Intelligent medical health data identification system

Abstract

The invention relates to an intelligent medical health data identification system. The system comprises an information capture assembly, a content capture assembly, a data analysis assembly, an anemia identification mechanism and a state display mechanism. The device is reasonable in structural design and can be massively popularized in the medical field.

Description

Technical Field

[0001] The present invention relates to the field of smart medical care, and in particular to a smart medical health data identification system. Background Art

[0002] In the healthcare sector, good living habits are crucial to good health. People should get enough sleep and avoid staying up late; quit smoking and limit alcohol consumption to reduce adverse effects on the body; maintain personal hygiene, wash hands frequently, and ventilate the room to prevent bacterial growth; maintain an optimistic attitude and proactively face the pressures and challenges of life. Regular physical examinations are also an important tool for disease prevention. Through physical examinations, people can promptly identify any abnormalities and take appropriate treatment measures. Generally, a comprehensive physical examination is recommended at least once a year, including checks of blood pressure, blood sugar, blood lipids, and an electrocardiogram.

[0003] Anemia testing is crucial in physical examinations, as many diseases and symptoms are associated with anemia. However, anemia testing is typically performed through a destructive method, requiring a needle to draw blood. This method is not only harmful but also inconvenient and slow to produce results. A non-destructive, efficient anemia detection mechanism is needed that can provide at least a preliminary diagnosis of anemia, thereby improving the speed and efficiency of physical examinations. Summary of the Invention

[0004] In order to solve technical problems in related fields, the present invention provides an intelligent medical health data identification system, which can use the total number of anemia days of the current patient in the previous year, the various inherent parameters of the current patient, and the cyan component value, magenta component value, yellow component value and black component value of each pixel point in the image block where the skin is located in the real-time facial pattern of the current patient as input data of the feedforward neural network model one by one, to execute the feedforward neural network model, and obtain the output of the feedforward neural network model to indicate whether the current patient is currently in an anemic state, thereby completing the intelligent preliminary identification of the anemia state based on the various basic information of the current patient and the visual information of the current patient's facial skin, and providing a preliminary screening mechanism for blood test.

[0005] According to the present invention, a smart medical health data identification system is provided, the system comprising: An information acquisition component is used to obtain various inherent parameters of the current patient, wherein the various inherent parameters of the current patient are the current patient's residence altitude, age value, gender marker, and years of anemia; a content capture component for capturing a frontal face image of the current patient in real time to obtain a live captured image, detecting various facial imaging areas in the live captured image based on human facial imaging features, and outputting the largest facial imaging area among the facial imaging areas as the real-time facial pattern of the current patient; a data parsing component connected to the content capture component, configured to parse the image block where the skin is located in the current patient's real-time facial pattern based on skin imaging features, and obtain a cyan component value, a magenta component value, a yellow component value, and a black component value for each pixel point in the image block where the skin is located in the current patient's real-time facial pattern; an anemia identification mechanism, connected to the information capture component, the content capture component, and the data parsing component, respectively, for using the total number of days of anemia in the previous year for the current patient, various inherent parameters of the current patient, and the cyan component value, magenta component value, yellow component value, and black component value of each pixel point in the image block where the skin is located in the real-time facial pattern of the current patient as input data for a feedforward neural network model, so as to execute the feedforward neural network model and obtain an indication of whether the current patient is currently in an anemic state as output by the feedforward neural network model; a status display mechanism connected to the anemia identification mechanism, configured to receive and display in real time anemia status information corresponding to an identification indicating whether the patient is currently in an anemia state; The total number of days of anemia in the previous year of the current patient, various inherent parameters of the current patient, and the cyan component value, magenta component value, yellow component value, and black component value of each pixel point of the image block where the skin is located in the real-time facial pattern of the current patient are used as input data of the feedforward neural network model one by one, so as to execute the feedforward neural network model, and obtain an identification of whether the current patient is currently in an anemic state output by the feedforward neural network model, including: the feedforward neural network model is a feedforward neural network that has undergone multiple learning actions, and the number of feedforward neural network learning actions is inversely correlated with the number of years of anemia; Among them, the content capture component is used to perform real-time image capture of the front face of the current patient to obtain a live capture image, detect each facial imaging area in the live capture image based on the human facial imaging features, and output the facial imaging area with the largest area among the various facial imaging areas as the real-time facial pattern of the current patient, including: the human facial imaging features are the facial patterns of each benchmark population corresponding to various populations.

[0006] It can be seen that the present invention has at least the following main inventive concepts: Inventive concept A: The total number of days of anemia in the previous year for the current patient, various inherent parameters of the current patient, and the cyan component value, magenta component value, yellow component value, and black component value of each pixel point in the image block where the skin of the current patient's real-time facial pattern is located are used as input data for a feedforward neural network model one by one, so as to execute the feedforward neural network model and obtain an output from the feedforward neural network model indicating whether the current patient is currently in an anemic state, thereby completing intelligent preliminary identification of the anemia state based on various basic information of the current patient and visual information of the current patient's facial skin, and providing a preliminary screening mechanism for blood test. Inventive Concept B: The feedforward neural network model employed is a feedforward neural network that has undergone multiple learning actions, and the number of learning actions of the feedforward neural network is inversely correlated with the duration of anemia, thereby satisfying the characteristic that the shorter the duration of anemia, the more difficult it is to initially identify anemia. Inventive concept C: Real-time image capture is performed on the front face of the current patient to obtain a live capture image, each facial imaging area in the live capture image is detected based on human facial imaging features, the facial imaging area with the largest area among the various facial imaging areas is used as the real-time facial pattern of the current patient, the image block where the skin is located in the real-time facial pattern of the current patient is analyzed based on skin imaging features, and the cyan component value, magenta component value, yellow component value and black component value of each pixel point of the image block where the skin is located in the real-time facial pattern of the current patient are obtained, thereby completing the targeted analysis of the facial skin visualization information of the current patient. DETAILED DESCRIPTION

[0007] The following is a detailed description of an embodiment of the smart medical health data identification system of the present invention.

[0008] Example 1

[0009] The smart medical and health data identification system shown in Example 1 of the present invention includes: An information acquisition component is used to obtain various inherent parameters of the current patient, wherein the various inherent parameters of the current patient are the current patient's residence altitude, age value, gender marker, and years of anemia; Specifically, the information capture component is used to obtain various inherent parameters of the current patient, wherein the various inherent parameters of the current patient are the current patient's residence altitude, age value, gender marker, and anemia suffering years. The information capture component includes a plurality of information capture units, which are used to respectively capture the current patient's residence altitude, age value, gender marker, and anemia suffering years. a content capture component for capturing a frontal face image of the current patient in real time to obtain a live captured image, detecting various facial imaging areas in the live captured image based on human facial imaging features, and outputting the largest facial imaging area among the facial imaging areas as the real-time facial pattern of the current patient; a data parsing component connected to the content capture component, configured to parse the image block where the skin is located in the current patient's real-time facial pattern based on skin imaging features, and obtain a cyan component value, a magenta component value, a yellow component value, and a black component value for each pixel point in the image block where the skin is located in the current patient's real-time facial pattern; an anemia identification mechanism, connected to the information capture component, the content capture component, and the data parsing component, respectively, for using the total number of days of anemia in the previous year for the current patient, various inherent parameters of the current patient, and the cyan component value, magenta component value, yellow component value, and black component value of each pixel point in the image block where the skin is located in the real-time facial pattern of the current patient as input data for a feedforward neural network model, so as to execute the feedforward neural network model and obtain an indication of whether the current patient is currently in an anemic state as output by the feedforward neural network model; a status display mechanism connected to the anemia identification mechanism, configured to receive and display in real time anemia status information corresponding to an identification indicating whether the patient is currently in an anemia state; The total number of days of anemia in the previous year of the current patient, various inherent parameters of the current patient, and the cyan component value, magenta component value, yellow component value, and black component value of each pixel point of the image block where the skin is located in the real-time facial pattern of the current patient are used as input data of the feedforward neural network model one by one, so as to execute the feedforward neural network model, and obtain an identification of whether the current patient is currently in an anemic state output by the feedforward neural network model, including: the feedforward neural network model is a feedforward neural network that has undergone multiple learning actions, and the number of feedforward neural network learning actions is inversely correlated with the number of years of anemia; The content capture component is configured to capture a frontal face image of the current patient in real time to obtain a live captured image, detect various facial imaging areas in the live captured image based on human facial imaging features, and output the largest facial imaging area among the facial imaging areas as the real-time facial pattern of the current patient, wherein the human facial imaging features are facial patterns of various reference groups corresponding to various groups of people; wherein, parsing an image block where the skin is located in the real-time facial pattern of the current patient based on skin imaging features, and obtaining a cyan component value, a magenta component value, a yellow component value, and a black component value of each pixel point in the image block where the skin is located in the real-time facial pattern of the current patient, includes: the cyan component value, the magenta component value, the yellow component value, and the black component value of each pixel point are the C component value, the M component value, the Y component value, and the K component value of each pixel point in a CMYK color space; wherein the total number of days of anemia of the current patient in the previous year, various inherent parameters of the current patient, and the cyan component value, magenta component value, yellow component value, and black component value of each pixel point of the image block where the skin of the current patient's real-time facial pattern is located are used as input data of the feedforward neural network model one by one to execute the feedforward neural network model, and obtaining an indication of whether the current patient is currently in an anemic state output by the feedforward neural network model further includes: when the indication of whether the current patient is currently in an anemic state output by the feedforward neural network model is 0B11, it indicates that the current patient is currently in an anemic state; And wherein, the total number of days of anemia of the current patient in the previous year, the various inherent parameters of the current patient, and the cyan component value, magenta component value, yellow component value and black component value of each pixel point of the image block where the skin is located in the real-time facial pattern of the current patient are used as input data of the feedforward neural network model item by item to execute the feedforward neural network model, and obtain the identification of whether the current patient is currently in an anemic state output by the feedforward neural network model, including: when the identification of whether the current patient is currently in an anemic state output by the feedforward neural network model is 0B10, it indicates that the current patient is currently in a non-anemic state.

[0010] Example 2

[0011] The smart medical and health data identification system according to embodiment 2 of the present invention may further include: a fault self-detection device for providing a fault self-detection service for the step of using the total number of days of anemia of the current patient in the previous year, various inherent parameters of the current patient, and the cyan component value, magenta component value, yellow component value, and black component value of each pixel point of an image block where the skin of the current patient's real-time facial pattern is located as input data to a feedforward neural network model one by one, executing the feedforward neural network model, and obtaining an indication of whether the current patient is currently in an anemic state as output by the feedforward neural network model; The fault self-detection device is configured to provide a fault self-detection service for the steps of using the total number of anemic days of the current patient in the previous year, various inherent parameters of the current patient, and the cyan component value, magenta component value, yellow component value, and black component value of each pixel point of the image block where the skin is located in the real-time facial pattern of the current patient as input data to the feedforward neural network model one by one, so as to execute the feedforward neural network model and obtain an identification of whether the current patient is currently in an anemic state output by the feedforward neural network model. The fault self-detection device includes a plurality of fault self-detection units; The fault self-detection device is used to provide a fault self-detection service for the step of using the total number of anemic days of the current patient in the previous year, various inherent parameters of the current patient, and the cyan component value, magenta component value, yellow component value, and black component value of each pixel point of the image block where the skin is located in the real-time facial pattern of the current patient as input data of the feedforward neural network model one by one, so as to execute the feedforward neural network model and obtain an identification of whether the current patient is currently in an anemic state output by the feedforward neural network model. The step also includes: the multiple fault self-detection units respectively provide fault self-detection services for multiple processes running synchronously at the same time; The fault self-detection device is configured to provide a fault self-detection service for the step of using the total number of anemic days of the current patient in the previous year, various inherent parameters of the current patient, and the cyan component value, magenta component value, yellow component value, and black component value of each pixel point of the image block where the skin is located in the real-time facial pattern of the current patient as input data to the feedforward neural network model one by one, so as to execute the feedforward neural network model and obtain an identification of whether the current patient is currently in an anemic state output by the feedforward neural network model. The step also includes: the multiple fault self-detection units are respectively implemented using different logic devices; Among them, the fault self-detection device is used to use the total number of anemia days of the current patient in the previous year, the various inherent parameters of the current patient, the cyan component value, magenta component value, yellow component value and black component value of each pixel point of the image block where the skin is located in the real-time facial pattern of the current patient as input data of the feedforward neural network model one by one, so as to execute the feedforward neural network model, and obtain the output of the feedforward neural network model to indicate whether the current patient is currently in an anemic state. The step of providing a fault self-detection service also includes: using an ASIC chip to implement the fault self-detection device.

[0012] Example 3

[0013] The smart medical and health data identification system shown in Example 3 of the present invention may further include: a throughput identification mechanism, configured to provide an identification operation of data throughput per unit time for the step of using the total number of days of anemia of the current patient in the previous year, various inherent parameters of the current patient, and the cyan component value, magenta component value, yellow component value, and black component value of each pixel point of the image block where the skin of the current patient's real-time facial pattern is located as input data to the feedforward neural network model one by one, executing the feedforward neural network model, and obtaining an identification of whether the current patient is currently in an anemic state as output by the feedforward neural network model; And wherein, the throughput identification mechanism is used to provide the data throughput per unit time for the step of using the total number of anemia days of the current patient in the previous year, the various inherent parameters of the current patient, and the cyan component value, magenta component value, yellow component value and black component value of each pixel point of the image block where the skin of the current patient's real-time facial pattern is located as input data of the feedforward neural network model one by one, so as to execute the feedforward neural network model and obtain the identification of whether the current patient is currently in an anemic state output by the feedforward neural network model. The identification operation includes: using the total number of anemia days of the current patient in the previous year, the various inherent parameters of the current patient, and the cyan component value, magenta component value, yellow component value and black component value of each pixel point of the image block where the skin of the current patient's real-time facial pattern is located as input data of the feedforward neural network model one by one The data throughput per unit time of the step of using the cyan component value and the black component value as the input data of the feedforward neural network model item by item to execute the feedforward neural network model and obtain the identification of whether the current patient is currently in an anemic state output by the feedforward neural network model is the total number of anemia days of the current patient in the previous year, the various inherent parameters of the current patient, and the cyan component value, magenta component value, yellow component value and black component value of each pixel point of the image block where the skin is located in the real-time facial pattern of the current patient as the input data of the feedforward neural network model item by item to execute the feedforward neural network model and obtain the identification of whether the current patient is currently in an anemic state output by the feedforward neural network model.

[0014] In addition, in the smart medical health data identification system, the feedforward neural network model is a feedforward neural network after multiple learning actions and the number of feedforward neural network learning actions is inversely correlated with the years of anemia, including: using an information conversion function to represent the information conversion relationship in which the number of feedforward neural network learning actions is inversely correlated with the years of anemia, in which the years of anemia are input data and the number of feedforward neural network learning actions corresponding to the years of anemia are output data.

[0015] The intelligent medical health data identification system of the present invention addresses the technical problem of inefficient anemia detection in the existing technology. By adopting a feedforward neural network model with a customized structure, it intelligently determines whether the current patient is in an anemic state based on multiple related data of the current patient, and completes the intelligent preliminary identification of the anemia state based on the basic information of the current patient and the visual information of the current patient's facial skin, thereby improving the speed and efficiency of physical examinations and solving the above technical problems.

[0016] As many apparently widely different embodiments of the present invention can be made without departing from the spirit and scope thereof, it is to be understood that the invention is not limited to the specific embodiments except as defined in the appended claims.

Claims

1. A smart medical health data recognition system, characterized by: The system includes: An information capture component is used to obtain various inherent parameters of the current patient, wherein the various inherent parameters of the current patient are the current patient's residence altitude, age value, gender marker, and years of anemia; A content capture component is used to capture the frontal face of the current patient in real time to obtain a live captured image, detect various facial imaging areas in the live captured image based on human facial imaging features, and output the largest facial imaging area among the various facial imaging areas as the real-time facial pattern of the current patient, including: facial patterns of various benchmark populations corresponding to the human facial imaging features; a data parsing component connected to the content capture component, for parsing the image block where the skin is located in the current patient's real-time facial pattern based on skin imaging features, and obtaining the cyan component value, magenta component value, yellow component value, and black component value of each pixel point in the image block where the skin is located in the current patient's real-time facial pattern; an anemia identification mechanism, connected to the information capture component, the content capture component, and the data analysis component, respectively, and configured to use the total number of days of anemia in the previous year for the current patient, various inherent parameters of the current patient, and the cyan component value, magenta component value, yellow component value, and black component value of each pixel point in the image block where the skin is located in the real-time facial pattern of the current patient as input data for a feedforward neural network model one by one, so as to execute the feedforward neural network model and obtain an identification of whether the current patient is currently in an anemic state output by the feedforward neural network model, including: the feedforward neural network model is a feedforward neural network that has undergone multiple learning actions, and the number of feedforward neural network learning actions is inversely correlated with the number of years of anemia; The status display mechanism is connected to the anemia identification mechanism and is used to receive and display in real time the anemia status information corresponding to the identification of whether the current patient is in an anemia state.

2. The smart medical health data recognition system according to claim 1, wherein: parsing an image block where the skin of the current patient's real-time facial pattern is located based on skin imaging features, and obtaining a cyan component value, a magenta component value, a yellow component value, and a black component value of each pixel point in the image block where the skin of the current patient's real-time facial pattern is located, including: the cyan component value, the magenta component value, the yellow component value, and the black component value of each pixel point are the C component value, the M component value, the Y component value, and the K component value of each pixel point in a CMYK color space; wherein the total number of days of anemia of the current patient in the previous year, various inherent parameters of the current patient, and the cyan component value, magenta component value, yellow component value, and black component value of each pixel point of the image block where the skin of the current patient's real-time facial pattern is located are used as input data of the feedforward neural network model one by one to execute the feedforward neural network model, and obtaining an indication of whether the current patient is currently in an anemic state output by the feedforward neural network model further includes: when the indication of whether the current patient is currently in an anemic state output by the feedforward neural network model is 0B11, it indicates that the current patient is currently in an anemic state; Among them, the total number of days of anemia of the current patient in the previous year, the various inherent parameters of the current patient, and the cyan component value, magenta component value, yellow component value and black component value of each pixel point of the image block where the skin is located in the real-time facial pattern of the current patient are used as input data of the feedforward neural network model item by item to execute the feedforward neural network model, and obtain the output of the feedforward neural network model to indicate whether the current patient is currently in an anemic state, including: when the output of the feedforward neural network model to indicate whether the current patient is currently in an anemic state is 0B10, it indicates that the current patient is currently in a non-anemic state.

3. The smart medical health data recognition system according to claim 2, characterized in that: The system includes: A fault self-detection device is used to provide a fault self-detection service for the steps of using the total number of days of anemia of the current patient in the previous year, the various inherent parameters of the current patient, and the cyan component value, magenta component value, yellow component value, and black component value of each pixel point of the image block where the skin is located in the real-time facial pattern of the current patient as input data of the feedforward neural network model one by one, so as to execute the feedforward neural network model and obtain an identification of whether the current patient is currently in an anemic state output by the feedforward neural network model.

4. The smart medical health data recognition system according to claim 3, wherein: A fault self-detection device is used to provide a fault self-detection service for the steps of using the total number of anemia days of the current patient in the previous year, the various inherent parameters of the current patient, and the cyan component value, magenta component value, yellow component value, and black component value of each pixel point of the image block where the skin is located in the real-time facial pattern of the current patient as input data of the feedforward neural network model one by one, so as to execute the feedforward neural network model and obtain an identification of whether the current patient is currently in an anemic state output by the feedforward neural network model. The fault self-detection device includes multiple fault self-detection units.

5. The smart medical health data recognition system according to claim 4, wherein: A fault self-detection device is used to provide a fault self-detection service for the step of using the total number of days of anemia of the current patient in the previous year, the various inherent parameters of the current patient, and the cyan component value, magenta component value, yellow component value, and black component value of each pixel point of the image block where the skin is located in the real-time facial pattern of the current patient as input data of the feedforward neural network model one by one, so as to execute the feedforward neural network model and obtain an identification of whether the current patient is currently in an anemic state output by the feedforward neural network model. The step also includes: multiple fault self-detection units provide fault self-detection services for multiple processes running synchronously at the same time.

6. The smart medical health data recognition system according to claim 5, wherein: A fault self-detection device is used to provide a fault self-detection service for the step of using the total number of days of anemia of the current patient in the previous year, the various inherent parameters of the current patient, and the cyan component value, magenta component value, yellow component value, and black component value of each pixel point of the image block where the skin is located in the real-time facial pattern of the current patient as input data of the feedforward neural network model one by one, so as to execute the feedforward neural network model and obtain an identification of whether the current patient is currently in an anemic state output by the feedforward neural network model. The step also includes: multiple fault self-detection units are implemented using different logic devices.

7. The smart medical health data recognition system according to claim 5, wherein: A fault self-detection device is used to provide a fault self-detection service for the steps of using the total number of days of anemia of the current patient in the previous year, the various inherent parameters of the current patient, and the cyan component value, magenta component value, yellow component value, and black component value of each pixel point of the image block where the skin is located in the real-time facial pattern of the current patient as input data for a feedforward neural network model one by one, so as to execute the feedforward neural network model and obtain an identification of whether the current patient is currently in an anemic state output by the feedforward neural network model. The output also includes: using an ASIC chip to implement the fault self-detection device.

8. The smart medical health data recognition system according to claim 2, wherein: The system also includes: The throughput identification mechanism is used to provide an identification operation of data throughput per unit time for the step of using the total number of days of anemia of the current patient in the previous year, various inherent parameters of the current patient, and the cyan component value, magenta component value, yellow component value, and black component value of each pixel point of the image block where the skin is located in the real-time facial pattern of the current patient as input data of the feedforward neural network model one by one, so as to execute the feedforward neural network model and obtain an identification of whether the current patient is currently in an anemic state output by the feedforward neural network model.

9. The smart medical health data recognition system according to claim 8, wherein: The throughput recognition mechanism is used to provide a data throughput per unit time for the step of using the total number of anemia days of the current patient in the previous year, the various inherent parameters of the current patient, and the cyan component value, magenta component value, yellow component value, and black component value of each pixel point of the image block where the skin of the current patient's real-time facial pattern is located as input data of the feedforward neural network model one by one, so as to execute the feedforward neural network model and obtain the identification of whether the current patient is currently in an anemic state output by the feedforward neural network model. The identification operation includes: using the total number of anemia days of the current patient in the previous year, the various inherent parameters of the current patient, and the cyan component value, magenta component value, yellow component value, and black component value of each pixel point of the image block where the skin of the current patient's real-time facial pattern is located as input data of the feedforward neural network model one by one The black component value is used as the input data of the feedforward neural network model item by item to execute the feedforward neural network model, and the data throughput per unit time of the step of obtaining the output of the feedforward neural network model to identify whether the current patient is currently in an anemic state is the total number of days of anemia of the current patient in the previous year, the various inherent parameters of the current patient, and the cyan component value, magenta component value, yellow component value and black component value of each pixel point of the image block where the skin is located in the real-time facial pattern of the current patient as the input data of the feedforward neural network model item by item to execute the feedforward neural network model, and obtain the output of the feedforward neural network model to identify whether the current patient is currently in an anemic state is the sum of the amount of data received and the amount of data sent per unit time.