An artificial intelligence-based lying position health risk assessment method

Abstract

The application discloses a lying behavior health risk assessment method based on artificial intelligence and relates to the technical field of health monitoring. Lying bed adaptation data of a lying bed old person is collected, image cleaning is performed on the lying bed adaptation data, standard action images are obtained, three-dimensional marking is performed on the standard action images, boundary feature points are obtained, position marking is performed on the boundary feature points by constructing additional three-dimensional coordinates, boundary feature coordinates are obtained, a monitoring cycle is set to perform cycle collection on the lying bed three-dimensional action images, a comprehensive cycle mapping graph is obtained, an offset threshold is set to perform intelligent analysis on the comprehensive mapping cycle graph, and lying bed change results are obtained, state transition is performed on the lying bed adaptation data, behavior adaptation monitoring graphs are obtained, risk assessment is performed on the lying bed change results through the behavior adaptation monitoring graphs, and individualized intervention schemes are obtained. The lying behavior health risk assessment method realizes group health risk analysis, helps to maintain the life quality of the old person and improves the pertinence and effectiveness of nursing.

Description

Technical Field

[0001] This invention relates to the field of health monitoring technology, specifically an artificial intelligence-based method for assessing health risks associated with lying posture. Background Technology

[0002] With societal development and an aging population, the demand for health management is increasing. Lying position, as an important part of daily life, is closely related to multiple health indicators such as sleep quality, cardiovascular health, and chronic disease risk. However, traditional health risk assessment methods often rely on questionnaires and physiological indicator testing, which have certain limitations. They either require active user participation or support from professional medical equipment and technicians, making it difficult to achieve large-scale, real-time, and continuous health monitoring.

[0003] In recent years, the rapid development of artificial intelligence technology has brought new opportunities to the field of health monitoring. Smart devices and sensors can non-invasively and in real-time monitor an individual's physiological and behavioral data. Developing an AI-based method for assessing health risks related to bedridden behavior involves collecting multi-dimensional data on bedridden behavior, processing and analyzing the data to obtain the behavioral health characteristics of elderly bedridden individuals, and then conducting intelligent assessments based on these characteristics to determine risk levels. This approach not only improves the convenience and accuracy of health monitoring but also provides users with timely health feedback and interventions, which is of great significance for improving public health. Summary of the Invention

[0004] The purpose of this invention is to provide an artificial intelligence-based method for assessing health risks related to lying posture, in order to solve the problems mentioned in the background art, such as the inability to provide real-time data in a timely manner and the difficulty in personalizing assessments.

[0005] An AI-based method for assessing health risks associated with lying posture includes the following steps:

[0006] Step S1: Collect bed rest adaptation data of bedridden elderly and construct a communication virtual space;

[0007] Step S2: Perform image cleaning on the bed rest adaptation data to obtain standardized action images, perform 3D labeling on the standardized action images to obtain boundary feature points, and mark the positions of the boundary feature points by constructing additional 3D coordinates to obtain boundary feature coordinates;

[0008] Step S3: Set the monitoring period to perform periodic aggregation on the bedridden three-dimensional motion chart to obtain a comprehensive periodic mapping chart, and set the offset threshold. Based on the set offset threshold, perform intelligent analysis on the comprehensive mapping periodic chart to obtain the bedridden change results.

[0009] Step S4: Perform state transition on the bed rest adaptation data to obtain a behavior adaptation monitoring map. Use the behavior adaptation monitoring map to conduct a risk assessment on the bed rest change results and obtain a personalized intervention plan.

[0010] Preferably, the process of collecting bedridden elderly people's bedridden adaptation data includes:

[0011] Set up the data collection device for bedridden elderly people to obtain a portable data collection device;

[0012] Data is collected from bedridden elderly people using a portable data collection device to obtain bed rest adaptation data;

[0013] A communication virtual space is constructed based on the information obtained about bedridden elderly individuals, and the obtained bedridden adaptation data is uploaded to the communication virtual space.

[0014] Preferably, the process of image cleaning for bedridden adaptation data includes:

[0015] Initial cleaning of bed rest adaptation data is performed based on the communication virtual space to obtain net bed rest data;

[0016] Perform motion separation on net asset value bed rest data to obtain bed rest motion data and bed rest monitoring data;

[0017] The obtained bedridden movement data is image normalized to obtain standardized movement images.

[0018] Preferably, the process of performing three-dimensional labeling on standardized motion images includes:

[0019] By converting standardized motion images into three dimensions through a virtual communication space, a three-dimensional motion image of bedridden person can be obtained.

[0020] Boundary recognition is performed on the obtained 3D motion images of bedridden patients to obtain boundary feature points;

[0021] Based on the standardized motion images, the obtained bedridden 3D motion images are used to set basic points to obtain marked original points. Based on the marked original points, the obtained bedridden 3D motion images are reconstructed to obtain additional 3D coordinates. Within the additional 3D coordinates, the obtained boundary feature points are marked to obtain boundary feature coordinates.

[0022] Preferably, the process of obtaining the comprehensive cycle mapping diagram includes:

[0023] The monitoring cycle is set based on the bed rest adaptation data, and the bed rest three-dimensional motion map is obtained based on the monitoring cycle. The obtained bed rest three-dimensional motion map is then aggregated and mapped to obtain a comprehensive mapping cycle map.

[0024] Based on the obtained comprehensive mapping periodic map, the boundary feature points are tracked in real state to obtain the periodic offset trajectory;

[0025] Set the offset threshold based on the boundary feature coordinates.

[0026] Preferably, the process for obtaining results of changes in bed rest includes:

[0027] The movement of the periodic offset trajectory is determined based on the offset threshold to obtain the initial distance difference;

[0028] The initial distance difference is summarized based on the periodic offset trajectory to obtain the comprehensive distance difference;

[0029] The obtained comprehensive distance difference is compared with the offset threshold to obtain the limited offset result;

[0030] The monitoring period is continuously compared based on the defined offset results to obtain the results of changes in bed rest.

[0031] Preferably, the process of state transition for bedridden adaptation data includes:

[0032] Bed rest monitoring data is obtained based on the monitoring cycle and recorded as matched monitoring data;

[0033] Based on the monitoring cycle, the obtained matching monitoring data is transformed into a form to obtain a behavior adaptation monitoring map;

[0034] Upload the bed rest change results to the behavior adaptation monitoring map and mark the corresponding bed rest change results on the behavior adaptation monitoring map.

[0035] Preferably, the process of risk assessment of bed rest changes using behavioral adaptation monitoring maps includes:

[0036] Set a balance axis based on the behavior adaptation monitoring map, upload the obtained balance axis to the behavior adaptation monitoring map, and perform a balance check on the behavior adaptation monitoring map through the balance axis to obtain the check monitoring results;

[0037] Based on the changes in bed rest, the inspection and monitoring results are homomorphically determined to obtain behavioral judgment results;

[0038] Based on the behavioral assessment results, the bedridden behavior of elderly people is risk-assessed to obtain a health risk level, and a personalized intervention plan for the elderly people is generated based on the obtained health risk level.

[0039] Compared with the prior art, the beneficial effects of the present invention are:

[0040] 1. Collect comprehensive data on bedridden elderly individuals and construct a virtual storage space for them to store the real-time collected comprehensive data. Centralized data storage facilitates management and analysis, while improving data security and protecting personal privacy. Data cleaning and processing are also performed to separate behavioral and action images for analysis, which helps to more accurately identify potential health risks and assists medical professionals in better monitoring and managing the health status of bedridden elderly individuals, preventing potential health problems.

[0041] 2. Feature extraction is performed on the separated behavioral motion images, and key node coordinates are set to obtain the node movement trajectory within the monitoring period. This helps to identify abnormal behaviors and potential health risks, facilitates the determination of the behavior of bedridden elderly within a unit of time, obtains information on prolonged bed rest and frequent movement, and classifies health risks accordingly. Automated monitoring and risk assessment can reduce the workload of caregivers and improve nursing efficiency; and generate corresponding personalized intervention plans, providing personalized health advice and nursing plans based on the individual differences and behavioral patterns of the elderly, thereby improving the pertinence and effectiveness of nursing care. Attached Figure Description

[0042] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0043] Figure 1 This is a schematic diagram of the present invention. Detailed Implementation

[0044] The technical solution of the present invention will be clearly and completely described below with reference to the embodiments. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0045] like Figure 1 As shown, an artificial intelligence-based method for assessing health risks related to lying posture includes the following steps:

[0046] Step S1: Collect bed rest adaptation data of bedridden elderly and construct a communication virtual space;

[0047] Step S2: Perform image cleaning on the bed rest adaptation data to obtain standardized action images, perform 3D labeling on the standardized action images to obtain boundary feature points, and mark the positions of the boundary feature points by constructing additional 3D coordinates to obtain boundary feature coordinates;

[0048] Step S3: Set the monitoring period to perform periodic aggregation on the bedridden three-dimensional motion chart to obtain a comprehensive periodic mapping chart, and set the offset threshold. Based on the set offset threshold, perform intelligent analysis on the comprehensive mapping periodic chart to obtain the bedridden change results.

[0049] Step S4: Perform state transition on the bed rest adaptation data to obtain a behavior adaptation monitoring map. Use the behavior adaptation monitoring map to conduct a risk assessment on the bed rest change results and obtain a personalized intervention plan.

[0050] It should be further explained that, in the specific implementation process, bedridden behavior, as an important part of people's daily lives, is closely related to multiple health indicators such as sleep quality, cardiovascular health, and chronic disease risk. Therefore, the bedridden behavior of the elderly requires closer observation and assessment to obtain health risks associated with it, greatly improving the health level of individuals and groups. By processing and analyzing the collected data, personalized closed-loop intervention plans are generated based on the analysis results and pushed to caregivers in real time via mobile terminals or nursing management platforms. This enables group health risk analysis and intelligent nursing resource scheduling, thus forming a complete intelligent nursing system from real-time monitoring and risk prediction to closed-loop intervention and trend optimization. The process of collecting bedridden elderly bedridden adaptation data includes:

[0051] Set up the data collection device for bedridden elderly people to obtain a portable data collection device;

[0052] The aforementioned data collection terminal is a port set up for each bedridden elderly person to collect information, i.e., a personal data collection terminal, which can collect all data information of the bedridden elderly person during their bedridden period. In this embodiment, the personal data collection terminal is set at the four feet of the bed where the bedridden elderly person is lying. The bedridden elderly person is a bedridden elderly person who needs to be monitored for health risks. In particular, the personal data collection terminal for collecting physiological data of the bedridden elderly person is a physiological indicator monitoring device set at the bedside. When the bedridden elderly person's physical condition permits, wearable micro sensors can be used to collect physiological data.

[0053] Data is collected from bedridden elderly people using a portable data collection device to obtain bedridden adaptation data, and the obtained bedridden adaptation data is associated with the corresponding bedridden elderly people.

[0054] The data collection refers to the collection of relevant data information generated by bedridden elderly people through a portable data collection device to obtain bed rest adaptation data. The bed rest adaptation data includes physiological data, activity data, lifestyle data, medical history data, and environmental data. Among them, physiological data includes, but is not limited to, the bedridden elderly person's continuous weight, body mass index, basal metabolic rate, heart rate, blood pressure, blood oxygen saturation, and body temperature; activity data includes, but is not limited to, bed rest time, changes in body position, and exercise ability; lifestyle data includes, but is not limited to, nutritional status, emotional stability, and changes in facial features and voice; medical history data includes, but is not limited to, medication use, past medical history, and surgical history; and environmental data includes, but is not limited to, bedroom temperature, humidity, and light.

[0055] In particular, the form of the portable data collection device includes, but is not limited to, sensors and cameras. The data collected for bedridden patients is not only in the form of text data, but also includes images and videos. This is used to collect information in all aspects, provide more comprehensive data monitoring, and facilitate the management of the collected behavioral data of bedridden elderly people in order to analyze health risks.

[0056] A communication virtual space is constructed based on the information obtained about bedridden elderly individuals. This communication virtual space is a virtual space used to store and process the bedridden elderly individuals' bedridden adaptation data, and the stored bedridden adaptation data can be accessed at any time.

[0057] The obtained bed rest adaptation data is uploaded to the communication virtual space.

[0058] Image cleaning is performed on the bed rest adaptation data to obtain standardized action images. These standardized action images are then 3D-labeled to obtain boundary feature points. Finally, additional 3D coordinates are constructed to mark the positions of these boundary feature points, resulting in boundary feature coordinates. The specific process includes:

[0059] Acquire bed rest adaptation data, perform initial cleaning on the acquired bed rest adaptation data based on the communication virtual space, and obtain net bed rest data;

[0060] The initial cleaning refers to performing preliminary data cleaning on the bed rest adaptation data in the communication virtual space, deleting bed rest adaptation data with duplicate times, and clearing redundant abnormal data; wherein, based on the physiological data, activity data, lifestyle data, medical history data, and environmental data included in the bed rest adaptation data, the net value bed rest data includes net value physiological data, net value activity data, net value lifestyle data, net value medical history data, and net value environmental data, all of which are bed rest adaptation data after data cleaning.

[0061] The obtained net value bed rest data is subjected to action separation to obtain bed rest action data and bed rest monitoring data;

[0062] The action separation refers to extracting image-based data related to the movements of bedridden elderly individuals from the net bedridden data, which is called bedridden action data. The remaining data in the net bedridden data is then recorded as bedridden monitoring data, which includes the physiological data, vocal expressions, and other data of the bedridden elderly. For example, images of the elderly person's bedridden state before and after turning over, or images of changes in body position, are bedridden action data separated from image and video data collected by a portable acquisition device. In particular, since the bedridden action data is in image form, the current movement status of the bedridden elderly person can be obtained by processing the image data, such as the bedridden state, turning over state, or prolonged bedridden state.

[0063] The obtained bed rest movement data is image normalized to obtain standardized movement images;

[0064] The image normalization refers to normalizing bed rest motion data by setting constraints. These constraints include size constraints and display constraints. The size constraints include length constraints and width constraints. The length constraint represents the number of pixels in the horizontal direction, and the width constraint represents the number of pixels in the vertical direction. The display constraints represent the allowed colors. For example, if the display constraints are 3 colors, then only the three colors included in the display constraints are allowed to pass, and the remaining colors are filtered out.

[0065] By applying constraints to standardize the images of bedridden movement data, standardized movement images are obtained. This involves unifying the image size and clarity of the bedridden movement data to obtain standardized movement images of the same standard.

[0066] By converting standardized motion images into three dimensions through a virtual communication space, a three-dimensional motion image of bedridden person can be obtained.

[0067] The 3D transformation refers to the 3D transformation of the bedridden elderly in the standardized action image through the communication virtual space. That is, the 2D image is transformed into a 3D action model, which is the bedridden 3D action diagram. For example, by using multi-view geometry technology, the standardized action images of the bedridden elderly from different perspectives are analyzed and the point cloud in 3D space is reconstructed, which is the bedridden 3D action diagram. In particular, the position of the obtained bedridden 3D action diagram in the standardized action image is exactly the same as the position of the person before the 3D transformation. Only the form of the person's expression is transformed from 2D to 3D.

[0068] Boundary recognition is performed on the obtained 3D motion images of bedridden patients to obtain boundary feature points;

[0069] The boundary recognition refers to marking the boundaries of a three-dimensional model of a bedridden elderly person in the communication virtual space, that is, marking the action boundaries of the bedridden elderly person, which are boundary feature points, such as the head, shoulders, waist, hips, body joint nodes, and limb boundary points of the bedridden elderly person. The purpose of obtaining boundary feature points is to obtain the boundary information of each body part in the three-dimensional image of the bedridden elderly person's bedridden image information, which is used to determine the bedridden elderly person's actions and changes in bed.

[0070] Based on the standardized motion images, the obtained bedridden stereoscopic motion image is set with basic points to obtain marked original points. Based on the marked original points, the obtained bedridden stereoscopic motion image is reconstructed to obtain additional three-dimensional coordinates. Within the additional three-dimensional coordinates, the obtained boundary feature points are marked to obtain boundary feature coordinates.

[0071] It should be further explained that, in the specific implementation process, the setting of the base point means selecting a point in the standard motion diagram as the origin of the coordinate system. In this embodiment, the point at the lower left corner of the standard motion diagram is selected as the origin of the coordinate system, which is the marked original point. A three-dimensional coordinate system is constructed at the marked original point, which is the additional three-dimensional coordinate system. The coordinate axes in the three directions are the x-axis, y-axis, and z-axis, respectively. That is, a three-dimensional spatial coordinate system is constructed with the marked original point of the standard motion diagram as the origin, and the lower boundary of the standard motion diagram is the y-axis, and the left boundary of the standard motion diagram is the z-axis.

[0072] The coordinate marking indicates that the coordinates of each boundary feature point are recorded in the additional three-dimensional coordinate system to obtain the boundary feature coordinates. The obtained boundary feature coordinates are marked as (x, y, z), which is the coordinate information of each boundary feature point in the bed rest three-dimensional motion diagram represented by each boundary feature point in the additional three-dimensional coordinate system based on the standard feature motion diagram.

[0073] The monitoring cycle is set to aggregate the bedridden three-dimensional motion charts to obtain a comprehensive cycle mapping chart. An offset threshold is set, and the comprehensive mapping cycle chart is intelligently analyzed based on the set offset threshold to obtain the bedridden change results. The specific process includes:

[0074] The monitoring period is set based on the bed rest adaptation data. The monitoring period refers to a time period set according to the collected bed rest adaptation data. During this time period, the bed rest behavior of the elderly is monitored to see if there are any health-related issues. That is, each piece of bed rest adaptation data has a corresponding collection time. Therefore, the monitoring period also includes a number of bed rest adaptation data.

[0075] Based on the monitoring cycle, obtain bed rest three-dimensional motion images, perform set mapping on the obtained bed rest three-dimensional motion images, and obtain a comprehensive mapping cycle image;

[0076] The set mapping means that during the monitoring period, all bedridden stereoscopic motion images are acquired and mapped in chronological order, with the bedridden stereoscopic motion image at the first time point as the origin. All the acquired bedridden stereoscopic motion images are mapped to the bedridden stereoscopic motion image at the first time point, and the origin of each bedridden stereoscopic motion image is the origin of the bedridden stereoscopic motion image at the first time point. This results in a comprehensive mapping periodic map, indicating that all bedridden stereoscopic motion images are aggregated at one origin point during the monitoring period. Therefore, within the comprehensive mapping periodic map, any selected time point can yield the corresponding bedridden stereoscopic motion image. Thus, during the monitoring period, changes in the behavior and movements of bedridden elderly can be dynamically observed, facilitating the analysis of whether bedridden elderly have been lying down for a long time or their bedridden movements have remained unchanged during the monitoring period.

[0077] Based on the obtained comprehensive mapping periodic map, the boundary feature points are tracked in real state to obtain the periodic offset trajectory;

[0078] It needs further explanation that, in the specific implementation process, the real-state tracking means that in the comprehensive mapping periodic diagram, a boundary feature point is randomly selected as the observation sample point. Within the monitoring period, the different positions of the observation sample point at different time points are highlighted. That is, all observation sample points at all time points are marked in red in the comprehensive mapping periodic diagram to highlight the observation sample points and facilitate the observation of positional changes within a monitoring period. According to the time sequence of the monitoring period, all observation sample points at all time points are connected to obtain the positional movement trajectory of the observation sample point, which is the periodic offset trajectory. Then, one of the remaining boundary feature points is randomly selected as the observation sample point, and the process of obtaining the periodic offset trajectory is repeated until all boundary feature points in the comprehensive mapping periodic diagram have obtained the corresponding periodic offset trajectory, thus completing the real-state tracking. In this way, the change path of each boundary feature point within the monitoring period can be obtained. By comprehensively analyzing the movement trajectories of all boundary feature points, the movement changes of bedridden elderly can be obtained, and it is possible to promptly detect whether the bedridden elderly have turned over or maintained the same movement for a long time, so as to take corresponding preventive and intervention measures to prevent health problems such as pressure sores.

[0079] An offset threshold is set based on the boundary feature coordinates. The offset threshold includes a first offset threshold and a second offset threshold, wherein the first offset threshold is less than the second offset threshold and both are greater than zero.

[0080] The offset threshold refers to the acceptable range of the offset of the corresponding boundary feature point at different time points based on the positional change of the boundary feature coordinates, which is the offset threshold.

[0081] The movement of the periodic offset trajectory is determined based on the obtained offset threshold to obtain the initial distance difference;

[0082] The movement determination means comparing the offset threshold with the periodic offset trajectory, that is, calculating the distance between two adjacent boundary feature points in the periodic offset trajectory and taking the absolute value to obtain the initial distance difference. For example, the initial distance difference between the first boundary feature point and the second boundary feature point in the periodic offset trajectory is calculated, the initial distance difference between the second boundary feature point and the third boundary feature point is calculated, and the initial distance difference between the third boundary feature point and the fourth boundary feature point is calculated.

[0083] The initial distance differences obtained from the periodic offset trajectory are summarized to obtain the comprehensive distance difference. The comprehensive summary means summing up all the initial distance differences obtained from the periodic offset trajectory to obtain the comprehensive distance difference.

[0084] The obtained comprehensive distance difference is compared with the offset threshold to obtain the limited offset result, which includes no change offset, first-level change offset and second-level change offset;

[0085] It should be further explained that, in the specific implementation process, the threshold comparison means comparing the comprehensive distance with the offset threshold to obtain the limited offset result;

[0086] When 0 ≤ comprehensive distance difference ≤ first offset threshold, the periodic offset trajectory corresponding to the comprehensive distance difference is recorded as unchanged offset;

[0087] When the first offset threshold < the comprehensive distance difference < the second offset threshold, the periodic offset trajectory corresponding to the comprehensive distance difference is recorded as the first-level change offset, indicating that the level of the bedridden elderly's movement change within the monitoring period is the first level, which is a moderate change.

[0088] When the second offset threshold is less than or equal to the comprehensive distance difference, the periodic offset trajectory corresponding to the comprehensive distance difference is recorded as the second-level change offset. This indicates that the level of the movement change of the bedridden elderly during the monitoring period is level two, exceeding the intensity of the change offset. That is, the activity intensity of the second-level change offset is higher than the activity intensity of the first-level change offset. The activity intensity includes, but is not limited to, the frequency of body position changes, joint angle changes, and limb length stability.

[0089] The obtained monitoring period is continuously compared based on the obtained limited offset results to obtain the bed rest change results, which include inactive bed rest and active bed rest.

[0090] It should be further explained that, in the specific implementation process, the continuous comparison means that if the limited offset result of the bedridden elderly is unchanged in the current monitoring period, then the limited offset result of the following two monitoring periods will be continuously monitored. If it is still unchanged, that is, the limited offset result of three consecutive monitoring periods is unchanged, the bedridden elderly's bedridden behavior in these three monitoring periods will be marked as inactive bedridden. This indicates that the bedridden elderly has no behavioral changes for a long time, that is, the bedridden elderly's head, shoulders, waist and other body positions have not changed, and there is no behavior such as turning over. In this case, the bedridden elderly are prone to health risks and need to be assessed to reduce the health risks of prolonged bed rest.

[0091] If the limit offset results for three consecutive monitoring periods are either Level 1 or Level 2, the bedridden elderly person's bedridden behavior during these three monitoring periods will be marked as active bedriddenness. In particular, if the limit offset results for three consecutive monitoring periods are all Level 2, it indicates that the bedridden elderly person's bedridden behavior during these three monitoring periods is continuously changing, with high-intensity and frequent changes in movement, which will also put pressure on the bedridden elderly person's health. In this case, the bedridden elderly person's bedridden behavior during these three monitoring periods will be marked as frequent active bedriddenness.

[0092] The bed rest adaptation data is used to perform state transitions to obtain a behavioral adaptation monitoring map. The behavioral adaptation monitoring map is then used to assess the risk of changes in bed rest outcomes, leading to personalized intervention plans. The specific process includes:

[0093] Bed rest monitoring data is obtained based on the monitoring period and is denoted as matched monitoring data. This means that the bed rest monitoring data corresponding to the bed rest movement data of the bed resting elderly within the monitoring period is obtained. For example, the bed rest movement data at each time point corresponds to the bed rest monitoring data, such as the heart rate and respiratory rate of the bed rest image at time point t1.

[0094] Based on the monitoring cycle, the obtained matching monitoring data is transformed into a form to obtain a behavior adaptation monitoring map;

[0095] The formal transformation involves constructing a two-dimensional Cartesian coordinate system with respect to time within the monitoring period, uploading the matched monitoring data to the two-dimensional Cartesian coordinate system in chronological order, generating behavioral monitoring curves based on the matched monitoring data, and marking the two-dimensional Cartesian coordinate system containing the behavioral monitoring curves as a behavioral adaptation monitoring map. In particular, depending on the data types included in the matched monitoring data, the behavioral adaptation monitoring map also includes behavioral monitoring curves corresponding to several data types. Through the behavioral monitoring curves, the physiological and psychological data changes of bedridden elderly people within the monitoring period can be observed intuitively, and the behavioral status of bedridden elderly people can be further identified.

[0096] The obtained bed rest change results are uploaded to the behavior adaptation monitoring map, and the corresponding bed rest change results are marked on the behavior adaptation monitoring map. This means that the bed rest change results for each monitoring period are marked on the behavior adaptation monitoring map, so that the bed rest change results for each monitoring period can be obtained intuitively through the behavior adaptation monitoring map.

[0097] An equilibrium axis is set on the obtained behavior adaptation monitoring map. The equilibrium axis is used to determine whether the changes in the behavior monitoring curves within the behavior adaptation monitoring map are static or fluctuating. It further determines whether the bedridden elderly have changes in behavior during the monitoring period. The distance between the upper and lower equilibrium limits must be small enough to identify matching monitoring data that has not changed. The equilibrium axis includes an upper equilibrium limit and a lower equilibrium limit, which are two straight lines parallel to the horizontal axis of the behavior adaptation monitoring map, with the upper equilibrium limit located above the lower equilibrium limit.

[0098] The obtained equilibrium axis is uploaded to the behavior adaptation monitoring map. The equilibrium of the behavior adaptation monitoring map is checked through the equilibrium axis to obtain the check monitoring results, which include equilibrium fluctuation results and behavior fluctuation results.

[0099] It should be further explained that, in the specific implementation process, the balance check means comparing the behavior monitoring curve with the balance axis in the behavior adaptation monitoring graph. The portion of the behavior monitoring curve that falls above the lower balance limit and below the upper balance limit is recorded as the balance docking curve. This indicates that the matching monitoring data of this portion of the curve has no significant fluctuations, or in other words, no change. The proportion of the balance docking curve to the behavior monitoring curve is calculated and recorded as the balance proportion. A proportion threshold is set and compared with the balance proportion. When the balance proportion is greater than the proportion threshold, the behavior monitoring curve is recorded as the balance fluctuation result, indicating that within the monitoring period, a large part of the matching monitoring data is balanced fluctuation, i.e., static fluctuation, without significant fluctuation amplitude. Here, the proportion threshold represents a threshold of the proportion ratio, for example, 90% or 95%.

[0100] Conversely, if the balance ratio is less than or equal to the ratio threshold, the behavior monitoring curve is recorded as a behavior fluctuation result, indicating that the physiological and psychological data fluctuates greatly during the monitoring period, and there may be changes in behavior. By matching the bed rest change results at the corresponding time, it is determined whether it is a behavior change or a disease change.

[0101] Based on the obtained bed rest change results, the inspection and monitoring results are homomorphically determined to obtain behavioral determination results, which include static bed rest results and dynamic bed rest results;

[0102] Furthermore, the homomorphic determination means that in the behavior adaptation monitoring map, the inspection monitoring results are compared with the bed rest change results. When the bed rest change result is no movement in bed and the inspection monitoring result is a balanced fluctuation result, it means that the bedridden elderly in the corresponding monitoring period has not shown any behavior, that is, the elderly have been lying in bed for a long time and have maintained one movement. The behavior determination result of the bedridden elderly is recorded as the static bed rest result.

[0103] When the bed rest change result is active bed rest and the inspection and monitoring result is behavioral fluctuation result, it means that the bed resting elderly person has behavioral changes within the corresponding monitoring period, and the behavioral judgment result of the bed resting elderly person is recorded as dynamic bed rest result;

[0104] Based on the obtained behavioral assessment results, the bedridden behavior of elderly people is risk-assessed to obtain a health risk level, and a personalized intervention plan for elderly people is generated based on the obtained health risk level.

[0105] It should be further explained that, in the specific implementation process, the risk assessment means to conduct a health assessment of the behavior of bedridden elderly based on the behavior judgment results, obtain a health risk level, and record the bedridden elderly whose behavior judgment result is static bed rest as prolonged bed rest health risk. Based on the health problems caused by prolonged bed rest, the corresponding personalized intervention plan includes the treatment and nursing process for the bedridden patient to solve the health problems caused by prolonged bed rest during the monitoring period, and the plan corresponding to this process is recorded as a personalized intervention plan.

[0106] Specifically, when the behavior assessment result is dynamic bed rest, and the health risk level of frequent movement bed rest is recorded as high-frequency movement risk, it means that high-intensity movement of bed resting elderly people will also be detrimental to their recuperation. It is necessary to further generate corresponding nursing plans, namely personalized intervention plans, such as solving pain and breathing difficulties caused by excessive turning over.

[0107] The preferred embodiments of the present invention disclosed above are merely illustrative of the invention. These preferred embodiments do not exhaustively describe all details, nor do they limit the invention to any specific implementation. Clearly, many modifications and variations can be made based on the content of this specification. This specification selects and specifically describes these embodiments to better explain the principles and practical applications of the invention, thereby enabling those skilled in the art to better understand and utilize the invention. The invention is limited only by the claims and their full scope and equivalents.

Claims

1. A method for assessing health risks related to supine posture based on artificial intelligence, characterized in that, Includes the following steps: Step S1: Collect bed rest adaptation data of bedridden elderly and construct a communication virtual space; Step S2: Perform image cleaning on the bed rest adaptation data to obtain standardized action images, perform 3D labeling on the standardized action images to obtain boundary feature points, and mark the positions of the boundary feature points by constructing additional 3D coordinates to obtain boundary feature coordinates; Step S3: Set the monitoring period to perform periodic aggregation on the bedridden three-dimensional motion chart to obtain a comprehensive periodic mapping chart, and set the offset threshold. Based on the set offset threshold, perform intelligent analysis on the comprehensive mapping periodic chart to obtain the bedridden change results. Step S4: Perform state transition on the bed rest adaptation data to obtain a behavior adaptation monitoring map. Use the behavior adaptation monitoring map to conduct a risk assessment on the bed rest change results and obtain a personalized intervention plan.

2. The method for assessing health risks of supine behavior based on artificial intelligence according to claim 1, characterized in that, The process of collecting bedridden elderly people's bedridden adaptation data includes: Set up the data collection device for bedridden elderly people to obtain a portable data collection device; Data is collected from bedridden elderly people using a portable data collection device to obtain bed rest adaptation data; A communication virtual space is constructed based on the information obtained about bedridden elderly individuals, and the obtained bedridden adaptation data is uploaded to the communication virtual space.

3. The method for assessing health risks of supine behavior based on artificial intelligence according to claim 1, characterized in that, The process of image cleaning for bedridden adaptation data includes: Initial cleaning of bed rest adaptation data is performed based on the communication virtual space to obtain net bed rest data; Perform motion separation on net asset value bed rest data to obtain bed rest motion data and bed rest monitoring data; The obtained bedridden movement data is image normalized to obtain standardized movement images.

4. The method for assessing health risks of supine behavior based on artificial intelligence according to claim 1, characterized in that, The process of performing 3D labeling on standardized motion images includes: By converting standardized motion images into three dimensions through a virtual communication space, a three-dimensional motion image of bedridden person can be obtained. Boundary recognition is performed on the obtained 3D motion images of bedridden patients to obtain boundary feature points; Based on the standardized motion images, the obtained bedridden 3D motion images are used to set basic points to obtain marked original points. Based on the marked original points, the obtained bedridden 3D motion images are reconstructed to obtain additional 3D coordinates. Within the additional 3D coordinates, the obtained boundary feature points are marked to obtain boundary feature coordinates.

5. The method for assessing health risks of supine behavior based on artificial intelligence according to claim 4, characterized in that, The process of obtaining the comprehensive cycle mapping diagram includes: The monitoring cycle is set based on the bed rest adaptation data, and the bed rest three-dimensional motion map is obtained based on the monitoring cycle. The obtained bed rest three-dimensional motion map is then aggregated and mapped to obtain a comprehensive mapping cycle map. Based on the obtained comprehensive mapping periodic map, the boundary feature points are tracked in real state to obtain the periodic offset trajectory; Set the offset threshold based on the boundary feature coordinates.

6. The method for assessing health risks of supine behavior based on artificial intelligence according to claim 5, characterized in that, The process of obtaining results of changes in bed rest includes: The movement of the periodic offset trajectory is determined based on the offset threshold to obtain the initial distance difference; The initial distance difference is summarized based on the periodic offset trajectory to obtain the comprehensive distance difference; The obtained comprehensive distance difference is compared with the offset threshold to obtain the limited offset result; The monitoring period is continuously compared based on the defined offset results to obtain the results of changes in bed rest.

7. The method for assessing health risks of supine behavior based on artificial intelligence according to claim 3, characterized in that, The process of state transitioning bedridden fit data includes: Bed rest monitoring data is obtained based on the monitoring cycle and recorded as matched monitoring data; Based on the monitoring cycle, the obtained matching monitoring data is transformed into a form to obtain a behavior adaptation monitoring map; Upload the bed rest change results to the behavior adaptation monitoring map and mark the corresponding bed rest change results on the behavior adaptation monitoring map.

8. The method for assessing health risks of supine behavior based on artificial intelligence according to claim 1, characterized in that, The process of risk assessment of changes in bed rest outcomes using behavioral adaptation monitoring maps includes: Set a balance axis based on the behavior adaptation monitoring map, upload the obtained balance axis to the behavior adaptation monitoring map, and perform a balance check on the behavior adaptation monitoring map through the balance axis to obtain the check monitoring results; Based on the changes in bed rest, the inspection and monitoring results are homomorphically determined to obtain behavioral judgment results; Based on the behavioral assessment results, the bedridden behavior of elderly people is risk-assessed to obtain a health risk level, and a personalized intervention plan for the elderly people is generated based on the obtained health risk level.

Images