early warning of Alzheimer's disease can be performed; for example... Figures 5b-5d As shown, based on pulse waveform characteristics, the distribution and color characteristics of spots on the iris, and the distribution, color characteristics, and morphological characteristics of spots and blood vessels on the sclera, risk prediction and early warning for cardiovascular and cerebrovascular diseases, diabetes, tumors or cancers, polycystic ovary syndrome, etc., can be performed. Figure 5a In (1) healthy individuals, the iris microstructure and iris edge contour are clear, while in (2) individuals at high risk of Alzheimer's disease, the iris microstructure becomes blurred and unclear, the iris edge contour is severely hazy, and ring-shaped yellow or gray erosion circles appear. Normally, the P1, P2, and P3 layers of the pulse wave are clear in healthy individuals, while in (2) individuals at high risk of Alzheimer's disease, the iris microstructure becomes blurred and unclear, the iris edge contour is severely hazy, and ring-shaped yellow or gray erosion circles appear. Figure 5b In high-risk individuals for cardiovascular disease, those with aortic insufficiency exhibiting a secondary peak alongside the P1 peak in their pulse wave; ② those with coronary artery sclerosis showing a complete drop in the P2 peak of their pulse wave; ③ those with depression or emotional instability exhibiting severely erratic and jerky pulse waveforms; ④ those with arrhythmias showing excessively short intervals between two cardiac contractions in their pulse wave, with the second contraction being incomplete; ⑤ those with palpitations, chest tightness, and heart palpitations showing a decrease in the P2 peak and the disappearance of the P3 peak in their pulse wave. Normally, the sclera of healthy individuals is clean and clear, while... Figure 5c In individuals at high risk of diabetes, yellow haze or yellow spots may appear in the sclera. As diabetes worsens, the area affected by the yellow haze or spots increases in size and intensity. In contrast, healthy individuals typically have clear, uniformly sized, and uniformly colored blood vessels in the sclera. Figure 5dIn individuals at high risk of tumors or cancer, the blood vessels may swell, exhibit abnormal color, and show signs such as blood vessels appearing as if they are fogged up, hanging like beads, or leaking.

[0055] In summary, this invention achieves multi-parameter non-intrusive measurement of human visual diagnosis and sensing through a high-resolution variable-magnification dual-focal-plane imaging optical detection subsystem 1, a human sensory area automatic tracking clear magnification imaging and positioning segmentation acquisition subsystem 2, a wearable and / or external sensing subsystem 3, and an AI-based human multi-physiological indicator health analysis and risk prediction and early warning subsystem 4. It adopts an optical detection system structure that maintains sufficient working distance to achieve visual discomfort-free and non-contact measurement, and uses wearable and / or placement within the human activity range for sensing detection. This enables non-pressure-free perception and rapid dynamic measurement of multiple physiological indicators through multi-modal visual diagnosis, without unintentional interference with daily life and work.

[0056] The various embodiments in this specification are described in a progressive manner. Similar or identical parts between embodiments can be referred to interchangeably. Each embodiment focuses on describing the differences from other embodiments. In the description of this specification, the terms "a preferred embodiment," "furthermore," "specifically," "in this embodiment," etc., refer to specific features, structures, materials, or characteristics described in connection with that embodiment or example, which are included in at least one embodiment or example of the embodiments in this specification. In this specification, the 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 can be combined in any suitable manner in one or more embodiments or examples. Moreover, without contradiction, those skilled in the art can combine and integrate the different embodiments or examples described in this specification, as well as the features of different embodiments or examples.

[0057] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A high-resolution intelligent non-intrusive multi-mode detection and human health monitoring device, characterized in that, This includes a high-resolution variable-magnification dual-focal-plane imaging optical detection subsystem, a human sensory area automatic tracking clear magnification imaging and positioning segmentation acquisition subsystem, a wearable and / or external sensing subsystem, and an AI-based human multi-physiological indicator health analysis and risk prediction and early warning subsystem, among which: The high-resolution variable-magnification dual-focal-plane imaging optical detection subsystem is used to obtain high-resolution detection images of different sensory areas of the human body, and realize non-invasive and rapid measurement of multi-modal diagnostic information of the human body. The automatic tracking, clear magnification imaging, positioning, and segmentation acquisition subsystem for human body regions of interest is used to identify regions of interest or targets, and sequentially perform automatic tracking and positioning of regions of interest or targets, automatic focusing and clear imaging, and segmentation of regions of interest or targets to achieve high-resolution clear imaging of regions of interest or targets. The wearable and / or external sensing subsystem is worn on important parts of the body and / or in external contact with important parts of the body to acquire physiological parameters of the human body. The AI-based human multi-physiological indicator health analysis and risk prediction and early warning subsystem uses large-scale models, big data and AI analysis based on human multi-modal diagnostic information and human physiological parameter indicators to achieve prediction and early warning of human health risks and disease diagnosis.

2. The high-resolution intelligent non-contact multi-mode detection and human health monitoring device according to claim 1, characterized in that, The high-resolution variable-magnification dual-focal-plane imaging optical detection subsystem includes a front lens group, an intermediate lens group, a rear lens group, an automatic zoom control system, and a CCD detector. The front lens group is positioned near the focal plane for human detection. The intermediate lens group moves back and forth under the control of the automatic zoom control system to adjust the magnification. The CCD detector's sensing surface is located at the focal plane of the rear lens group to acquire human imaging signals. Different detection areas of the human body share a single optical detection subsystem, obtaining high-resolution detection images of different sensory regions of the human body, enabling non-invasive, rapid, and comprehensive measurement of multi-modal diagnostic information.

3. The high-resolution intelligent non-contact multi-mode detection and human health monitoring device according to claim 2, characterized in that, The high-resolution variable-magnification dual-focal-plane imaging optical detection subsystem is also equipped with a position sensor and a microprocessor AI analysis system. The position sensor is used to measure the distance between the human body and the CCD detector and transmit the distance value to the microprocessor AI analysis system. The microprocessor AI analysis system controls the imaging lens set in front of the CCD detector to complete the corresponding adjustment control based on the distance value, so that the human body can finally obtain a clear image on the CCD detector.

4. The high-resolution intelligent non-contact multi-mode detection and human health monitoring device according to claim 3, characterized in that, A light source is placed between the front lens and the human body. By designing the shape, size and number of the light source, it is used to achieve planar shadowless illumination, three-dimensional shadowless illumination or four-way independent controllable backlight shadowless illumination for the human body.

5. The high-resolution intelligent non-contact multi-mode detection and human health monitoring device according to claim 1, characterized in that, The identification of regions of interest or targets is achieved through the YOLOv5 small target identification and analysis algorithm, and the optimal sharpness adjustment is automatically performed by searching for the maximum grayscale variance based on the hill climbing algorithm, so as to achieve high-resolution clear imaging only for regions of interest or targets.

6. The high-resolution intelligent non-contact multi-mode detection and human health monitoring device according to claim 1, characterized in that, The wearable and / or external sensing subsystem is a wearable and / or sensor placed within the range of human activity. The sensor is used to perform rapid dynamic measurement of physiological parameters in the human sensory area through wearable contact or human stepping, without being unintentional or interfering with daily life and work.

7. The high-resolution intelligent non-contact multi-mode detection and human health monitoring device according to claim 6, characterized in that, The sensors, worn on vital parts of the body, acquire physiological indicators of regions of interest, including one or more of pressure sensors, photoelectric sensors, gas sensors, temperature and humidity sensors, inertial attitude sensors, and GPS position sensors, for acquiring signals of human physiological indicators, body posture, and location; or / and, The sensors are installed in the human activity area to acquire physiological indicators of the human body's region of interest. They include one or more of the following: pressure sensor, sound sensor, ultrasonic or laser rangefinder, infrared sensor, and infrared CCD. When human body pressure, sound, distance, or temperature is detected, the sensors automatically trigger the acquisition of human physiological indicator signals, body posture, and location signals.

8. The high-resolution intelligent non-contact multi-mode detection and human health monitoring device according to claim 7, characterized in that, When the sensor is a pressure sensor or a photoelectric sensor, by using only one of the pressure sensor or photoelectric sensor and establishing a matching theoretical model and AI algorithm analysis method, it can simultaneously measure and provide multiple physiological indicators of the human body: pulse, blood pressure, heart rate, electrocardiogram related information, respiration and blood oxygen; it can also be combined with the gas sensor, temperature and humidity sensor, inertial attitude sensor and GPS position sensor to provide monitoring of disabled persons' turning over in bed, falls, urinary and fecal incontinence, and locate their current location information.

9. The high-resolution intelligent non-contact multi-mode detection and human health monitoring device according to claim 1, characterized in that, Based on multi-modal visual diagnosis information and physiological parameters of the human body, large-scale models, big data, and AI analysis are used to predict and warn of human health risks and diagnose diseases, including: Deep learning algorithms are used to extract features from multimodal visual diagnosis information and physiological parameters of the human body. The extracted features are compared using a knowledge graph of a big data expert system that pre-defines human physiological and pathological characteristics. Based on the similarity of features, AI analysis provides diagnostic and early warning results, enabling the prediction and early warning of human health risks and disease diagnosis, and providing suggestions on medication, diet therapy, and physical therapy.

10. The high-resolution intelligent non-contact multi-mode detection and human health monitoring device according to claim 9, characterized in that, The AI-based human multi-physiological indicator health analysis and risk prediction and early warning subsystem can be deployed on cloud servers, local servers, single computers or mobile terminals.

Images