A method and system for pressure sore risk assessment and automatic body position adjustment of an intelligent nursing bed

Abstract

This invention discloses a method and system for pressure ulcer risk assessment and automatic body positioning adjustment on an intelligent nursing bed, relating to the field of intelligent medical and nursing equipment technology. The method includes: collecting surface contact pressure distribution data and microenvironmental temperature and humidity data of the pressure interface from a bedridden patient; mapping the surface contact pressure distribution data to a pressure topology map; calculating the local skin tolerance attenuation coefficient; weighting and correcting the pressure topology map to generate a multidimensional coupled risk distribution matrix; identifying risk extreme value regions in the multidimensional coupled risk distribution matrix; obtaining their cumulative risk values ​​and comparing them with dynamic intervention thresholds to determine the corresponding target airbag group; and generating asymmetric wave-like charging and discharging control commands based on the distribution pattern of the risk extreme value regions to adjust the local geometric contour of the intelligent nursing bed's support surface. This invention enables refined assessment of pressure ulcer risk and risk-driven body positioning adjustment.

Description

Technical Field

[0001] This invention relates to the field of intelligent medical and nursing equipment technology, and in particular to an intelligent nursing bed pressure ulcer risk assessment and automatic body position adjustment method and system. Background Technology

[0002] With the deepening of population aging and the continuous increase in the number of long-term bedridden patients, pressure ulcers, as a secondary injury caused by prolonged local tissue pressure, microcirculatory disturbances, and decreased skin tolerance, have become one of the complications that urgently need to be controlled in the field of clinical nursing. In recent years, intelligent nursing beds and anti-decubitus mattress technologies have gradually developed towards digitalization, sensing, and intelligence. By integrating pressure sensors, temperature and humidity sensors, and adjustable support structures into the bed support surface, they can assist in the intervention of pressure status and positional changes of bedridden patients. Existing research shows that the occurrence of pressure ulcers is not only related to the instantaneous pressure peak, but also closely related to the coupling effect of multiple factors such as the duration of pressure, changes in local skin microenvironment temperature and humidity, and the decline in tissue tolerance. However, most current intelligent nursing bed systems still use a single pressure threshold or empirical rules as the basis for risk assessment, and there is insufficient modeling of the coupling relationship between factors such as pressure, temperature, and humidity, making it difficult to truly reflect the damage evolution process at the tissue level. At the same time, positional adjustment often adopts symmetrical, static, or regular airbag inflation and deflation strategies, making it difficult to make fine interventions based on the spatial distribution pattern of risk areas, resulting in a certain lag and uncertainty in the protective effect.

[0003] CN121401062A discloses an intelligent pressure-adaptive anti-bedsore care mattress system. This system uses a flexible pressure-sensing layer on the mattress surface to monitor the pressure distribution of different parts of the body in real time. Combined with a multi-chamber dynamic adjustment structure, it adjusts the pressure in the air chambers according to control commands generated by an edge intelligent control module. Simultaneously, it incorporates temperature and humidity sensors and a microenvironment control device to collaboratively improve the local environment. This solution achieves a certain degree of linkage between body pressure monitoring, risk prediction, and support adjustment, possessing basic anti-bedsore functions. However, the risk assessment model of this technology is mainly based on pressure distribution characteristics or artificial intelligence prediction results. Temperature and humidity data are used more as independent control parameters, without quantitatively modeling the impact of temperature and humidity on the decline of skin tolerance, nor feeding this impact back into the pressure risk distribution correction process. This makes it difficult to form a coupled evaluation result that reflects the actual risk of tissue damage. Furthermore, its air chamber adjustment strategy aims at regional pressure balance, without considering the differences in the shape of risk areas and the direction of force. The body position adjustment method is relatively coarse-grained, and the problem of local risk accumulation still exists in complex pressure scenarios.

[0004] CN120661333A discloses a method for dynamic adjustment of bedridden patient positioning based on infrared sensing. This method constructs a monitoring network using a multi-channel infrared thermal imaging sensor and a 3D depth camera to fuse the patient's surface temperature distribution with their body contour, thereby identifying areas of concentrated pressure and providing early warning of potential pressure injury risks. This approach offers certain technical advantages in non-contact monitoring, precise temperature field perception, and early risk detection, reducing the workload of nursing staff. However, this method primarily relies on changes in surface temperature and geometric contours to infer areas of concentrated pressure, without directly collecting data on the distribution of pressure on the body surface, making it difficult to accurately depict the true stress state of the support interface. Furthermore, its risk identification logic focuses on detecting abnormal areas, lacking a quantitative description of the persistence of pressure and dynamic changes in skin tolerance. It also fails to construct a unified risk coupling model based on multi-source information such as pressure, temperature, and humidity, resulting in positioning decisions still being driven by empirical rules, making it difficult to achieve targeted and gradual risk mitigation control.

[0005] In summary, existing intelligent nursing beds and related postural adjustment technologies generally suffer from insufficient characterization of the pressure ulcer risk formation mechanism, a single dimension of risk assessment, and inadequate matching between postural adjustment strategies and risk spatial distribution, making it difficult to achieve accurate assessment and proactive intervention of actual tissue damage risk. This invention proposes a method and system for pressure ulcer risk assessment and automatic postural adjustment in intelligent nursing beds. By constructing a multimodal sensing system on the support surface of the intelligent nursing bed, it jointly models the surface pressure distribution and pressure microenvironment parameters to form a coupled risk distribution result reflecting the actual tissue damage risk. Based on this, it drives the dynamic adjustment of the support structure to alleviate the accumulation of local risks. Summary of the Invention

[0006] The purpose of this section is to outline some aspects of the embodiments of the present invention and to briefly introduce some preferred embodiments. Some simplifications or omissions may be made in this section, as well as in the abstract and title of the present application, to avoid obscuring the purpose of this section, the abstract and the title of the invention. Such simplifications or omissions shall not be used to limit the scope of the present invention.

[0007] Given that existing intelligent nursing beds and anti-bedsore technologies often focus on a single pressure or temperature indicator in the process of pressure ulcer risk assessment, lack comprehensive modeling of changes in tissue tolerance under the coupled influence of multiple factors such as pressure, temperature, and humidity, and that the matching degree between body position adjustment strategies and the spatial distribution of high-risk areas is insufficient, making it difficult to conduct refined and dynamic intervention on pressure ulcer risk, this invention is proposed.

[0008] Therefore, the problem to be solved by this invention is how to construct a coupled risk assessment model that reflects the actual risk of tissue damage based on multimodal perception, and generate a body position adjustment and control strategy that matches the risk distribution pattern, so as to reduce the accumulation of local risks and improve the pertinence and reliability of pressure ulcer prevention and control.

[0009] To solve the above-mentioned technical problems, the present invention provides the following technical solution: In a first aspect, embodiments of the present invention provide a method for intelligent nursing bed pressure ulcer risk assessment and automatic body positioning adjustment, comprising, By laying a multimodal sensor array on the support surface of the intelligent nursing bed, the body surface contact pressure distribution data, micro-environment temperature data and micro-environment humidity data of the bedridden object are collected simultaneously, and the body surface contact pressure distribution data is mapped into a pressure topology map. Based on the microenvironment temperature data and microenvironment humidity data, the local skin tolerance attenuation coefficient is calculated, and the corresponding area in the pressure topology map is weighted and corrected using the local skin tolerance attenuation coefficient to generate a multidimensional coupled risk distribution matrix. Identify the extreme risk regions in the multidimensional coupled risk distribution matrix, extract the cumulative risk value of the extreme risk regions, and compare the cumulative risk value with a preset dynamic intervention threshold to determine the target airbag group; Based on the distribution pattern of the risk extreme value region, an asymmetric wave-like charge-discharge control command is generated for the target airbag group. The asymmetric wave-like charge-discharge control command is executed to change the local geometric contour of the intelligent nursing bed support surface until the value of the updated risk extreme value region in the multidimensional coupled risk distribution matrix is ​​lower than the dynamic intervention threshold.

[0010] Secondly, embodiments of the present invention provide an intelligent nursing bed pressure ulcer risk assessment and automatic body position adjustment system, comprising: The multimodal data acquisition and pressure mapping module is used to simultaneously acquire surface contact pressure distribution data, microenvironmental temperature data, and microenvironmental humidity data of the bedridden object through a multimodal sensor array laid on the support surface of the intelligent nursing bed, and to map the surface contact pressure distribution data into a pressure topology map. The risk coupling assessment module calculates the local skin tolerance attenuation coefficient based on the microenvironment temperature data and microenvironment humidity data, and uses the local skin tolerance attenuation coefficient to perform weighted correction on the corresponding area in the pressure topology map to generate a multidimensional coupling risk distribution matrix. The risk area identification and airbag positioning module is used to identify the extreme risk areas in the multidimensional coupled risk distribution matrix, extract the cumulative risk value of the extreme risk areas, and compare the cumulative risk value with a preset dynamic intervention threshold to determine the target airbag group. The body position adjustment control module generates an asymmetric wave-like charge-discharge control command for the target airbag group based on the distribution pattern of the risk extreme value region. The module executes the asymmetric wave-like charge-discharge control command to change the local geometric contour of the intelligent nursing bed support surface until the value of the updated risk extreme value region in the multidimensional coupled risk distribution matrix is ​​lower than the dynamic intervention threshold.

[0011] Thirdly, embodiments of the present invention provide a computer device, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement any step of the above-described intelligent nursing bed pressure ulcer risk assessment and automatic body position adjustment method.

[0012] Fourthly, embodiments of the present invention provide a computer-readable storage medium having a computer program stored thereon, wherein: when the computer program is executed by a processor, it implements any step of the above-described intelligent nursing bed pressure ulcer risk assessment and automatic body position adjustment method.

[0013] Compared with existing technologies, the beneficial effects of this invention are as follows: By deploying a multimodal sensor array to synchronously collect pressure, temperature, and humidity data, and mapping the pressure data into a pressure topology map, integrated in-situ monitoring and spatial visualization of pressure ulcer risk factors are achieved; based on temperature and humidity data, the local skin tolerance attenuation coefficient is calculated, and the pressure topology map is weighted and corrected to generate a multidimensional coupled risk distribution matrix, thereby quantifying the physiological and pathological mechanisms into a dynamic risk field model, realizing a scientific upgrade from monitoring a single physical quantity to predicting multi-factor coupled risks; extreme value regions in the risk matrix are identified and their cumulative risk values ​​are extracted, and the target airbag group is determined by comparing with the dynamic intervention threshold, establishing an on-demand triggering decision-making mechanism based on real-time risk load, achieving precise and personalized intervention; based on the distribution pattern of the risk area, an asymmetric wave-like charging and discharging control command is generated, and the support surface contour is adjusted through a progressive pressure transfer method simulating ergonomics, which efficiently releases local pressure while minimizing interference to the bedridden patient, ultimately achieving proactive, dynamic, and intelligent prevention of pressure ulcer risk. Attached Figure Description

[0014] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the description of the embodiments 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. Wherein: Figure 1 A flowchart for a method of pressure ulcer risk assessment and automatic body positioning adjustment for intelligent nursing beds; Figure 2This is a structural diagram of an intelligent nursing bed pressure ulcer risk assessment and automatic body position adjustment system. Detailed Implementation

[0015] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments.

[0016] Based on the embodiments of this invention, all other embodiments obtained by those skilled in the art without inventive effort should fall within the scope of protection of this invention.

[0017] Many specific details are set forth in the following description in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways different from those described herein, and those skilled in the art can make similar extensions without departing from the spirit of the invention. Therefore, the invention is not limited to the specific embodiments disclosed below.

[0018] As mentioned in the background section, existing intelligent nursing beds and related pressure ulcer prevention technologies often rely on single pressure or temperature indicators in pressure ulcer risk assessment. They lack systematic modeling of the coupling effects of multiple factors, such as the distribution of pressure on the body surface and the temperature and humidity of the pressure microenvironment. Furthermore, the matching degree between body positioning strategies and the spatial distribution characteristics of high-risk areas is insufficient, making it difficult to accurately characterize and intervene in the dynamic evolution of pressure ulcer risk. To address these problems, this invention provides a method for pressure ulcer risk assessment and automatic body positioning on an intelligent nursing bed.

[0019] Reference Figure 1 , Figure 1 This is a flowchart illustrating a method for pressure ulcer risk assessment and automatic body positioning adjustment on an intelligent nursing bed according to an embodiment of the present invention. Figure 1 As shown, a method for assessing pressure ulcer risk and automatically adjusting body position on an intelligent nursing bed includes: S1: Through a multimodal sensor array laid on the support surface of the intelligent nursing bed, the surface contact pressure distribution data, micro-environment temperature data and micro-environment humidity data of the bedridden object are collected simultaneously, and the surface contact pressure distribution data is mapped into a pressure topology map. S1.1: A multimodal sensor array is laid out inside the mattress layer of the intelligent nursing bed support surface according to a preset grid layout. It should be noted that the multimodal sensor array includes uniformly distributed pressure sensing units, temperature sensing units, and humidity sensing units; the spatial resolution of the pressure sensing unit is set to four independent measuring points per square decimeter; the temperature sensing unit and the humidity sensing unit form a one-to-one spatial correspondence with the pressure sensing unit, so that each pressure measuring point is equipped with a corresponding temperature measuring point and humidity measuring point.

[0020] S1.2: When the bedridden person lies on the support surface of the intelligent nursing bed, each sensing unit in the multimodal sensor array starts the synchronous acquisition mode; Specifically, the pressure sensing unit outputs the pressure values ​​at each measuring point to form the surface contact pressure distribution data, the temperature sensing unit outputs the temperature values ​​at each measuring point to form the micro-environment temperature data of the pressure interface, and the humidity sensing unit outputs the humidity values ​​at each measuring point to form the micro-environment humidity data of the pressure interface. The data acquisition frequency of all sensing units is uniformly set to perform five sampling operations per second.

[0021] S1.3: Arrange the pressure values ​​of each measuring point in the body surface contact pressure distribution data according to the two-dimensional coordinate positions corresponding to the preset grid layout to construct a two-dimensional data matrix that reflects the spatial distribution characteristics of pressure. The row index and column index of the two-dimensional data matrix correspond to the grid coordinates of the intelligent nursing bed support surface along the length and width directions, respectively. S1.4: The pressure values ​​in the two-dimensional data matrix are interpolated. The pressure estimate in the middle position is filled between adjacent measuring points using a bilinear interpolation algorithm. The bilinear interpolation algorithm calculates the pressure value of the interpolation point based on the pressure values ​​of four adjacent measuring points according to the distance weight. The grid density of the data matrix after interpolation is increased to four times that of the original measuring point density. S1.5: Convert the interpolated data matrix into a visual pressure topology map, where each pixel in the pressure topology map carries the pressure value information of the corresponding grid coordinate position; Preferably, the pressure topology map uses a color gradient mapping method to express the difference in pressure values, where high pressure areas are mapped to warm color levels and low pressure areas are mapped to cool color levels.

[0022] S1.6: Add a unified timestamp to the surface contact pressure distribution data, the microenvironment temperature data of the pressure interface, and the microenvironment humidity data of the pressure interface. Furthermore, timestamps record the precise moment when each sensing unit completes a single sampling operation. The timestamps establish the temporal correspondence between the three types of sensing data, ensuring the time synchronization of data from each measurement point during subsequent multi-dimensional data fusion analysis.

[0023] S2: Based on microenvironment temperature and humidity data, calculate the local skin tolerance attenuation coefficient, and use the local skin tolerance attenuation coefficient to perform weighted correction on the corresponding area in the pressure topology map to generate a multidimensional coupled risk distribution matrix. S2.1: Extract the temperature values ​​at each measuring point in the microenvironment temperature data of the pressure interface, and calculate the temperature deviation parameter at each measuring point; Furthermore, the temperature deviation parameter is obtained by calculating the difference between the temperature value of each measuring point and the normal skin temperature of the human body. If the temperature deviation parameter is positive, it indicates that there is a local temperature rise at the measuring point; if the temperature deviation parameter is negative, it indicates that there is a local temperature drop at the measuring point.

[0024] S2.2: Extract the humidity values ​​at each measuring point in the microenvironment humidity data of the pressure interface, and calculate the humidity saturation parameter at each measuring point; Specifically, the humidity saturation parameter is obtained by calculating the ratio of the humidity value at each measuring point to the upper limit of the comfortable humidity range on the skin surface. If the humidity saturation parameter is greater than the unit value, it means that the humidity environment at that measuring point exceeds the skin's tolerance range. The larger the humidity saturation parameter value, the more severe the local humidity.

[0025] S2.3: For each measurement point in the multimodal sensor array, perform nonlinear coupling calculation on the temperature deviation parameter and humidity saturation parameter corresponding to this measurement point to obtain the temperature-humidity coupling factor. Furthermore, an exponential function-based enhancement term is introduced during the calculation of the temperature-humidity coupling factor. If both the temperature deviation parameter and the humidity saturation parameter are in the unfavorable range, the exponential function-based enhancement term will amplify the temperature-humidity coupling factor. The unfavorable range is defined as the state where the temperature deviation parameter is greater than the preset temperature threshold and the humidity saturation parameter is greater than the preset humidity threshold.

[0026] S2.4: Based on the temperature and humidity coupling factor, the local skin tolerance attenuation coefficient corresponding to each measuring point is obtained by querying the pre-stored skin physiological response curve database; Preferably, the skin physiological response curve database stores the mapping relationship between the values ​​of different temperature and humidity coupling factors and the degree of skin tolerance decay. The value range of the local skin tolerance decay coefficient is set between zero and one. The closer the value is to zero, the more severe the skin tolerance decay at that measurement point.

[0027] S2.5: Multiply the pressure value carried by each pixel in the pressure topology map with the local skin tolerance attenuation coefficient at the grid coordinate position corresponding to the pixel to obtain the corrected pressure risk value of the pixel. It should be noted that the corrected pressure risk value reflects the combined effect of the pressure load and the degree of skin tolerance reduction. If the pressure value at a certain measuring point is high and the local skin tolerance attenuation coefficient is small, the corrected pressure risk value at that measuring point will increase significantly.

[0028] S2.6: Replace the pressure values ​​of all pixels in the pressure topology map with the corresponding corrected pressure risk values, and construct a risk value matrix with the same spatial dimension structure as the pressure topology map. The row and column indices in the risk value matrix are consistent with those in the pressure topology map, and each matrix element stores the corrected pressure risk value at the corresponding grid coordinate position. S2.7: Normalize the risk numerical matrix and use the normalized risk numerical matrix as the multidimensional coupled risk distribution matrix; Specifically, the numerical range of all elements in the risk value matrix is ​​uniformly scaled to the interval between zero and one hundred. The magnitude of the values ​​in the multidimensional coupled risk distribution matrix intuitively reflects the relative risk of tissue damage at each measurement point. The closer the value is to one hundred, the higher the risk of pressure sores forming at that location.

[0029] S2.8: For each matrix element in the multidimensional coupled risk distribution matrix, add the original measured values ​​of the surface contact pressure distribution data, the microenvironment temperature data of the pressure interface, the microenvironment humidity data of the pressure interface, and the local skin tolerance attenuation coefficient as traceability information. The traceability information is recorded in the extended attribute field of the matrix element.

[0030] In an optional embodiment, when an elderly patient is lying supine on a smart nursing bed, a multimodal sensor array detects a high-risk area in the sacral and coccygeal region. The surface contact pressure distribution data for this area shows a local pressure as high as 80 mmHg, while microenvironmental temperature data indicates that the skin temperature at this location is 2.5°C higher than the surrounding area, and microenvironmental humidity data shows a humidity saturation parameter of 1.3 (above the upper limit of skin comfort). The system calculates the local skin tolerance attenuation coefficient and weights the pressure topology map of this area to generate a multidimensional coupled risk distribution matrix. The matrix shows that the risk value of this area is 85 (full scale 100), identified by the connected component detection algorithm as an independent extreme risk region. Its cumulative risk value exceeds the dynamically calculated intervention threshold for this patient's individual needs; for example, due to their advanced age and low Braden score, the threshold is dynamically lowered.

[0031] S3: Identify the extreme risk regions in the multidimensional coupled risk distribution matrix, extract the cumulative risk value of the extreme risk regions, and compare the cumulative risk value with the preset dynamic intervention threshold to determine the target airbag group; S3.1: Perform region segmentation on the multidimensional coupled risk distribution matrix and use a connected component detection algorithm to identify regions in the matrix where risk values ​​are continuously distributed, as risk extreme value regions; S3.1.1: Set risk threshold judgment criteria, and calculate the mean parameter and standard deviation parameter of the matrix element values ​​by statistically analyzing the numerical distribution characteristics of all matrix elements in the multidimensional coupled risk distribution matrix. Specifically, the mean parameter is obtained by traversing each element of the multidimensional coupled risk distribution matrix, summing the risk values ​​of all matrix elements, and dividing by the total number of matrix elements. The squared deviation between the risk value of each matrix element and the mean parameter is calculated, and the square root of the sum of all squared deviations is obtained by dividing by the total number of matrix elements. The mean parameter reflects the central trend of the overall risk distribution, and the standard deviation parameter reflects the dispersion of the risk values.

[0032] S3.1.2: Calculate the dynamic risk segmentation threshold based on the mean parameter and standard deviation parameter; Furthermore, the product of the mean parameter and the standard deviation parameter is added together to obtain the dynamic risk segmentation threshold, where the weighting coefficient of the standard deviation parameter is set to 1.5 times. The dynamic risk segmentation threshold is higher than the mean parameter and is used to screen out high-risk measurement point locations in the multidimensional coupled risk distribution matrix that deviate significantly from the normal level.

[0033] Preferably, if the overall pressure distribution of the bedridden object is relatively uniform, the standard deviation parameter value is small, the dynamic risk segmentation threshold is correspondingly reduced, and the sensitivity to the identification of local risk points is improved; if the pressure distribution of the bedridden object is significantly uneven, the standard deviation parameter value increases, the dynamic risk segmentation threshold is correspondingly increased, and normal pressure fluctuations are avoided from being misjudged as high-risk areas.

[0034] S3.1.3: Traverse each element of the multidimensional coupled risk distribution matrix and compare the risk value of the matrix element with the dynamic risk segmentation threshold; It should be noted that if the risk value of a matrix element is greater than or equal to the dynamic risk segmentation threshold, the matrix element is marked as a high-risk pixel and a high-risk identifier is written into the marking attribute field of the matrix element; if the risk value of a matrix element is less than the dynamic risk segmentation threshold, the matrix element is marked as a low-risk pixel and a low-risk identifier is written into the marking attribute field of the matrix element; after the traversal operation is completed, all matrix elements in the multidimensional coupled risk distribution matrix carry the corresponding risk level identifier.

[0035] S3.1.4: Construct spatial adjacency relationship determination rules. For each matrix element marked as a high-risk pixel in the multidimensional coupling risk distribution matrix, detect whether there are other high-risk pixels within the spatial neighborhood of this matrix element in the multidimensional coupling risk distribution matrix. Preferably, the spatial neighborhood range is defined using an eight-neighborhood connection pattern. That is, with the row index and column index of the current matrix element as the center, the positions of the eight adjacent matrix elements in the four diagonal directions, where the row index is incremented by one, the row index is decremented by one, the column index is incremented by one, the column index is decremented by one, and the positions of the eight adjacent matrix elements are included in the spatial neighborhood range. The eight adjacent matrix elements in the spatial neighborhood range are traversed, and the risk level identifier in the label attribute field of each adjacent matrix element is read. If at least one adjacent matrix element carries a high-risk identifier, it is determined that there is a spatial connectivity relationship between the current matrix element and that adjacent matrix element.

[0036] S3.1.5: Based on spatial connectivity, perform connected component labeling operation to merge all high-risk pixels in the multidimensional coupled risk distribution matrix that are interconnected through spatial connectivity into the same connected component set; Specifically, a seed-filling algorithm is used to traverse the multidimensional coupled risk distribution matrix. Starting with the first detected high-risk pixel, it is used as a seed point. The algorithm recursively checks the adjacent matrix elements within the eight-neighbor range of the seed point. If an adjacent matrix element is also marked as a high-risk pixel and has not yet been merged into any connected component set, then this adjacent matrix element is added to the current connected component set, and this adjacent matrix element is used as a new seed point to continue expanding the check outwards. If it is impossible to continue expanding to new high-risk pixels, a unique connected component number is assigned to the current connected component set, and the connected component number is written into the extended attribute field of all matrix elements in the connected component set. The above operation is repeated until all high-risk pixels in the multidimensional coupled risk distribution matrix are merged into the corresponding connected component sets.

[0037] S3.1.6: Count the number of matrix elements contained in each connected component set to obtain the area parameters of each connected component set; It should be noted that the area parameter is obtained by accumulating the number of matrix elements with the same connected component number within the connected component set. The value of the area parameter directly reflects the spatial coverage of the high-risk area. If the area parameter value of a connected component set is large, it indicates that there is a large area of ​​continuous high pressure load or a significant decrease in skin tolerance, and the risk of tissue damage is significantly increased.

[0038] S3.1.7: Calculate the arithmetic mean of the risk values ​​of all matrix elements in each connected component set to obtain the regional average risk intensity parameter for each connected component set; Furthermore, the risk value stored in each matrix element within the connected domain set is extracted. The risk values ​​of all matrix elements within the connected domain set are summed and then divided by the area parameter of the connected domain set to obtain the regional average risk intensity parameter. The regional average risk intensity parameter reflects the central tendency of the overall risk level within the connected domain set. The closer the value is to one hundred, the more likely the measurement points within the connected domain set are in a high-risk state.

[0039] S3.1.8: For each set of connected components, multiply the area parameter of the region of this set of connected components with the average risk intensity parameter of the region to obtain the comprehensive risk weight factor of the set of connected components. Preferably, the comprehensive risk weighting factor considers both the spatial scale and risk intensity level of high-risk areas. When the area parameter of the connected region set is large and the average risk intensity parameter is high, the comprehensive risk weighting factor value increases significantly, indicating that the body part corresponding to the connected region set faces a serious threat of pressure ulcer formation. Introducing the comprehensive risk weighting factor avoids misjudgments based solely on a single-dimensional indicator. For example, certain small but extremely high-risk local pressure concentration points, as well as certain large but moderately high-risk broad pressure areas, can all be reasonably prioritized using the comprehensive risk weighting factor.

[0040] S3.1.9: Set connected component filtering conditions to filter out the set of connected components where the comprehensive risk weight factor is greater than the preset weight threshold, and use them as risk extreme value regions; It should be noted that the weight thresholds are set based on the critical intervention triggering conditions defined in the clinical pressure ulcer care standards, and the weight values ​​that are significantly correlated with the probability of pressure ulcer occurrence are obtained by fitting a large amount of clinical trial data.

[0041] Specifically, all connected component sets are traversed, and the comprehensive risk weight factor value of each connected component set is read. If the comprehensive risk weight factor is greater than the weight threshold, the connected component set is marked as a risk extreme value region, and the risk extreme value identifier is written into the extended attribute field of all matrix elements in this connected component set. After the screening operation is completed, the connected component set marked as a risk extreme value region in the multidimensional coupled risk distribution matrix is ​​the target region that needs to be immediately subjected to postural adjustment intervention.

[0042] S3.1.10: Extract the grid coordinate positions corresponding to all matrix elements within each risk extreme value region, calculate the minimum row index, maximum row index, minimum column index, and maximum column index of the grid coordinate positions, and construct the minimum bounding rectangle bounding box surrounding the risk extreme value region; It should be noted that the four boundary lines of the minimum bounding rectangle are determined by the minimum row index, the maximum row index, the minimum column index, and the maximum column index, respectively. The minimum bounding rectangle completely covers the spatial distribution range of the risk extreme value area. The geometric parameters of the minimum bounding rectangle provide a precise spatial mapping reference for locating the target airbag group on the support surface of the intelligent nursing bed in subsequent steps.

[0043] S3.1.11: Encapsulate the geometric parameters, connected component numbers, region area parameters, region average risk intensity parameters, and comprehensive risk weight factors of the minimum bounding rectangle of each risk extreme value region into a risk region feature data structure; Furthermore, the risk area feature data structure is stored in a structured data format, where the geometric parameter field records the minimum and maximum values ​​of the row index, the minimum and maximum values ​​of the column index; the identifier field records the connected component number; the area field records the area parameter; the intensity field records the average risk intensity parameter of the area; and the weight field records the comprehensive risk weight factor.

[0044] S3.2: Extract the risk values ​​of all matrix elements within the risk extreme value region, and sum the risk values ​​of all matrix elements within the risk extreme value region to obtain the cumulative risk value of this region; Preferably, the cumulative risk value comprehensively reflects the superposition effect of the risk area range and risk intensity level of the region. When the area of ​​a region is large or the risk intensity is high, the cumulative risk value of the region increases accordingly.

[0045] S3.3: Retrieve the dynamic intervention threshold stored in the control system and determine whether the cumulative risk value exceeds the dynamic intervention threshold; S3.2.1: Establish a traceability linked list for the risk extreme value region, and traverse all matrix elements carrying risk extreme value identifiers within the risk extreme value region; Specifically, the connected component numbers stored in the extended attribute field of each matrix element are read. Matrix elements with the same connected component number are arranged in ascending order of row index and then in ascending order of column index within the same row, constructing an ordered sequence of matrix element accesses. This ordered sequence ensures the spatial continuity of subsequent risk value extraction operations and avoids the reduced data reading efficiency caused by random access.

[0046] S3.2.2: Access each matrix element in the ordered matrix element access sequence in sequence, extract the risk value stored in the matrix element, and add it to the cumulative risk value variable in the risk extreme value region; It should be noted that the cumulative risk value variable is initialized to zero. After extracting the risk value of each matrix element, it is immediately added to the current cumulative risk value variable, and the result is updated to the cumulative risk value variable. After completing the entire traversal of the ordered matrix element access sequence, the value stored in the cumulative risk value variable is the sum of the risk values ​​of all measurement points within the risk extreme value region.

[0047] S3.2.3: Simultaneously count the number of matrix elements within the risk extreme value region to obtain the risk measurement point count parameter, and calculate the ratio of the cumulative risk value to the risk measurement point count parameter to obtain the regional mean risk density of the risk extreme value region; Furthermore, the risk point count parameter is obtained by accumulating the number of matrix elements in the ordered matrix element access sequence. The regional mean risk density reflects the average risk contribution intensity of a unit point location within the risk extreme value region. If the cumulative risk value is large but the risk point count parameter is also large, the regional mean risk density may be at a moderate level; if the cumulative risk value is large and the risk point count parameter is small, the regional mean risk density increases significantly, indicating that there is a high-intensity risk concentration phenomenon in the region.

[0048] S3.3: Retrieve the dynamic intervention threshold stored in the control system and determine whether the cumulative risk value exceeds the dynamic intervention threshold; S3.3.1: Read the preset benchmark intervention threshold and threshold adjustment coefficient library from the threshold parameter storage module of the control system; Preferably, the baseline intervention threshold is set according to the emergency intervention triggering conditions defined in the clinical pressure ulcer care guidelines, representing the upper limit of risk accumulation for bedridden subjects of standard body type in the standard lying position; the threshold adjustment coefficient library stores threshold correction coefficients corresponding to different influencing factors, including the body mass index classification coefficient, age segment coefficient, skin condition assessment coefficient, and continuous bed rest duration coefficient for bedridden subjects.

[0049] S3.3.2: Retrieve individualized physiological profile data of the bedridden subject, and extract the body mass index, age, skin Braden score, and current duration of continuous bed rest. Specifically, individualized physiological profile data is pre-stored in the patient information database of the control system, including basic physiological parameters collected upon admission of bedridden patients and health status assessment results updated regularly by nursing staff; the body mass index is obtained by dividing the bedridden patient's weight measurement by the square of the height measurement; the skin Braden score is calculated based on the assessment results of six dimensions: sensory ability, moisture level, activity level, mobility, nutritional status, and friction and shear force; and the current duration of continuous bed rest is automatically accumulated by the intelligent nursing bed usage time recording module.

[0050] S3.3.3: Query the corresponding body mass index (BMI) grading coefficient in the threshold adjustment coefficient library based on the BMI value; It should be noted that the threshold adjustment coefficient library predefines the mapping relationship between body mass index (BMI) value ranges and BMI grading coefficients. When the BMI value is below the normal range, the corresponding BMI grading coefficient is less than one; when the BMI value is within the normal range, the corresponding BMI grading coefficient is equal to one; and when the BMI value is above the normal range, the corresponding BMI grading coefficient is greater than one. Bedridden subjects with a low BMI are more prone to tissue ischemia under pressure due to insufficient buffering capacity of subcutaneous fat tissue; therefore, the intervention threshold needs to be lowered to increase protective sensitivity. Bedridden subjects with a high BMI also require a higher threshold correction level because their body weight increases local pressure load.

[0051] S3.3.4: Query the corresponding age segmentation coefficient in the threshold adjustment coefficient library based on the age value; Furthermore, the threshold adjustment coefficient library divides age values ​​into multiple age ranges, with each age range corresponding to a different age segment coefficient. The older the bedridden subject, the weaker their skin elasticity and microcirculation compensatory ability, and the smaller the corresponding age segment coefficient value. Therefore, the intervention threshold needs to be lowered to provide more proactive preventive protection.

[0052] S3.3.5: Query the corresponding skin condition assessment coefficient in the threshold adjustment coefficient library based on the skin Braden score; Preferably, the lower the Braden score, the higher the intrinsic risk of pressure ulcers in bedridden subjects, and the smaller the corresponding skin condition assessment coefficient. The skin condition assessment coefficient directly reflects the current skin health baseline and damage resistance reserve of bedridden subjects. When the skin condition assessment coefficient is significantly less than one, it indicates that the intervention threshold needs to be significantly reduced to accommodate the fragile skin tolerance of bedridden subjects.

[0053] S3.3.6: Query the corresponding continuous bed rest duration coefficient in the threshold adjustment coefficient library based on the current continuous bed rest duration value; Specifically, the continuous bed rest duration coefficient decreases as the bed rest duration increases, reflecting the cumulative effect of tissue fatigue caused by maintaining a fixed position for a long time. When the current continuous bed rest duration exceeds the preset fatigue accumulation threshold, the continuous bed rest duration coefficient decreases significantly, indicating that even if the risk value does not reach the standard threshold level, intervention should be triggered in advance due to the time-dependent decay of tissue tolerance.

[0054] S3.3.7: Multiply the baseline intervention threshold with the body mass index grading coefficient, age segmentation coefficient, skin condition assessment coefficient, and continuous bed rest duration coefficient to obtain the dynamic intervention threshold for the current bedridden subject; It should be noted that the dynamic intervention threshold, through a composite correction of multi-dimensional coefficients, transforms the standardized baseline intervention threshold into a customized judgment standard adapted to the current physiological characteristics and bedridden status of the patient. The calculation process of the dynamic intervention threshold realizes the transformation from a group-based nursing standard to a precise individual protection strategy, avoiding the problems of over-intervention or under-intervention caused by using a uniform threshold.

[0055] S3.3.8: When the cumulative risk value exceeds the dynamic intervention threshold, the emergency intervention process is initiated to identify the target airbag group covering the extreme risk area and trigger local pressure release operations, specifically including: The deviation between the cumulative risk value and the dynamic intervention threshold is calculated to obtain the risk exceedance index. The risk exceedance index is obtained by subtracting the dynamic intervention threshold from the cumulative risk value and then dividing by the dynamic intervention threshold to obtain the relative deviation percentage value. The risk exceedance index reflects the severity of the current risk level exceeding the safety boundary. The larger the value, the more urgent and stronger the intervention measures are required. Based on the risk exceedance level index, the corresponding intervention priority level code is queried from the preset intervention intensity grading table. The intervention intensity grading table divides the risk exceedance level index into multiple intervals, each interval corresponding to a different intervention priority level code. If the risk exceedance level index is less than 20%, a Level 1 intervention priority level code is assigned, and normal intensity postural adjustment is performed. If the risk exceedance level index is between 20% and 50%, a Level 2 intervention priority level code is assigned, and moderate intensity postural adjustment is performed with a shortened adjustment cycle. If the risk exceedance level index exceeds 50%, a Level 3 intervention priority level code is assigned, and maximum intensity postural adjustment is performed while simultaneously triggering an alarm notification for nursing staff. The geometric parameters of the minimum bounding rectangle of the risk extreme value region are mapped to the spatial coordinates of the airbag distribution map of the smart nursing bed to determine the target airbag group covering the risk extreme value region. The smart nursing bed airbag distribution map pre-stores the grid coordinate range and airbag number of all airbag units on the mattress support surface. By comparing the row index range and column index range of the minimum bounding rectangle with the grid coordinate range of each airbag unit, the airbag units with overlapping grid coordinates are selected, and the airbag numbers of these airbag units are summarized to form the target airbag group. An intervention command package containing the target airbag group number, intervention priority level code, and risk extreme value area characteristic data structure is sent to the body position adjustment execution module. This triggers a local pressure release operation targeting the risk extreme value area. The intervention priority level code in the intervention command package determines the airbag inflation and deflation speed parameters and pressure adjustment amplitude parameters. The area morphology information in the risk extreme value area characteristic data structure guides the selection of the coordinated action mode of the airbag group. Upon receiving the intervention command package, the body position adjustment execution module immediately interrupts the current routine monitoring task and prioritizes the execution of the emergency intervention process.

[0056] In an optional embodiment, the system then determines the target airbag group covering the sacral and coccygeal risk area. In step S4, the system analyzes that the risk area is approximately elliptical in shape, with its major axis forming an angle of about 15 degrees with the longitudinal axis of the bed. The three airbag units located at the core of the risk area in the target airbag group are classified as the core pressure release airbag group and sorted according to their modified pressure risk values. The central airbag unit with the highest risk is the first to deflate. At the same time, the six airbag units on the periphery of the target airbag group are classified as the peripheral auxiliary adjustment airbag group and different inflation phase parameters are assigned to them according to their orientation relative to the risk center point (e.g., upper left, lower right, etc.).

[0057] S3.3.9: If multiple extreme risk regions exist and their corresponding cumulative risk values ​​are all greater than the dynamic intervention threshold, then the intervention priorities for these multiple extreme risk regions shall be ranked, specifically including: Extract the comprehensive risk weight factor value for each risk extreme value region, and arrange multiple risk extreme value regions in descending order of comprehensive risk weight factor value; create a priority queue data structure for risk extreme value regions, inserting the connected component number and corresponding comprehensive risk weight factor value of each risk extreme value region as queue elements, and using the quicksort algorithm to sort them in descending order based on the comprehensive risk weight factor value. After sorting, the risk extreme value region at the head of the queue is the region with the highest comprehensive risk level and the one that needs to be dealt with most first; Check whether the risk extreme value regions of adjacent sorting positions have spatial overlap or close proximity; read the geometric parameters of the minimum bounding rectangle bounding boxes of two adjacent risk extreme value regions, and calculate the Euclidean distance between the coordinates of the center points of the two minimum bounding rectangle bounding boxes; if the Euclidean distance is less than the preset spatial association judgment threshold, it is determined that the two risk extreme value regions are spatially close and may belong to different parts of the same large-scale pressure area, and need to be considered for coordinated processing when planning intervention strategies; Multiple spatially adjacent high-risk regions are merged and marked to generate a set of joint intervention regions. The target airbag group numbers of multiple spatially adjacent high-risk regions are combined to expand and form an extended target airbag group corresponding to the set of joint intervention regions. The intervention operation of the set of joint intervention regions avoids the problem of secondary pressure concentration in the boundary area that may be caused by independent adjustment of adjacent regions. A smoother pressure gradient transition is achieved through large-scale coordinated adjustment. According to the priority queue of risk extreme value areas, intervention instruction packages corresponding to each risk extreme value area or set of joint intervention areas are sequentially issued to the body position adjustment execution module. A mandatory delay waiting time is set after each intervention operation is executed. The mandatory delay waiting time is used to ensure that the body position adjustment action of the previous risk extreme value area is completely completed and the mattress pressure distribution reaches a stable state before starting the intervention operation of the next risk extreme value area. The setting of the mandatory delay waiting time avoids the risk of violent shaking or instability caused to the bedridden person by performing large adjustment actions in multiple areas at the same time, ensuring the smoothness and safety of the body position adjustment process.

[0058] S3.3.10: When the cumulative risk value is less than or equal to the dynamic intervention threshold, the risk trend monitoring process is initiated, specifically including: Calculate the safety margin parameter between the cumulative risk value and the dynamic intervention threshold. The safety margin parameter is obtained by subtracting the cumulative risk value from the dynamic intervention threshold and then dividing by the dynamic intervention threshold to obtain the relative safety margin percentage. The safety margin parameter reflects the remaining buffer space between the current risk level and the triggering intervention condition. The larger the value, the safer the current state. The smaller the value, the closer the state is to the intervention threshold, and the more critical the state is, the more monitoring frequency is needed. The historical cumulative risk value sequence corresponding to the risk extreme value area is retrieved from the historical monitoring data storage module, and the cumulative risk value data points within the most recent five consecutive monitoring periods are extracted. The historical cumulative risk value sequence records the trajectory of the cumulative risk value change of the risk extreme value area in the past multiple monitoring periods in the order of timestamps. The most recent five data points are extracted to form a short-term risk trend analysis sample set, which can reflect the recent dynamic evolution characteristics of risk values.

[0059] Linear regression was performed on five cumulative risk value data points in the short-term risk trend analysis sample set to obtain the risk growth rate index by calculating the slope parameter of the fitted line. The least squares method was then used to perform linear regression on the five data points, and the slope parameter of the fitted line is the risk growth rate index. If the risk growth rate index is positive and large, it indicates that the cumulative risk value in the extreme risk region is showing a continuous upward trend. Even if the current cumulative risk value has not exceeded the threshold, based on the current growth rate, it may reach a level requiring intervention in a short period of time. If the risk growth rate index is negative, it indicates that the cumulative risk value in the extreme risk region is showing a downward trend, suggesting that the pressure load on this region from the current postural state is being alleviated. Based on the safety margin parameter and the risk growth rate index, the remaining time window for predicting the intervention threshold is calculated. The difference in risk value corresponding to the safety margin parameter is divided by the risk growth rate index to obtain the estimated time required for the current cumulative risk value to grow to the dynamic intervention threshold according to the current growth trend, which is the remaining time window. If the remaining time window is less than the preset warning time threshold, it is determined that the risk extreme value area is in a state of rapid deterioration, and preventive intervention measures need to be taken in advance. If the remaining time window is less than the preset warning time threshold, a warning message containing the location information of the risk extreme value area and the remaining time window value is pushed to the nursing staff's terminal device. The warning message is displayed simultaneously on the visual monitoring panel of the nursing station and the mobile terminal device carried by the nursing staff, prompting the nursing staff to pay attention to the position of the bedridden person and to intervene in advance to manually adjust the position or check the skin condition of the area. The warning message realizes the upgrade of the monitoring mode from passive response to intervention threshold triggering to proactive prediction of risk trends. If the remaining time window is greater than or equal to the preset warning time threshold, the current monitoring cycle frequency will remain unchanged, and the sensor data of the multimodal sensor array will continue to be collected and the multidimensional coupled risk distribution matrix will be updated according to the standard time interval. The standard time interval is set according to clinical nursing guidelines and is usually a complete data collection and risk assessment cycle performed every 30 to 60 minutes. When the risk extreme value area is determined to be in a stable and safe state, there is no need to increase the monitoring frequency or trigger intervention operations. Maintaining the regular monitoring rhythm can effectively balance the monitoring effect and system resource consumption. The cumulative risk value of the current monitoring period for the risk extreme value area is appended to the end of the historical cumulative risk value sequence, and the oldest data point in the historical cumulative risk value sequence is deleted, ensuring that the historical cumulative risk value sequence always includes data from the most recent five monitoring periods. A first-in, first-out (FIFO) data update strategy is used to maintain the historical cumulative risk value sequence, ensuring that trend analysis is always based on the latest time window data, improving the timeliness and accuracy of risk trend judgment. The data update operation at the end of each monitoring period provides a continuously evolving data foundation for the trend analysis of the next period.

[0060] S4: Based on the distribution pattern of the risk extreme value region, generate asymmetric wave-like charge and discharge control commands for the target airbag group, execute the asymmetric wave-like charge and discharge control commands to change the local geometric contour of the intelligent nursing bed support surface, until the value of the updated risk extreme value region in the multidimensional coupled risk distribution matrix is ​​lower than the dynamic intervention threshold.

[0061] S4.1: Analyze the geometric shape characteristics of the risk extreme value region in the multidimensional coupled risk distribution matrix, and calculate the major axis direction and minor axis direction of this region; Specifically, the direction of maximum variance is obtained by principal component analysis of the coordinate positions of all matrix elements in the region along the major axis, while the direction of the minor axis is an orthogonal direction perpendicular to the major axis. The angle between the major axis and the length direction of the support surface of the intelligent nursing bed is recorded as the risk area deflection angle.

[0062] S4.2: Based on the long axis and short axis, the airbag units in the target airbag group are divided into the core pressure release airbag group and the peripheral auxiliary adjustment airbag group. Furthermore, the core pressure relief airbag group includes airbag units located within a preset radius around the peak coordinates of the risk extreme value region, and the peripheral auxiliary adjustment airbag group includes the remaining airbag units in the target airbag group other than the core pressure relief airbag group.

[0063] S4.3: Calculate the deflation priority coefficient based on the corrected pressure risk value of each airbag unit in the core pressure release airbag group at the corresponding position; Furthermore, the deflation priority coefficient is directly proportional to the corrected pressure risk value. The airbag units in the core pressure release airbag group are sorted from high to low according to the deflation priority coefficient, and the airbag unit with the highest priority coefficient performs the deflation operation first.

[0064] S4.4: Calculate the inflation phase parameters based on the spatial orientation relationship of each airbag unit in the peripheral auxiliary adjustment airbag group relative to the peak coordinates of the risk extreme value area; Preferably, the inflation phase parameter is determined by the difference between the angle of the line connecting each airbag unit to the peak coordinate and the deflection angle of the risk area. Airbag units located at different positions outside the risk extreme value area are assigned different inflation phase parameters.

[0065] S4.5: Based on the deflation priority coefficient and inflation phase parameters, construct an asymmetric wave-like charge and discharge timing table to generate the air pressure regulation parameters for each airbag unit; It should be noted that the asymmetric wave-like charge and discharge timing table records the start time and duration of the charge and discharge actions of each airbag unit in the target airbag group. For the airbag units in the core pressure release airbag group, the start time of the deflation is set sequentially according to the order of the deflation priority coefficient. The interval between the start times of the deflation of adjacent priority airbag units is set to a preset time interval. For the airbag units in the peripheral auxiliary adjustment airbag group, the start time of the inflation is set according to the inflation phase parameter. The start times of the inflation of airbag units with values ​​close to the inflation phase parameter are close, forming a wave-like inflation sequence that is gradually pushed from the outside to the inside.

[0066] Specifically, the air pressure regulation parameter of the airbag unit in the core pressure release airbag group is set to a negative value to indicate that a deflation operation is performed. The deflation amount is proportional to the corrected pressure risk value of the airbag unit. The air pressure regulation parameter of the airbag unit in the peripheral auxiliary adjustment airbag group is set to a positive value to indicate that an inflation operation is performed. The inflation amount is set inversely proportional to the distance between the airbag unit and the boundary of the risk extreme value area. The closer the distance, the smaller the inflation amount.

[0067] S4.7: Encapsulate the asymmetric wave-type charge / discharge timing table and air pressure regulation parameters into an asymmetric wave-type charge / discharge control command. It should be noted that the asymmetric wave-type charge and discharge control command is sent to the airbag control module, which drives the charge and discharge valves of each airbag unit to perform corresponding operations according to the command content. The asymmetric wave-type charge and discharge control command includes the airbag number, action type, start time, duration and adjustment value of each airbag unit.

[0068] In an optional embodiment, the system generates and executes asymmetric wave-like charge and discharge control commands. The core airbags deflate sequentially according to priority, rapidly reducing the direct pressure on the sacrum and coccyx. At the same time, the peripheral airbags do not inflate synchronously, but rather, according to their phase parameters, they inflate and lift slightly from the outside to the inside in an orderly manner, like waves. This gently pushes the patient's center of gravity to the side and upward without causing drastic changes in body position, effectively transferring and dispersing the pressure on the sacrum and coccyx.

[0069] S4.8: During the execution of the asymmetric wave-type charge and discharge control command, the multimodal sensor array continuously collects updated body surface contact pressure distribution data, micro-environment temperature data of the pressure interface, and micro-environment humidity data of the pressure interface, and regenerates the updated multidimensional coupling risk distribution matrix according to the processing flow of steps S1 and S2. It should be noted that the updated multidimensional coupled risk distribution matrix reflects the changes in pressure and risk distribution caused by airbag adjustment operations.

[0070] S4.9: Extract the risk value corresponding to the original risk extreme value region in the updated multidimensional coupled risk distribution matrix, and calculate the current updated cumulative risk value at that position; Specifically, the updated cumulative risk value is compared with the dynamic intervention threshold. If the updated cumulative risk value is still higher than the dynamic intervention threshold, steps S4.1 to S4.8 are repeated to generate a new asymmetric wave-type charge and discharge control command and continue to adjust.

[0071] Furthermore, when the cumulative risk value after the update decreases below the dynamic intervention threshold, the sending of new asymmetric wave-like charge and discharge control commands is stopped, the current inflation state of each airbag unit is maintained, and the total execution time of this body position adjustment operation, the number of airbag unit adjustments, and the magnitude of the risk value decrease are recorded as adjustment effect evaluation data. The adjustment effect evaluation data is used to optimize the dynamic intervention threshold and air pressure adjustment parameter settings in subsequent similar risk scenarios.

[0072] During the execution of adjustment commands, the sensor array continuously collects data. When the updated multidimensional coupled risk distribution matrix shows that the cumulative risk value of the original risk area has dropped below the dynamic intervention threshold (e.g., from 85 to 30), the system stops issuing new adjustment commands and records assessment data such as the adjustment time, the number of airbags involved, and the degree of risk reduction. If the risk value does not meet the target, the system will generate and fine-tune control commands again based on new pressure and microenvironment data until the risk is effectively mitigated.

[0073] In summary, this invention achieves integrated in-situ monitoring and spatial visualization of pressure ulcer risk factors by deploying a multimodal sensor array to simultaneously collect pressure, temperature, and humidity data, and mapping the pressure data into a pressure topology map. Based on temperature and humidity data, it calculates the local skin tolerance attenuation coefficient and performs weighted correction on the pressure topology map to generate a multidimensional coupled risk distribution matrix, thereby quantifying the physiological and pathological mechanisms into a dynamic risk field model. This represents a scientific upgrade from monitoring a single physical quantity to predicting risks coupled with multiple factors. It identifies extreme value regions in the risk matrix and extracts their cumulative risk values, determining the target airbag group by comparing them with dynamic intervention thresholds. This establishes an on-demand triggering decision-making mechanism based on real-time risk load, achieving precise and personalized intervention. Based on the distribution pattern of risk areas, it generates asymmetric wave-like charging and discharging control commands, adjusting the support surface contour through a progressive pressure transfer method simulating ergonomics. This efficiently releases local pressure while minimizing interference with bedridden individuals, ultimately achieving proactive, dynamic, and intelligent prevention of pressure ulcer risk.

[0074] Under the teachings of the above embodiments, such as Figure 2 As shown in the embodiments of the present invention, other aspects also disclose an intelligent nursing bed pressure ulcer risk assessment and automatic body position adjustment system, comprising: The multimodal data acquisition and pressure mapping module is used to simultaneously acquire the body surface contact pressure distribution data, micro-environment temperature data and micro-environment humidity data of the bedridden object through a multimodal sensor array laid on the support surface of the intelligent nursing bed, and map the body surface contact pressure distribution data into a pressure topology map. The risk coupling assessment module calculates the local skin tolerance attenuation coefficient based on microenvironment temperature and humidity data, and uses the local skin tolerance attenuation coefficient to perform weighted correction on the corresponding area in the pressure topology map to generate a multidimensional coupling risk distribution matrix. The risk area identification and airbag positioning module is used to identify the extreme risk areas in the multidimensional coupled risk distribution matrix, extract the cumulative risk value of the extreme risk areas, and compare the cumulative risk value with the preset dynamic intervention threshold to determine the target airbag group. The body position adjustment control module generates asymmetric wave-like charge and discharge control commands for the target airbag group based on the distribution pattern of the risk extreme value region. The asymmetric wave-like charge and discharge control commands are executed to change the local geometric contour of the support surface of the intelligent nursing bed until the value of the updated risk extreme value region in the multidimensional coupled risk distribution matrix is ​​lower than the dynamic intervention threshold.

[0075] This embodiment also provides a computer device applicable to the intelligent nursing bed pressure ulcer risk assessment and automatic body position adjustment method, including a memory and a processor; the memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions to realize the intelligent nursing bed pressure ulcer risk assessment and automatic body position adjustment method as proposed in the above embodiment.

[0076] The computer device can be a terminal, comprising a processor, memory, communication interface, display screen, and input devices connected via a system bus. The processor provides computing and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The communication interface is used for wired or wireless communication with external terminals; wireless communication can be achieved through Wi-Fi, carrier networks, NFC (Near Field Communication), or other technologies. The display screen can be an LCD screen or an e-ink screen. The input devices can be a touch layer covering the display screen, buttons, a trackball, or a touchpad on the computer device's casing, or an external keyboard, touchpad, or mouse.

[0077] This embodiment also provides a storage medium storing a computer program that, when executed by a processor, implements the method for intelligent nursing bed pressure ulcer risk assessment and automatic body position adjustment as proposed in the above embodiments.

[0078] The storage medium proposed in this embodiment and the data storage method proposed in the above embodiments belong to the same inventive concept. Technical details not described in detail in this embodiment can be found in the above embodiments, and this embodiment has the same beneficial effects as the above embodiments.

[0079] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A method for assessing pressure ulcer risk and automatically adjusting body position on an intelligent nursing bed, characterized in that: include, By laying a multimodal sensor array on the support surface of the intelligent nursing bed, the body surface contact pressure distribution data, micro-environment temperature data and micro-environment humidity data of the bedridden object are collected simultaneously, and the body surface contact pressure distribution data is mapped into a pressure topology map. Based on the microenvironment temperature data and microenvironment humidity data, the local skin tolerance attenuation coefficient is calculated, and the corresponding area in the pressure topology map is weighted and corrected using the local skin tolerance attenuation coefficient to generate a multidimensional coupled risk distribution matrix. Identify the extreme risk regions in the multidimensional coupled risk distribution matrix, extract the cumulative risk value of the extreme risk regions, and compare the cumulative risk value with a preset dynamic intervention threshold to determine the target airbag group; Based on the distribution pattern of the risk extreme value region, an asymmetric wave-like charge-discharge control command is generated for the target airbag group. The asymmetric wave-like charge-discharge control command is executed to change the local geometric contour of the intelligent nursing bed support surface until the value of the updated risk extreme value region in the multidimensional coupled risk distribution matrix is ​​lower than the dynamic intervention threshold.

2. The method for assessing pressure ulcer risk and automatically adjusting body position on an intelligent nursing bed as described in claim 1, characterized in that: The method for generating the asymmetric wave-type charge and discharge control command is as follows: Analyze the geometric shape characteristics of the risk extreme value region in the multidimensional coupled risk distribution matrix, and calculate the major axis and minor axis directions of this region; Based on the major axis and minor axis, each airbag unit in the target airbag group is divided into a core pressure release airbag group and a peripheral auxiliary adjustment airbag group. Calculate the deflation priority coefficient based on the corrected pressure risk value of each airbag unit in the core pressure release airbag group at the corresponding position. Based on the spatial orientation relationship of each airbag unit in the peripheral auxiliary adjustment airbag group relative to the peak coordinates of the risk extreme value region, the inflation phase parameters are calculated. Based on the deflation priority coefficient and the inflation phase parameter, an asymmetric wave-like charge and discharge timing table is constructed to generate the air pressure regulation parameters for each airbag unit. The asymmetric wave-shaped charge-discharge timing table and the air pressure regulation parameters are encapsulated into asymmetric wave-shaped charge-discharge control commands.

3. The method for assessing pressure ulcer risk and automatically adjusting body position on an intelligent nursing bed as described in claim 2, characterized in that: The asymmetric wave-type charge-discharge control command is sent to the airbag control module, wherein the airbag control module drives the charge-discharge valves of each airbag unit to perform corresponding operations according to the command content; the asymmetric wave-type charge-discharge control command includes the airbag number, action type, start time, duration and adjustment value of each airbag unit.

4. The method for assessing pressure ulcer risk and automatically adjusting body position on an intelligent nursing bed as described in claim 3, characterized in that: The cumulative risk value is compared with a preset dynamic intervention threshold to determine the target airbag group, including: Retrieve the dynamic intervention threshold stored in the control system and determine whether the cumulative risk value exceeds the dynamic intervention threshold; When the cumulative risk value exceeds the dynamic intervention threshold, the emergency intervention process is initiated to identify the target airbag group covering the extreme risk area and trigger a local pressure release operation. When the cumulative risk value is less than or equal to the dynamic intervention threshold, the risk trend monitoring process is initiated.

5. The method for pressure ulcer risk assessment and automatic body position adjustment of an intelligent nursing bed as described in claim 4, characterized in that: The method for obtaining the cumulative risk value is as follows: A region segmentation operation is performed on the multidimensional coupled risk distribution matrix, and a connected component detection algorithm is used to identify regions in the matrix where risk values ​​are continuously distributed, which are then used as risk extreme value regions. Extract the risk values ​​of all matrix elements within the risk extreme value region, and sum the risk values ​​of all matrix elements within the risk extreme value region to obtain the cumulative risk value of this region.

6. The method for pressure ulcer risk assessment and automatic body position adjustment of an intelligent nursing bed as described in claim 5, characterized in that: The method for generating the multidimensional coupled risk distribution matrix is ​​as follows: Extract the temperature values ​​at each measuring point from the microenvironment temperature data of the pressure interface, and calculate the temperature deviation parameter at each measuring point; Extract the humidity values ​​at each measuring point from the microenvironment humidity data of the pressure interface, and calculate the humidity saturation parameter at each measuring point; For each measurement point in the multimodal sensor array, the temperature deviation parameter and the humidity saturation parameter corresponding to this measurement point are nonlinearly coupled and calculated to obtain the temperature-humidity coupling factor. Based on the temperature and humidity coupling factor, the local skin tolerance attenuation coefficient corresponding to each measuring point is obtained by querying the pre-stored skin physiological response curve database. The pressure value carried by each pixel in the pressure topology map is multiplied by the local skin tolerance attenuation coefficient at the grid coordinate position corresponding to the pixel to obtain the corrected pressure risk value of the pixel. Replace the pressure values ​​of all pixels in the pressure topology map with the corresponding corrected pressure risk values ​​to construct a risk value matrix with the same spatial dimension structure as the pressure topology map. The risk numerical matrix is ​​normalized, and the normalized risk numerical matrix is ​​used as the multidimensional coupled risk distribution matrix.

7. The method for pressure ulcer risk assessment and automatic body position adjustment of an intelligent nursing bed as described in claim 6, characterized in that: The pressure topology map uses a color gradient mapping method to express the difference in pressure values, where high pressure areas are mapped to warm color levels and low pressure areas are mapped to cool color levels.

8. A smart nursing bed pressure ulcer risk assessment and automatic body position adjustment system, based on the smart nursing bed pressure ulcer risk assessment and automatic body position adjustment method according to any one of claims 1 to 7, characterized in that: include, The multimodal data acquisition and pressure mapping module is used to simultaneously acquire surface contact pressure distribution data, microenvironmental temperature data, and microenvironmental humidity data of the bedridden object through a multimodal sensor array laid on the support surface of the intelligent nursing bed, and to map the surface contact pressure distribution data into a pressure topology map. The risk coupling assessment module calculates the local skin tolerance attenuation coefficient based on the microenvironment temperature data and microenvironment humidity data, and uses the local skin tolerance attenuation coefficient to perform weighted correction on the corresponding area in the pressure topology map to generate a multidimensional coupling risk distribution matrix. The risk area identification and airbag positioning module is used to identify the extreme risk areas in the multidimensional coupled risk distribution matrix, extract the cumulative risk value of the extreme risk areas, and compare the cumulative risk value with a preset dynamic intervention threshold to determine the target airbag group. The body position adjustment control module generates an asymmetric wave-like charge-discharge control command for the target airbag group based on the distribution pattern of the risk extreme value region. The module executes the asymmetric wave-like charge-discharge control command to change the local geometric contour of the intelligent nursing bed support surface until the value of the updated risk extreme value region in the multidimensional coupled risk distribution matrix is ​​lower than the dynamic intervention threshold.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that: When the processor executes the computer program, it implements the steps of the intelligent nursing bed pressure ulcer risk assessment and automatic body position adjustment method according to any one of claims 1 to 7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by the processor, it implements the steps of the intelligent nursing bed pressure ulcer risk assessment and automatic body position adjustment method according to any one of claims 1 to 7.

Images