Medical health behavior data collection system and method

Abstract

The invention discloses a medical health behavior data collection system and method, and the system comprises a video data collection and preprocessing module which is used for collecting video data and carrying out the preprocessing of the video data; the key node extracting and marking module is used for extracting and marking key data nodes; the data transmission and storage module is used for transmitting and storing the data; and the data processing and application module is used for processing and applying the data. According to the method, a deep learning mode recognition algorithm is combined, frame-by-frame key points are integrated and analyzed, various postures of behavior subjects are accurately recognized, corresponding statistical information is generated, and rich and accurate data support is provided for medical diagnosis and treatment and health management decisions.

Description

Technical Field

[0001] The present invention belongs to the field of medical health technology, and specifically relates to a medical health behavior data collection system and method. Background Art

[0002] In recent years, driven by multiple factors, the medical and health big data industry has shown unprecedented development vitality and broad prospects. Core factors such as scientific and technological progress, policy support, surging market demand and demographic transformation have jointly shaped the current booming status of the industry.

[0003] In the modern medical decision-making system, the depth and breadth of information data have become an indispensable cornerstone. This trend has directly promoted the vigorous rise of the medical and health big data industry. The deepening of medical and health concepts is essentially the advancement of the boundaries of medical services, that is, the shift from the traditional disease treatment model to the preventive medical model with health management as the core. This transformation requires the seamless connection of in-hospital diagnosis and treatment information with out-of-hospital life data to build a comprehensive and continuous individual health portrait.

[0004] What is particularly critical is that the data of life at home after leaving the hospital, especially the patient behavior data, has become a bridge connecting health management inside and outside the hospital. As a direct reflection of individual health status and living habits, patient behavior data not only covers the activity trajectory and living habits in daily life outside the hospital, but also involves information such as interactive behaviors and treatment responses during the diagnosis and treatment process in the hospital. These data are crucial for the formulation of precision medicine and personalized health management plans.

[0005] However, the realistic recording of behavioral data faces many challenges. Although video recordings can intuitively show patient behavior, their huge data volume and high privacy sensitivity limit their application in long-term, large-scale health monitoring. In contrast, although text records, voice calls and other data are easy to obtain, it is often difficult to directly and comprehensively capture behavioral details. Effective analysis requires complex algorithm conversion and multi-dimensional information fusion, which places higher demands on technical implementation and data processing efficiency.

[0006] Therefore, in the field of medical and health big data, there is an urgent need to develop more efficient, accurate and less invasive data collection and analysis technologies, such as physiological parameter monitoring based on wearable devices and the application of environmental perception sensor networks. These technologies can capture patient behavior data more directly and continuously while ensuring the security and privacy of the data, providing solid data support for modern medical decision-making. Summary of the invention

[0007] In view of the above problems, the present invention is proposed to provide a medical health behavior data collection system and method that overcomes the above problems or at least partially solves the above problems.

[0008] To achieve the above object, the present invention adopts the following technical solutions:

[0009] A medical health behavior data collection system, the system comprising:

[0010] Video data acquisition and preprocessing module, used for acquiring video data and preprocessing the video data;

[0011] Key node extraction and marking module, used to extract and mark key data nodes;

[0012] Data transmission and storage module, used for transmitting and storing data;

[0013] Data processing and application module, used to process and apply data.

[0014] Optionally, the video data acquisition and preprocessing module includes:

[0015] A raw video capture unit, which uses a camera to capture video data in real time;

[0016] Edge computing and image analysis unit, used to analyze the collected video data.

[0017] Optionally, the edge computing and image analysis unit includes:

[0018] Transmit the captured video data to the chip built into the sensor;

[0019] Processing video data using chips;

[0020] Identify the human body area in the video data and calculate and analyze the data image matrix of the area.

[0021] Optionally, the key node extraction and marking module includes:

[0022] Generate key node data with clear location information;

[0023] Mark the information of each key node to ensure the structuring and traceability of data;

[0024] Traverse and mark the key point information of each frame image in the real-time video stream.

[0025] Optionally, the data transmission and storage module includes:

[0026] Transmit key point data information through network protocol;

[0027] The cloud server or dedicated application software database receives and stores data information in real time.

[0028] Optional data processing and application modules include:

[0029] The data receiving and distributing unit is used to receive data from sensors, store them locally and then push them;

[0030] A video simulation and display unit, used for video simulation and display of data;

[0031] The deep learning pattern recognition and statistics unit is used to integrate and analyze data and perform statistics.

[0032] Optionally, the video simulation and display unit includes:

[0033] Receive the pushed data, and use the key point video simulation algorithm to restore the discrete key point data frame by frame into a simulated compressed video;

[0034] Output analog video for external display, retaining key information of the original video and reducing storage and transmission burden.

[0035] Optionally, the deep learning pattern recognition and statistics unit includes:

[0036] After receiving the data, the key points of each frame are integrated and analyzed in combination with the medical and health knowledge graph;

[0037] Specific human postures are identified, and the simulated video is calculated frame by frame based on the processed data labels to generate statistical information for each posture.

[0038] The present invention also provides a method for collecting medical health behavior data, which is based on any of the above-mentioned systems and comprises the following steps:

[0039] S1, collect video data through the camera;

[0040] S2, performing edge computing and image analysis on the captured video data;

[0041] S3, generating key point data of human body posture;

[0042] S4, storing and transmitting the generated human body posture key point data;

[0043] S5, receiving data and distributing it;

[0044] S6, converting the key point data into a simulation video and displaying it;

[0045] S7. Calculate the simulated video frame by frame to generate statistical information of various human postures.

[0046] In summary, due to the adoption of the above technical solution, the beneficial effects of the present invention are:

[0047] 1. The present invention realizes long-term, continuous, desensitized and compressed video acquisition and output in multiple scenarios for individuals and families; through intelligent processing of real-time video data, the video content is converted into key point data representing human posture. This process not only effectively reduces the sensitivity of the data, but also can greatly reduce data storage requirements through efficient compression algorithms.

[0048] 2. The present invention uses a key point video simulation algorithm to restore discrete key point data to a simulated video, which not only retains the key information of the original video, but also realizes the desensitization of the data. At the same time, combined with a deep learning pattern recognition algorithm, the key points are integrated and analyzed frame by frame, and various postures of the subject, such as walking, sitting, falling, etc., are accurately identified, and corresponding statistical information is generated, providing rich and accurate data support for medical diagnosis and health management decisions.

[0049] 3. The present invention not only solves many pain points of traditional video acquisition methods in the application of medical and health fields, but also realizes the precision, desensitization and efficiency of data acquisition through technological innovation, opening up a new path for the digital transformation and intelligent upgrading of the medical and health industry. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] Figure 1 A flowchart of a medical health behavior data collection system provided in an embodiment of the present application. DETAILED DESCRIPTION

[0051] In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments in the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.

[0052] Embodiment 1:

[0053] See also Figure 1 , this embodiment provides a medical health behavior data collection system, including:

[0054] The video data acquisition and preprocessing module is used to acquire video data and preprocess the video data.

[0055] The video data acquisition and preprocessing module includes:

[0056] The original video capture unit uses a high-precision ordinary camera to capture video data in the scene in real time.

[0057] Edge computing and image analysis unit, used to analyze the collected video data.

[0058] The edge computing and image analysis unit includes:

[0059] The captured video data is transmitted to the chip built into the sensor.

[0060] The powerful computing power of the chip is used to process video data frame by frame, including image decomposition and segmentation into digital image data matrices.

[0061] A deep learning human pose estimation algorithm is applied to identify the human body area in the video data, and perform deep calculation and analysis on the data image matrix of the area.

[0062] The key node extraction and marking module is used to extract and mark key data nodes.

[0063] The key node extraction and marking module includes:

[0064] Generate key node data with clear location information;

[0065] According to the predefined data table format, the information of each key node is marked in detail to ensure the structure and traceability of the data;

[0066] Traverse and mark the key point information of each frame image in the real-time video stream.

[0067] Data transmission and storage module, used for transmitting and storing data;

[0068] The data transmission and storage module includes:

[0069] The key point data information is transmitted through the network protocol.

[0070] The cloud server or dedicated application software database receives and stores data information in real time.

[0071] Data processing and application module, used to process and apply data.

[0072] Data processing and application modules include:

[0073] The data receiving and distributing unit is used to receive data from sensors, store them locally, and then push them to different processing modules according to demand, such as the video simulation module and the deep learning pattern recognition algorithm module.

[0074] The video simulation and display unit is used to simulate and display the data in the form of video.

[0075] The deep learning pattern recognition and statistics unit is used to integrate and analyze data and perform statistics.

[0076] The video simulation and display unit comprises:

[0077] The video simulation module receives the data pushed by the database, and uses the key point video simulation algorithm to restore the discrete key point data frame by frame into a simulated compressed video.

[0078] Output analog video for external display, retaining key information of the original video and reducing storage and transmission burden.

[0079] The deep learning pattern recognition and statistics unit includes:

[0080] After receiving the data, the deep learning pattern recognition algorithm module combines the medical and health knowledge graph to integrate and analyze the key points of each frame.

[0081] Identify specific human postures (such as walking, sitting, falling, etc.), and calculate the simulated video frame by frame based on the processed data labels to generate statistical information for each posture.

[0082] This embodiment realizes long-term, continuous, desensitized and compressed video acquisition and output in multiple scenarios for individuals and families; through intelligent processing of real-time video data, the video content is converted into key point data representing human posture. This process not only effectively reduces the sensitivity of the data, but also can greatly reduce data storage requirements through efficient compression algorithms.

[0083] This embodiment uses a key point video simulation algorithm to restore discrete key point data to a simulated video, which not only retains the key information of the original video, but also achieves data desensitization. At the same time, combined with a deep learning pattern recognition algorithm, it integrates and analyzes key points frame by frame, accurately identifies various postures of the subject (such as walking, sitting, falling, etc.), and generates corresponding statistical information, providing rich and accurate data support for medical diagnosis and health management decisions.

[0084] This embodiment not only solves many pain points of traditional video acquisition methods in the application of the medical and health field, but also realizes the precision, desensitization and efficiency of data acquisition through technological innovation, opening up a new path for the digital transformation and intelligent upgrading of the medical and health industry.

[0085] Embodiment 2:

[0086] Embodiment 2 discloses a method for collecting medical health behavior data. The method is based on the above-mentioned medical health behavior data collection system, and the method includes the following steps:

[0087] S1. Collect video data through a camera.

[0088] S2. Perform edge computing and image analysis on the captured video data.

[0089] S3. Generate key point data of human body posture.

[0090] S4. Store and transmit the generated human body posture key point data.

[0091] S5. Receive data and distribute it.

[0092] S6. Convert key point data into simulated video and display it.

[0093] S7. Calculate the simulated video frame by frame to generate statistical information of various human postures.

[0094] The above contents are further detailed descriptions of the present invention in combination with specific preferred embodiments, and it cannot be determined that the specific implementation of the present invention is limited to these descriptions. For ordinary technicians in the technical field to which the present invention belongs, several simple deductions or substitutions can be made without departing from the concept of the present invention, which should be regarded as falling within the protection scope of the present invention.

Claims

1. A medical health behavior data collection system, characterized in that: The system comprises: Video data acquisition and preprocessing module, used for acquiring video data and preprocessing the video data; Key node extraction and marking module, used to extract and mark key data nodes; Data transmission and storage module, used for transmitting and storing data; Data processing and application module, used to process and apply data.

2. A medical health behavior data collection system as claimed in claim 1, characterized in that: The video data acquisition and preprocessing module includes: A raw video capture unit, which uses a camera to capture video data in real time; Edge computing and image analysis unit, used to analyze the collected video data.

3. A medical health behavior data collection system as claimed in claim 2, characterized in that: The edge computing and image analysis unit includes: Transmit the captured video data to the chip built into the sensor; Processing video data using chips; Identify the human body area in the video data and calculate and analyze the data image matrix of the area.

4. A medical health behavior data collection system as claimed in claim 1, characterized in that: The key node extraction and marking module includes: Generate key node data with clear location information; Mark the information of each key node to ensure the structuring and traceability of data; Traverse and mark the key point information of each frame image in the real-time video stream.

5. A medical health behavior data collection system as claimed in claim 1, characterized in that: The data transmission and storage module includes: Transmit key point data information through network protocol; The cloud server or dedicated application software database receives and stores data information in real time.

6. A medical health behavior data collection system as claimed in claim 1, characterized in that: Data processing and application modules include: The data receiving and distributing unit is used to receive data from sensors, store them locally and then push them; A video simulation and display unit, used for video simulation and display of data; The deep learning pattern recognition and statistics unit is used to integrate and analyze data and perform statistics.

7. A medical health behavior data collection system as claimed in claim 6, characterized in that: The video simulation and display unit comprises: Receive the pushed data, and use the key point video simulation algorithm to restore the discrete key point data frame by frame into a simulated compressed video; Output analog video for external display, retaining key information of the original video and reducing storage and transmission burden.

8. A medical health behavior data collection system as claimed in claim 6, characterized in that: The deep learning pattern recognition and statistics unit includes: After receiving the data, the key points of each frame are integrated and analyzed in combination with the medical and health knowledge graph; Specific human postures are identified, and the simulated video is calculated frame by frame based on the processed data labels to generate statistical information for each posture.

9. A method for collecting medical health behavior data, the method being based on a medical health behavior data collection system according to any one of claims 1 to 8, characterized in that: The method comprises the following steps: S1, collect video data through the camera; S2, performing edge computing and image analysis on the captured video data; S3, generating key point data of human body posture; S4, storing and transmitting the generated human body posture key point data; S5, receiving data and distributing it; S6, converting the key point data into a simulation video and displaying it; S7. Calculate the simulated video frame by frame to generate statistical information of various human postures.

Images