Pressure sore risk dynamic assessment and mattress pressure regulation system

Abstract

The invention provides a pressure sore risk dynamic assessment method and a mattress pressure adjusting system, and belongs to the technical field of medical instruments and intelligent nursing. The system comprises a multi-physical field sensing layer, a data preprocessing and feature extraction layer, a dynamic risk fusion calculation layer, a grading early warning and decision-making layer and a closed loop execution layer. The sensing layer is integrated with multiple types of sensors and an electronic medical record interface, and collects mechanics, physiology and microenvironment parameters of a pressed area and basic information of a patient in real time; the preprocessing layer is used for extracting characteristic parameters after processing the data, fitting a nutrition correction coefficient and calculating a temperature and humidity stimulation coefficient; the fusion calculation layer maps a multi-dimensional parameter into a dynamic risk index (DRI) through a weighted fusion algorithm model, and supports machine learning nonlinear mapping; the early warning decision layer executes graded early warning according to the DRI numerical value, the change trend and the deep tissue damage precursor mode, and identifies dominant risk factors; and the closed-loop execution layer executes pathogenesis targeted intervention according to the early warning level and the dominant risk factor.

Description

Technical Field

[0001] This invention belongs to the field of medical device and intelligent nursing technology, and more specifically, it relates to a dynamic assessment of pressure ulcer risk and a mattress pressure adjustment system. Background Technology

[0002] Pressure ulcers, also known as pressure injuries, are one of the most common complications in patients who are bedridden for a long time. Their formation is related to multiple factors such as vertical pressure, shear force, tissue tolerance, microenvironment climate, and cumulative mechanical load time, which seriously affect the patient's recovery process and increase the clinical nursing burden.

[0003] Current pressure ulcer risk assessment methods primarily rely on healthcare professionals periodically using tools such as the Braden and Norton scales for subjective scoring. This assessment method is detached from the specific pathophysiological mechanisms of pressure ulcer formation, relying solely on clinical experience to make macroscopic judgments, resulting in significant lag and an inability to capture the dynamic evolution of a patient's pressure ulcer risk in real time. In recent years, while some pressure-reducing mattresses have integrated pressure monitoring functions, their monitoring dimensions are limited. They can only trigger local decompression operations based on a single vertical pressure threshold (such as exceeding 32 mmHg), failing to simultaneously perceive key pathogenic factors of pressure ulcer formation such as shear force, tissue ischemia and hypoxia, and temperature and humidity of the contact surface. Furthermore, they lack the ability to integrate and analyze multidimensional monitoring data, making it impossible to achieve dynamic calculation and predictive intervention of pressure ulcer risk.

[0004] The core shortcomings of existing technologies are concentrated in four aspects: First, the perception dimension is singular, ignoring key factors such as shear force, friction force, tissue perfusion, and microenvironment temperature and humidity, forming a monitoring blind spot; second, risk assessment is lagging, and manual assessment at discrete time points cannot reflect real-time changes in risk; third, the adjustment strategy is blind, and decompression behavior is disconnected from the etiology of pressure ulcers. Simply reducing the average pressure by alternating inflation and deflation cannot specifically alleviate shear force or improve local blood supply; fourth, there is insufficient monitoring of deep tissue injury (DTI). Deep tissue injury mostly occurs in subcutaneous soft tissue, and surface pressure often does not show obvious abnormalities, making it difficult for existing technologies to achieve early identification and warning.

[0005] In summary, there is an urgent clinical need for a pressure ulcer risk assessment and regulation system that can achieve multi-dimensional real-time monitoring, dynamic quantitative evaluation, and etiological-targeted intervention to address the many shortcomings of existing technologies and improve the clinical effectiveness of pressure ulcer prevention and control. Summary of the Invention

[0006] The purpose of this application is to address the technical problems in existing pressure ulcer prevention and control technologies, such as limited perception dimensions, lagging risk assessment, blind adjustment strategies, and blind spots in monitoring deep tissue damage, which prevent the dynamic quantification of pressure ulcer risk and targeted and precise intervention for the causes of ulcers.

[0007] To achieve the above objectives, this application provides the following technical solution:

[0008] A dynamic assessment method for pressure ulcer risk and a mattress pressure adjustment system include a multi-physics sensing layer, a data preprocessing and feature extraction layer, a dynamic risk fusion calculation layer, a graded early warning and decision-making layer, and a closed-loop execution layer.

[0009] The multi-physical sensing layer is disposed in the flexible sensing area of ​​the pressure surface of the medical mattress. The multi-physical sensing layer is used to collect data and transmit it to the data preprocessing and feature extraction layer. The multi-physical sensing layer includes several sensors and interface units.

[0010] The sensors include a high-density flexible pressure sensor array, a two-dimensional shear force sensing unit, a multispectral tissue blood oxygen sensor, and a digital temperature and humidity sensor.

[0011] The interface unit includes a time counter and an electronic medical record interface;

[0012] The data preprocessing and feature extraction layer employs an embedded microprocessor unit to receive the raw data acquired by the multi-physics sensing layer and to complete data preprocessing, feature parameter extraction, and correction / stimulation coefficient calculation.

[0013] The dynamic risk fusion calculation layer is used to calculate the dynamic risk index based on the data output by the data preprocessing and feature extraction layer, and to normalize the calculated dynamic risk index.

[0014] The hierarchical early warning and decision-making layer conducts risk warnings based on a dynamic risk index and identifies dominant risk factors, providing a basis for intervention decisions for the closed-loop execution layer, while transmitting the early warning information to the nursing station terminal.

[0015] The closed-loop execution layer performs targeted and precise intervention operations based on the early warning levels and dominant risk factors of the hierarchical early warning and decision-making layers.

[0016] Preferably, the high-density flexible pressure sensor array uses capacitive flexible pressure sensors, with an arrangement density of 2 sensors / cm². 2 The detection range is 0-100 mmHg, and the detection accuracy is ±1 mmHg. It covers the pressure area of ​​the mattress in a grid pattern. The two-dimensional shear force sensing unit adopts a piezoelectric shear force sensor. The two-dimensional shear force sensing unit is arranged in the same area as the high-density flexible pressure sensor array. The multispectral tissue oxygenation sensor adopts photoplethysmography technology and emits 660nm / 940nm dual-wavelength light to monitor the blood oxygen saturation and blood perfusion status of the subcutaneous tissue at the pressure site in real time.

[0017] The digital temperature and humidity sensor is used to collect the real-time temperature and relative humidity of the contact surface between the patient and the mattress; the time technology is used to record the cumulative duration of mechanical load and the duration of adverse microclimate; the electronic medical record interface adopts the HL7FHIR standard protocol to interface with the hospital's electronic medical record system.

[0018] Preferably, the sampling frequency of each unit in the multi-physical sensing layer is 10 Hz, the clock synchronization accuracy is controlled within 10ms, and the data collected by all sensing units is transmitted to the data preprocessing and feature extraction layer via Bluetooth 5.0 wireless communication, with a data transmission rate ≥1 Mbps and a transmission distance ≤10 m.

[0019] Preferably, the data preprocessing and feature extraction layer processing steps are as follows:

[0020] S1: Raw data preprocessing:

[0021] The raw analog signals collected by the multiphysics sensing layer are filtered, denoised, and interpolated to eliminate invalid data caused by environmental interference and sensor drift, and to fill in missing sampling points to obtain standardized digital signals.

[0022] S2: Feature parameter extraction:

[0023] From the preprocessed data, core characteristic parameters for pressure ulcer risk assessment were extracted, including: mechanical characteristics, physiological characteristics, microenvironmental characteristics, and time characteristics.

[0024] S3: Calculation of Correction Factor and Stimulus Factor

[0025] Based on the patient's basic information retrieved from electronic medical records, a nutritional correction coefficient was fitted; based on temperature and humidity monitoring data, a temperature and humidity stimulation coefficient was calculated. Both coefficients were normalized to the 0-1 range. The specific calculation formula is as follows:

[0026] Nutritional correction factor = 0.2 × serum albumin level (g / L) + 0.3 × hemoglobin concentration (g / L) / 10; if the patient has diabetes, the factor is multiplied by 0.8 for correction; if the patient has edema (≥ grade 2), the factor is multiplied by 0.7 for correction.

[0027] Temperature and humidity stimulation coefficient = 0.4 × (actual temperature - 32℃) + 0.6 × (actual humidity - 60%) / 10; when temperature ≤ 32℃ and humidity ≤ 60%RH, the corresponding item is 0, and the coefficient is 0 overall; positive stimulation value is calculated only when temperature / humidity exceeds the threshold, reflecting the degree of harm of the microenvironment.

[0028] Preferably, the dynamic risk fusion calculation layer is used to calculate the dynamic risk index (DRI). The basic algorithm for the DRI calculation method is an explicit weighted fusion formula, and all input parameters are normalized to the 0-1 range to ensure the consistency of the calculation results. The formula is as follows:

[0029] DRI = [(Integral of vertical pressure × 0.6) + (Shear intensity × 0.4)] × Cumulative duration ÷ [(Tissue oxygen saturation × 0.7 + Perfusion index × 0.3) × Nutritional correction factor] + Temperature and humidity stimulation factor

[0030] In the above formula, "vertical pressure integral × 0.6 + shear force intensity × 0.4" represents the equivalent mechanical stress, which reflects the core mechanical factors in the formation of pressure ulcers; "tissue oxygen saturation × 0.7 + blood perfusion index × 0.3" represents the tissue perfusion status, which reflects the degree of tissue ischemia and hypoxia at the pressure site; "cumulative duration" is the time-dimensional weighted average of the mechanical load; "nutritional correction coefficient" is the personalized correction for individual physiological state; and "temperature and humidity stimulation coefficient" is an additional risk factor of the microenvironment.

[0031] Preferably, the dynamic risk fusion calculation layer also employs machine learning methods, using large clinical sample data (including multi-physics field perception data and pressure ulcer occurrence outcome data of patients with different weights, underlying diseases, and bedridden durations) as the training set, and using the pressure ulcer occurrence probability as the label to train a nonlinear mapping model;

[0032] After the model is deployed, it can receive feature parameters and output DRI in real time. It also supports incremental learning, continuously optimizing model parameters based on actual clinical data to improve the accuracy of risk prediction.

[0033] Preferably, the graded early warning and decision-making layer, based on the DRI output by the dynamic risk fusion calculation layer, combined with the DRI change trend and the early warning pattern of deep tissue damage, executes a secondary risk warning. The secondary risk warning determination rules are as follows:

[0034] (1) Primary warning: The warning level is yellow if any of the following conditions are met, indicating a low to moderate risk of pressure ulcers and requiring enhanced monitoring:

[0035] ① The Dynamic Risk Index (DRI) remains at ≥2.0 (the baseline level is 1.0, which means it exceeds the baseline level by 1.5 times) and lasts for ≥5 minutes.

[0036] ② Pressure-time integral (PTI) ≥ 32 mmHg·2h (threshold for reversible tissue damage), i.e., vertical pressure ≥ 32 mmHg and duration ≥ 2h;

[0037] (2) Advanced warning: The warning level is red when any of the following conditions are met, indicating a high risk of pressure ulcers and requiring immediate intervention:

[0038] ① Identify early signs of deep tissue damage: shear force ≥ 20N and local tissue oxygen saturation < 40% (even if surface pressure < 32mmHg);

[0039] ② Abnormal DRI first derivative analysis: The first derivative of DRI is calculated every 5 minutes. The slope is ≥0.1 for three consecutive sampling periods (15min), which means that DRI is showing a rapid upward trend and the risk is continuously deteriorating.

[0040] ③ If the DRI is ≥3.5 and lasts for ≥3 minutes, the risk of pressure ulcers is high.

[0041] Once an alert is triggered, the system identifies the dominant risk factors for pressure ulcer formation through feature parameter contribution rate analysis. These dominant risk factors are categorized into five types, which can exist individually or in combination:

[0042] Excessive vertical pressure: peak pressure ≥ 60 mmHg or pressure-time integral ≥ 32 mmHg·2h, contribution rate ≥ 40%;

[0043] Excessive shear force: Shear force intensity ≥ 20 N or shear force-time integral ≥ 10 N·min, contribution rate ≥ 40%;

[0044] Tissue ischemia: tissue oxygen saturation <60% or perfusion index <0.5, contribution rate ≥40%;

[0045] Humid microenvironment: humidity ≥ 70%RH and duration ≥ 30 min, or temperature and humidity stimulation coefficient ≥ 0.5, contribution rate ≥ 30%;

[0046] High-risk compound: Two or more of the above risk factors exceed the standard at the same time, with a single factor contributing ≥20%.

[0047] Preferably, the closed-loop execution layer includes an airbag array pressure regulation module, an electric backplate angle adjustment module, a microenvironment adjustment module, and a nursing terminal prompt module.

[0048] Preferably, the airbag array pressure regulating module includes an airbag array with 16 independent zones located inside the mattress. The single airbag pressure adjustment range is 5-40 mmHg, and the adjustment accuracy is ±0.5 mmHg. Each zone can be adjusted independently, synchronously, or with gradient pressure, corresponding to different pressure points of the patient.

[0049] Preferably, the specific intervention instructions, parameters, and execution steps of the closed-loop execution layer are as follows:

[0050] The specific intervention instructions, parameters, and execution steps are as follows:

[0051] Intervention instructions for excessive vertical pressure are: control the local decompression of the corresponding area airbag to 15-20 mmHg, or perform alternating inflation and deflation operations of 5 minutes inflation and 3 minutes deflation;

[0052] The intervention command for excessive shear force is: automatically reduce the headboard angle to ≤30° and adjust the airbag array to form a gradient support structure with a gradient of 5mmHg from the feet to the head to counteract the downward trend of the torso's gravity.

[0053] The intervention instructions for tissue ischemia are: to perform high-frequency micro-motion intervention with a micro-motion amplitude of ±5mm and a frequency of 10 times / minute on the ischemic area to interrupt the continuous ischemic process and prevent reperfusion injury;

[0054] The intervention instructions for the humid microenvironment are as follows: activate the miniature fan and semiconductor cooling module to adjust the local temperature and humidity of the pressure area to 28-32℃ and 40%-60%RH, and at the same time push a reminder message to the nursing terminal to change the bed sheet;

[0055] The intervention instructions for complex high-risk cases are as follows: comprehensively execute all intervention operations corresponding to a single risk factor, and push comprehensive prompt information including risk factor type, intervention measures, and nursing recommendations to the nursing terminal, and simultaneously update it to the hospital's electronic medical record system.

[0056] Compared with the prior art, this application has at least the following beneficial effects:

[0057] The system provided in this application integrates a pressure / shear force composite sensor array, a tissue oxygenation sensor, and a temperature and humidity sensor. It simultaneously quantifies the four essential conditions for pressure ulcer formation—mechanical load, tissue perfusion, microenvironment climate, and time accumulation—on the same platform. For the first time, it achieves online real-time monitoring of shear force and local tissue oxygenation, effectively filling the monitoring blind spot of deep tissue injury (DTI), providing a reliable early warning method for deep tissue injury, and solving the core problem of the single sensing dimension in existing technologies.

[0058] The system proposed in this application can accurately assess the dynamics of risk, adapt to individual differences, and abandon traditional subjective scoring scales. It constructs a dynamic risk fusion model based on the pathophysiological mechanism of pressure ulcers, combining real-time collected tissue mechanics and physiological parameters with basic information such as the patient's serum albumin, hemoglobin, history of diabetes, and edema level. It fits personalized nutritional correction coefficients and calculates a continuously output dynamic risk index (DRI). At the same time, it supports machine learning methods such as fuzzy logic and random forest to optimize the model, so that the risk index can truly and in real time reflect the individual pressure ulcer risk status of different patients, avoiding the lag and one-sidedness of the assessment.

[0059] The system proposed in this application enables precise targeted intervention strategies, improving intervention efficiency. The system can accurately identify the dominant risk factors for pressure ulcer formation (excessive vertical pressure, excessive shear force, tissue ischemia, moist microenvironment, or a combination of high risks), and perform etiological-level targeted interventions for different risk factors, such as local airbag decompression, gradient support to counteract shear force, high-frequency micro-motion to improve tissue perfusion, and active ventilation to regulate the microenvironment. This avoids the drawbacks of blind decompression in existing technologies, significantly improving the effectiveness and precision of intervention, while reducing unnecessary frequent repositioning that interferes with the patient's sleep and recovery process.

[0060] The system proposed in this application achieves closed-loop intelligent control, reducing the nursing burden. It constructs a closed-loop intelligent control system covering the entire process of "monitoring-pretreatment-risk calculation-tiered early warning-precise intervention-effect feedback," with each level working collaboratively to automate and intelligently control pressure ulcer risks. It eliminates the need for real-time manual intervention by medical staff, effectively reducing the workload of clinical nurses and improving nursing efficiency. At the same time, it records all monitoring data, intervention operations, and effects in real time, providing objective and quantitative evidence for the dynamic adjustment of clinical pressure ulcer prevention and control plans.

[0061] The system provided in this application has good compatibility and adaptability. It is easy to implement in clinical settings and is compatible with standard hospital beds of 90-120cm. The sensors are encapsulated in medical-grade silicone to meet clinical disinfection and usage requirements. The electronic medical record interface supports mainstream medical data protocols and can seamlessly connect with electronic medical record systems of different brands of hospitals. The modules use standardized wireless communication methods, eliminating the need for large-scale modifications to existing clinical nursing equipment and providing excellent conditions for clinical promotion and implementation.

[0062] The system provided in this application has significant clinical prevention and control effects, improving patient prognosis. In addition, this invention can effectively block high-risk clinical processes such as deep tissue damage and reperfusion injury, reduce the probability of pressure ulcer occurrence from the source, significantly reduce the incidence of hospital-acquired pressure ulcers, reduce the risk of complications caused by pressure ulcers, improve the clinical prognosis of long-term bedridden patients, and reduce the medical cost of pressure ulcer treatment. It has high clinical application value and social value. Attached Figure Description

[0063] Figure 1 This is a schematic diagram of the method flow of the present invention. Detailed Implementation

[0064] Please see Figure 1 This invention provides a dynamic assessment system for pressure ulcer risk and a mattress pressure adjustment system. The core of the system is to achieve real-time monitoring of the pathological mechanism of pressure ulcer risk, dynamic quantitative assessment, and targeted closed-loop intervention for the cause. This solves the technical defects of existing technologies, such as single perception dimension, lagging risk assessment, blind adjustment strategies, and blind spots in monitoring deep tissue damage.

[0065] The pressure ulcer risk dynamic assessment and mattress pressure adjustment system comprises a five-layer architecture: a multi-physics sensing layer, a data preprocessing and feature extraction layer, a dynamic risk fusion calculation layer, a hierarchical early warning and decision-making layer, and a closed-loop execution layer. These layers work collaboratively to achieve full automation and intelligence throughout the entire process from multi-dimensional data collection to risk intervention. The system described below, in conjunction with the accompanying drawings, provides a detailed explanation:

[0066] I. Specific configuration of hardware and functions at each layer of the system:

[0067] (a) Multiphysics sensing layer:

[0068] The multi-physical sensing layer is the core of the system's data acquisition. It is set in the flexible sensing area on the pressure-bearing surface of the medical mattress. In one embodiment, the multi-physical sensing layer is adapted to a standard hospital bed with a width of 90-120 cm, and all sensors are encapsulated in medical-grade silicone with an IP67 protection rating, capable of withstanding medical alcohol wiping and disinfection.

[0069] Specifically, it includes the following sensing units and interfaces:

[0070] High-density flexible pressure sensor array: Employs capacitive flexible pressure sensors with an arrangement density of 2 sensors / cm². 2 With a detection range of 0-100 mmHg and a detection accuracy of ±1 mmHg, it covers the pressure area of ​​the mattress in a grid pattern and is used to collect the vertical pressure distribution, peak pressure, average pressure and 90th percentile pressure of the part of the patient's body in contact with the mattress in real time, accurately capturing the pressure load differences of different body parts.

[0071] Two-dimensional shear force sensing unit: It adopts a piezoelectric shear force sensor with a detection range of 0-50N and a detection accuracy of ±0.5N. It is arranged in the same area as the pressure sensor array and is used to collect the tangential shear force value and shear force distribution between the patient's body and the mattress, filling the gap in existing shear force monitoring technology.

[0072] Multispectral tissue oxygenation sensor: Employing photoplethysmography technology, it emits dual-wavelength light of 660nm / 940nm to detect local tissue oxygen saturation (StO2), with a range of 0-100% and an accuracy of ±2%. The perfusion index (PI) ranges from 0.02-20 and has an accuracy of ±0.01. It is used to monitor the blood oxygen saturation and perfusion status of subcutaneous tissue at pressure sites in real time, reflecting the degree of tissue ischemia and hypoxia, and enabling early monitoring of deep tissue damage.

[0073] Digital temperature and humidity sensor: temperature detection range 20-40℃, accuracy ±0.2℃; humidity detection range 30%-95%RH, accuracy ±2%RH. It is placed in the breathable areas of each zone of the mattress to collect real-time temperature and relative humidity of the patient's contact surface with the mattress, and to quantify the impact of the microenvironment climate on pressure ulcer formation.

[0074] Time counter: Linked with each sensing unit, it is used to record the cumulative duration of mechanical load and the duration of adverse microclimate (temperature > 32℃ / humidity > 60%RH), thereby quantifying risk factors in the time dimension.

[0075] Electronic Medical Record Interface: It adopts the HL7FHIR standard protocol to interface with the hospital's electronic medical record system, with a data transmission rate of ≥1Mbps. It is used to read the patient's basic clinical information in real time, including serum albumin level (g / L), hemoglobin concentration (g / L), history of diabetes, and edema grade (1-4), which serve as individual nutritional correction factors for subsequent risk calculations to achieve personalized risk assessment.

[0076] The sampling frequency of each unit in the multi-physical sensing layer is 10 Hz, the clock synchronization accuracy is controlled within 10 ms, and the data collected by all sensing units is transmitted to the data preprocessing and feature extraction layer via Bluetooth 5.0 wireless communication. The data transmission rate is ≥1 Mbps and the transmission distance is ≤10 m, ensuring the real-time performance and stability of data transmission.

[0077] (II) Data Preprocessing and Feature Extraction Layer:

[0078] The data preprocessing and feature extraction layer uses an embedded microprocessor unit (MCU is STM32H743) to receive the raw data collected by the multiphysics sensing layer and complete data preprocessing, feature parameter extraction and correction coefficient / stimulation coefficient calculation through built-in algorithms.

[0079] The data preprocessing and feature extraction layer processes are as follows:

[0080] S1: Raw data preprocessing:

[0081] The original analog signals such as pressure, shear force, blood oxygen, temperature and humidity are processed by filtering (5th order Butterworth low-pass filtering), denoising (median filtering method), and interpolation (linear interpolation method) to eliminate invalid data caused by environmental interference and sensor drift, and to fill in missing sampling points to obtain standardized digital signals.

[0082] S2: Feature parameter extraction:

[0083] From the preprocessed data, the core feature parameters for pressure ulcer risk assessment were extracted, including:

[0084] Mechanical characteristics: peak pressure, average pressure, 90th percentile pressure, shear strength, pressure-time integral (PTI, in mmHg·min), shear-time integral (STI, in N·min);

[0085] Physiological characteristics: local tissue oxygen saturation (StO2), perfusion index (PI);

[0086] Microenvironment characteristics: average temperature and humidity of the contact surface, duration of adverse microclimate;

[0087] Time characteristics: cumulative duration of mechanical load, duration of continuous exceedance of a single risk factor.

[0088] S3: Calculation of Correction Factor and Stimulus Factor

[0089] Based on the patient's basic information retrieved from electronic medical records, a nutritional correction coefficient was fitted; based on temperature and humidity monitoring data, a temperature and humidity stimulation coefficient was calculated. Both coefficients were normalized to the 0-1 range. The specific calculation formula is as follows:

[0090] Nutritional correction factor = 0.2 × serum albumin level (g / L) + 0.3 × hemoglobin concentration (g / L) / 10; if the patient has diabetes, the factor is multiplied by 0.8 for correction; if the patient has edema (≥ grade 2), the factor is multiplied by 0.7 for correction.

[0091] Temperature and humidity stimulation coefficient = 0.4 × (actual temperature - 32℃) + 0.6 × (actual humidity - 60%) / 10; when temperature ≤ 32℃ and humidity ≤ 60%RH, the corresponding item is 0, and the coefficient is 0 overall; positive stimulation value is calculated only when temperature / humidity exceeds the threshold, reflecting the degree of harm of the microenvironment.

[0092] The feature parameters, nutrition correction coefficient, and temperature and humidity stimulation coefficients processed by the data preprocessing and feature extraction layers are transmitted in real time to the dynamic risk fusion calculation layer as input data for risk index calculation.

[0093] (III) Dynamic Risk Fusion Calculation Layer:

[0094] The dynamic risk fusion calculation layer is the core layer of the system. It adopts an embedded artificial intelligence chip (NPU is RK1808), and has a built-in weighted fusion algorithm model based on the pathophysiological mechanism of pressure ulcers. It also supports machine learning methods such as fuzzy logic, random forest, and long short-term memory network (LSTM) (nonlinear mapping function is obtained by training with large clinical sample data). It maps the input multi-dimensional feature parameters into a continuously output dynamic risk index (DRI) to realize the quantification and dynamic assessment of pressure ulcer risk.

[0095] In one embodiment, the Dynamic Risk Index (DRI) is calculated as follows:

[0096] The basic algorithm is an explicit weighted fusion formula, where all input parameters are normalized to the 0-1 range to ensure consistency of the calculation results. The formula is as follows:

[0097] DRI = [(Integral of vertical pressure × 0.6) + (Shear intensity × 0.4)] × Cumulative duration ÷ [(Tissue oxygen saturation × 0.7 + Perfusion index × 0.3) × Nutritional correction factor] + Temperature and humidity stimulation factor

[0098] In the above formula, "vertical pressure integral × 0.6 + shear force intensity × 0.4" represents the equivalent mechanical stress, which reflects the core mechanical factors in the formation of pressure ulcers; "tissue oxygen saturation × 0.7 + blood perfusion index × 0.3" represents the tissue perfusion status, which reflects the degree of tissue ischemia and hypoxia at the pressure site; "cumulative duration" is the time-dimensional weighted average of the mechanical load; "nutritional correction coefficient" is the personalized correction for individual physiological state; and "temperature and humidity stimulation coefficient" is an additional risk factor of the microenvironment.

[0099] In one embodiment, the dynamic risk fusion calculation layer also employs machine learning methods, using large clinical sample data (including multi-physics field perception data and pressure ulcer occurrence outcome data of patients with different weights, underlying diseases, and bedridden durations) as the training set, and using the pressure ulcer occurrence probability as the label to train a nonlinear mapping model;

[0100] After the model is deployed, it can receive feature parameters and output DRI in real time. It also supports incremental learning, continuously optimizing model parameters based on actual clinical data to improve the accuracy of risk prediction.

[0101] After the DRI is calculated, it needs to be normalized to the 0-5 range. The higher the value, the higher the risk of pressure ulcers, providing a quantitative basis for subsequent graded early warning.

[0102] (iv) Tiered early warning and decision-making level:

[0103] The tiered early warning and decision-making layer is based on the DRI output by the dynamic risk fusion calculation layer. It combines the DRI change trend and the early warning pattern of deep tissue damage to execute a secondary risk warning. It also identifies the dominant risk factors for the current pressure ulcer formation through a built-in algorithm, providing a basis for intervention decision-making for the closed-loop execution layer. The early warning information is simultaneously pushed to the nursing station terminal. The early warning signal includes audible and visual warnings and text prompts. The nursing terminal can view the patient's DRI change curve, risk factor type and raw sensor data in real time.

[0104] Level 2 Risk Warning Judgment Rules: The warning judgment combines the absolute value of DRI, single parameter threshold, DRI change trend, and precursors of deep tissue damage to achieve early detection and early warning. The specific rules are as follows:

[0105] (1) Primary warning: The warning level is yellow if any of the following conditions are met, indicating a low to moderate risk of pressure ulcers and requiring enhanced monitoring:

[0106] ① The Dynamic Risk Index (DRI) remains at ≥2.0 (the baseline level is 1.0, which means it exceeds the baseline level by 1.5 times) and lasts for ≥5 minutes.

[0107] ② Pressure-time integral (PTI) ≥ 32 mmHg·2h (threshold for reversible tissue damage), i.e., vertical pressure ≥ 32 mmHg and duration ≥ 2h.

[0108] (2) Advanced warning: The warning level is red when any of the following conditions are met, indicating a high risk of pressure ulcers and requiring immediate intervention:

[0109] ① Identify early signs of deep tissue damage: shear force ≥ 20N and local tissue oxygen saturation < 40% (even if surface pressure < 32mmHg);

[0110] ② Abnormal DRI first derivative analysis: The first derivative of DRI is calculated every 5 minutes. The slope is ≥0.1 for three consecutive sampling periods (15min), which means that DRI is showing a rapid upward trend and the risk is continuously deteriorating.

[0111] ③ If the DRI is ≥3.5 and lasts for ≥3 minutes, the risk of pressure ulcers is considered high.

[0112] Once an alert is triggered, the system identifies the dominant risk factors for pressure ulcer formation through feature parameter contribution rate analysis. These dominant risk factors are categorized into five types, which can exist individually or in combination:

[0113] Excessive vertical pressure: peak pressure ≥ 60 mmHg or pressure-time integral ≥ 32 mmHg·2h, contribution rate ≥ 40%;

[0114] Excessive shear force: Shear force intensity ≥ 20 N or shear force-time integral ≥ 10 N·min, contribution rate ≥ 40%;

[0115] Tissue ischemia: tissue oxygen saturation <60% or perfusion index <0.5, contribution rate ≥40%;

[0116] Humid microenvironment: humidity ≥ 70%RH and duration ≥ 30 min, or temperature and humidity stimulation coefficient ≥ 0.5, contribution rate ≥ 30%;

[0117] High-risk compound: Two or more of the above risk factors exceed the standard at the same time, with a single factor contributing ≥20%.

[0118] (v) Closed-loop execution layer:

[0119] The closed-loop execution layer is the core of the system's intervention execution. This layer includes an airbag array pressure adjustment module, a motorized backrest angle adjustment module, a microenvironment adjustment module, and a nursing terminal prompt module. Based on the warning levels and dominant risk factors of the graded warning and decision-making layers, the closed-loop execution layer performs targeted and precise intervention operations. After intervention, it collects mattress status and patient parameters in real time and feeds them back to the data preprocessing and feature extraction layer to update the DRI, forming a closed-loop control of monitoring-assessment-early warning-intervention-feedback. This avoids blind pressure reduction and improves intervention efficiency. The airbag array pressure adjustment module in this layer is the core execution unit. The mattress has a built-in 16-zone airbag array, with a single airbag pressure adjustment range of 5-40 mmHg and an adjustment accuracy of ±0.5 mmHg. Each zone can be adjusted independently, synchronously, or with gradients, corresponding to different pressure points such as the patient's head, shoulders, back, waist, buttocks, thighs, calves, and feet.

[0120] In summary, this invention, through the collaborative operation of a five-layer architecture, achieves a technological leap in pressure ulcer risk assessment and intervention, and compared with existing technologies, it possesses the following core advantages:

[0121] Comprehensive Perception Dimensions: Simultaneously quantifies the four essential conditions for pressure ulcer formation: mechanical load, tissue perfusion, microenvironment, and time accumulation, filling the technological gap in online monitoring of shear force and tissue blood oxygenation, and enabling early warning of deep tissue damage;

[0122] Dynamic and precise risk assessment: The traditional macro-static scale scoring is upgraded to a dynamic risk index (DRI) based on real-time tissue mechanics and physiological parameters, combined with individual nutritional correction factors, to achieve personalized, pathological mechanism-level risk quantification, and significantly improve prediction accuracy.

[0123] The intervention strategy is precise and efficient: based on the dominant risk factors, the intervention is targeted at the underlying cause, avoiding blind stress reduction, and reducing unnecessary frequent changes in body position that may interfere with the patient's sleep and recovery.

[0124] Closed-loop intelligent control: It realizes full-process automation from data collection, risk assessment, graded early warning to precise intervention and effect feedback, reduces the workload of medical staff, improves nursing efficiency, and can significantly reduce the incidence of hospital-acquired pressure ulcers.

[0125] II. Targeted intervention strategies for different dominant risk factors:

[0126] The core of the closed-loop execution layer is etiological intervention, which involves executing corresponding intervention operations based on the identified dominant risk factors. The specific intervention instructions, parameters, and execution steps for each risk factor are as follows:

[0127] (a) The dominant risk factor is excessive vertical pressure:

[0128] Intervention objective: To rapidly reduce vertical pressure at the compressed site, achieve pressure redistribution, and prevent ischemia caused by prolonged tissue pressure;

[0129] Specific intervention instructions: The system instructs the airbag in the corresponding pressure area to perform a two-option pressure adjustment operation, which is automatically determined by the system based on the patient's weight.

[0130] Local decompression: Directly adjust the corresponding airbag pressure to 15-20 mmHg (safe pressure range), and continue to decompress until DRI < 1.5;

[0131] Alternating inflation and deflation: Perform alternating inflation (pressure 30-35 mmHg) for 5 minutes and deflation (pressure 15-20 mmHg) for 3 minutes until DRI < 1.5. This method is suitable for patients weighing > 80 kg and avoids patient positional shift caused by single decompression.

[0132] (ii) The dominant risk factor is excessive shear stress:

[0133] Intervention goal: To counteract the downward trend of the patient's torso and reduce the tangential shear force between the body and the mattress;

[0134] Specific intervention instructions:

[0135] The electric backrest angle is automatically adjusted to reduce the headboard angle to ≤30° (when the original angle is >30°), reducing the gravitational force that causes the torso to slide down.

[0136] The airbag array is adjusted to form a gradient support from the feet to the head, with a gradient difference of 5 mmHg, i.e., the pressure of the foot airbag > calf > thigh > hip > waist > back. The gradient support counteracts the tendency of the torso to slide down until the shear force is <10N.

[0137] (iii) The dominant risk factor is tissue ischemia:

[0138] Intervention goals: to interrupt the continuous ischemic process of the compressed tissue, promote the restoration of local blood supply, prevent reperfusion injury, and avoid disturbing the patient's sleep;

[0139] Specific intervention instructions: Perform high-frequency micro-motion intervention on the airbag in the ischemic area. The specific parameters are: micro-motion amplitude ±5mm and micro-motion frequency 10 times / minute. That is, the airbag is adjusted at a frequency of 10 times / minute with a height of ±5mm based on the original pressure to achieve slight traction of local tissue, promote blood perfusion, until the tissue blood oxygen saturation is ≥60% and PI≥0.5.

[0140] (iv) The dominant risk factor is a humid microenvironment:

[0141] Intervention goal: To rapidly reduce the humidity of the surface in contact with the mattress, adjust it to a suitable temperature and humidity range, and improve the microenvironment;

[0142] Specific intervention instructions:

[0143] Activate the mattress's built-in 5W micro fan and semiconductor cooling module. Adjust the fan speed to medium (1500r / min) and the cooling module temperature to 28℃ to achieve active ventilation and low-temperature drying of the contact surface.

[0144] Send a text message to the nursing terminal to remind nursing staff to change damp sheets in a timely manner;

[0145] Continue intervention until the temperature and humidity of the contact surface reach a suitable range of 28-32℃ and 40%-60%RH, and the temperature and humidity stimulation coefficient drops to 0.

[0146] (v) The dominant risk factor is a complex high-risk factor:

[0147] Intervention goal: To comprehensively alleviate various risk factors, achieve multi-dimensional and simultaneous intervention, and rapidly reduce the overall risk of pressure ulcers;

[0148] Specific intervention instructions:

[0149] All intervention operations corresponding to a single risk factor are carried out in a comprehensive manner, with each operation performed simultaneously, and the intervention priority is adjusted according to the contribution rate of the risk factor (factors with higher contribution rates are intervened first).

[0150] Push comprehensive nursing recommendations to the nursing terminal, including all current risk factor types, intervention measures, and key points of manual nursing care;

[0151] During the intervention, each characteristic parameter is monitored in real time. When a certain risk factor reaches the target, the corresponding intervention operation is stopped until DRI < 1.0, at which point all interventions are terminated.

[0152] III. Specific Implementation Examples in Different Clinical Scenarios:

[0153] To further illustrate the technical solution of the present invention, three specific embodiments are given in combination with patients of different weights and underlying diseases commonly seen in clinical practice. In each embodiment, the system parameters and intervention strategies are adjusted according to the individual patient conditions. Those skilled in the art can reproduce the technical effects of the present invention based on these embodiments.

[0154] Example 1: A 50kg bedridden patient (no underlying diseases, serum albumin 40 g / L, hemoglobin 120 g / L):

[0155] Patient basic information: Female, 65 years old, bedridden after femoral neck fracture surgery, no diabetes or edema, good nutritional status, nutritional correction factor = 0.2×40 + 0.3×120 / 10 = 8 + 3.6 = 11.6 (normalized to 1.0).

[0156] Monitoring data: Peak pressure in the pressure area of ​​the patient's buttocks was 70 mmHg, pressure-time integral was 35 mmHg·2h, shear force was 15N, tissue oxygen saturation was 75%, humidity was 65%RH, and calculated DRI value was 2.2.

[0157] Warning determination: DRI=2.2≥2.0, triggering a primary warning (yellow);

[0158] Dominant risk factor: Excessive vertical pressure (contribution rate 45%)

[0159] Intervention strategy: Perform local decompression of the hip airbag, adjusting the airbag pressure to 18 mmHg;

[0160] Intervention effect: 5 minutes after intervention, the peak pressure in the buttocks dropped to 25 mmHg, the DRI dropped to 1.2, the initial warning was lifted, and the system resumed continuous monitoring.

[0161] Example 2: An 80kg diabetic bedridden patient (type 2 diabetes, serum albumin 30g / L, hemoglobin 100g / L, grade 1 edema):

[0162] Patient basic information: Male, 70 years old, bedridden due to cerebral infarction, type 2 diabetes, grade 1 edema, fair nutritional status, nutritional correction coefficient = (0.2×30+0.3×100 / 10)×0.8 (diabetes correction) = 6+3×0.8=7.2 (normalized to 0.6).

[0163] Monitoring data: Shear force intensity in the pressure area of ​​the patient's sacrum and coccyx was 25N, tissue oxygen saturation was 38%, peak pressure was 28mmHg, calculated DRI value was 3.8, and the slope of the first derivative of DRI for three consecutive cycles was 0.12.

[0164] Warning criteria: If the shear force is ≥20N and the blood oxygen is <40%, a high-level warning (red) is triggered.

[0165] Dominant risk factors: excessive shear force + tissue ischemia (compound high risk, each contributing 40%).

[0166] Intervention strategies: ① Reduce the headboard angle from 45° to 25°, adjust the airbag array to form a 5mmHg gradient support, and reduce the shear force to 8N; ② Perform high-frequency micro-motion intervention (±5mm, 10 times / minute) on the sacral and coccygeal airbags, and the blood oxygen level rises to 65%;

[0167] Intervention effect: 8 minutes after the intervention, the DRI dropped to 1.1, the advanced warning was lifted, and the system resumed continuous monitoring.

[0168] Example 3: A 100kg bedridden patient with severe edema (grade 3 edema, serum albumin 25g / L, hemoglobin 90g / L):

[0169] Patient basic information: Male, 68 years old, bedridden due to heart failure, grade 3 edema, no diabetes, poor nutritional status, nutritional correction coefficient = (0.2×25+0.3×90 / 10)×0.7 (edema correction) = 5+2.7×0.7=5.39 (normalized to 0.5).

[0170] Monitoring data: Peak pressure in the pressure area on the patient's back was 80 mmHg, humidity was 75%RH, duration was 40 min, tissue oxygen saturation was 55%, and DRI calculated value was 3.0;

[0171] Warning determination: DRI=3.0≥2.0, triggering a primary warning (yellow);

[0172] The dominant risk factors are: excessive vertical pressure, humid microenvironment, and tissue ischemia (compound high risk, with contribution rates of 35%, 30%, and 25%, respectively).

[0173] Intervention strategy: ① The back airbag is alternately inflated and deflated (5 minutes of inflation to 32 mmHg / 3 minutes of deflation to 18 mmHg); ② A miniature fan and semiconductor cooling are activated to reduce the humidity to 55%RH; ③ The back airbag is simultaneously subjected to high-frequency micro-motion intervention (±5mm, 10 times / minute).

[0174] Intervention effect: 10 minutes after intervention, the peak pressure dropped to 30 mmHg, blood oxygen rose to 62%, DRI dropped to 1.3, the primary warning was lifted, and the system resumed continuous monitoring.

[0175] IV. System Compatibility and Adaptability:

[0176] The system described in this invention has good clinical compatibility and adaptability, and can be directly applied to existing hospital nursing scenarios without the need for large-scale modifications to existing equipment. Specifically:

[0177] The system provided in this application can be adapted to standard hospital beds with a width of 90-120cm, and the airbag array can be finely adjusted in zones according to the size of the hospital bed;

[0178] The electronic medical record interface in this application supports mainstream medical data protocols such as HL7FHIR and DICOM, and can be connected with electronic medical record systems of different brands of hospitals to realize real-time reading and updating of patients' basic information.

[0179] The system provided in this application uses Bluetooth 5.0 wireless communication between its modules, and also supports wired Ethernet communication (as a backup) to meet the communication environment requirements of different wards;

[0180] In addition, the warning information can be pushed to the nurse station computer, nursing handheld terminal (PDA), and hospital nursing management platform, so that it can be viewed simultaneously on multiple terminals.

Claims

1. A method for dynamic assessment of pressure ulcer risk and a mattress pressure adjustment system, characterized in that: It includes a multiphysics sensing layer, a data preprocessing and feature extraction layer, a dynamic risk fusion calculation layer, a hierarchical early warning and decision-making layer, and a closed-loop execution layer. The multi-physical sensing layer is disposed in the flexible sensing area of ​​the pressure surface of the medical mattress. The multi-physical sensing layer is used to collect data and transmit it to the data preprocessing and feature extraction layer. The multi-physical sensing layer includes several sensors and interface units. The sensors include a high-density flexible pressure sensor array, a two-dimensional shear force sensing unit, a multispectral tissue blood oxygen sensor, and a digital temperature and humidity sensor. The interface unit includes a time counter and an electronic medical record interface; The data preprocessing and feature extraction layer employs an embedded microprocessor unit to receive the raw data acquired by the multi-physics sensing layer and to complete data preprocessing, feature parameter extraction, and correction / stimulation coefficient calculation. The dynamic risk fusion calculation layer is used to calculate the dynamic risk index based on the data output by the data preprocessing and feature extraction layer, and to normalize the calculated dynamic risk index. The hierarchical early warning and decision-making layer conducts risk warnings based on a dynamic risk index and identifies dominant risk factors, providing a basis for intervention decisions for the closed-loop execution layer, while transmitting the early warning information to the nursing station terminal. The closed-loop execution layer performs targeted and precise intervention operations based on the early warning levels and dominant risk factors of the hierarchical early warning and decision-making layers.

2. The method for dynamic assessment of pressure ulcer risk and the mattress pressure adjustment system according to claim 1, characterized in that: The high-density flexible pressure sensor array employs capacitive flexible pressure sensors, with an arrangement density of 2 sensors / cm². 2 The detection range is 0-100 mmHg, and the detection accuracy is ±1 mmHg. It covers the pressure area of ​​the mattress in a grid pattern. The two-dimensional shear force sensing unit adopts a piezoelectric shear force sensor. The two-dimensional shear force sensing unit is arranged in the same area as the high-density flexible pressure sensor array. The multispectral tissue oxygenation sensor adopts photoplethysmography technology and emits 660nm / 940nm dual-wavelength light to monitor the blood oxygen saturation and blood perfusion status of the subcutaneous tissue at the pressure site in real time. The digital temperature and humidity sensor is used to collect the real-time temperature and relative humidity of the contact surface between the patient and the mattress; the time technology is used to record the cumulative duration of mechanical load and the duration of adverse microclimate; the electronic medical record interface adopts the HL7FHIR standard protocol to interface with the hospital's electronic medical record system.

3. The method for dynamic assessment of pressure ulcer risk and the mattress pressure adjustment system according to claim 2, characterized in that: The sampling frequency of each unit in the multi-physical sensing layer is 10 Hz, the clock synchronization accuracy is controlled within 10 ms, and the data collected by all sensing units is transmitted to the data preprocessing and feature extraction layer via Bluetooth 5.0 wireless communication with a data transmission rate of ≥1 Mbps and a transmission distance of ≤10 m.

4. The method for dynamic assessment of pressure ulcer risk and the mattress pressure adjustment system according to claim 1, characterized in that: The data preprocessing and feature extraction layer processes are as follows: S1: Raw data preprocessing: The raw analog signals collected by the multiphysics sensing layer are filtered, denoised, and interpolated to eliminate invalid data caused by environmental interference and sensor drift, and to fill in missing sampling points to obtain standardized digital signals. S2 Feature parameter extraction: From the preprocessed data, core characteristic parameters for pressure ulcer risk assessment were extracted, including: mechanical characteristics, physiological characteristics, microenvironmental characteristics, and time characteristics. S3: Calculation of Correction Factor and Stimulus Factor Based on the patient's basic information retrieved from electronic medical records, a nutritional correction coefficient was fitted; based on temperature and humidity monitoring data, a temperature and humidity stimulation coefficient was calculated. Both coefficients were normalized to the 0-1 range. The specific calculation formula is as follows: Nutritional correction factor = 0.2 × serum albumin level (g / L) + 0.3 × hemoglobin concentration (g / L) / 10; if the patient has diabetes, the factor is multiplied by 0.8 for correction; if the patient has edema (≥ grade 2), the factor is multiplied by 0.7 for correction. Temperature and humidity stimulation coefficient = 0.4 × (actual temperature - 32℃) + 0.6 × (actual humidity - 60%) / 10; when temperature ≤ 32℃ and humidity ≤ 60%RH, the corresponding item is 0, and the coefficient is 0 overall; positive stimulation value is calculated only when temperature / humidity exceeds the threshold, reflecting the degree of harm of the microenvironment.

5. The method for dynamic assessment of pressure ulcer risk and the mattress pressure adjustment system according to claim 4, characterized in that: The dynamic risk fusion calculation layer is used to calculate the Dynamic Risk Index (DRI). The basic algorithm for the DRI calculation method is an explicit weighted fusion formula. All input parameters are normalized to the 0-1 range to ensure the consistency of the calculation results. The formula is as follows: DRI = [(Vertical pressure integral × 0.6) + (Shear force intensity × 0.4)] × Cumulative duration ÷ [(Tissue oxygen saturation × 0.7 + Perfusion index × 0.3) × Nutritional correction factor] + Temperature and humidity stimulation factor In the above formula, "vertical pressure integral × 0.6 + shear force intensity × 0.4" is the equivalent mechanical stress, which is used to reflect the core mechanical factors of pressure ulcer formation; "tissue oxygen saturation × 0.7 + blood perfusion index × 0.3" is the tissue perfusion state, which is used to reflect the degree of tissue ischemia and hypoxia at the pressure site; "cumulative duration" is the time dimension weighted mechanical load; "nutritional correction coefficient" is the personalized correction of individual physiological state; and "temperature and humidity stimulation coefficient" is the additional risk factor of the microenvironment.

6. The method for dynamic assessment of pressure ulcer risk and the mattress pressure adjustment system according to claim 5, characterized in that: The dynamic risk fusion calculation layer also employs machine learning methods, using large clinical sample data (including multi-physical field perception data and pressure ulcer occurrence outcome data of patients with different weights, underlying diseases, and bed rest durations) as the training set, and using the pressure ulcer occurrence probability as the label to train a nonlinear mapping model. After the model is deployed, it can receive feature parameters and output DRI in real time. It also supports incremental learning, continuously optimizing model parameters based on actual clinical data to improve the accuracy of risk prediction.

7. The method for dynamic assessment of pressure ulcer risk and the mattress pressure adjustment system according to claim 6, characterized in that: The tiered early warning and decision-making layer, based on the DRI output by the dynamic risk fusion calculation layer, combined with the DRI change trend and the early warning pattern of deep tissue damage, executes a secondary risk warning. The determination rules for the secondary risk warning are as follows: (1) Primary warning: The warning level is yellow if any of the following conditions are met, indicating a low to moderate risk of pressure ulcers and requiring enhanced monitoring: ① The Dynamic Risk Index (DRI) remains at ≥2.0 (baseline level is 1.0, i.e., more than 1.5 times the baseline level) for a duration of ≥5 minutes; ② Pressure-time integral (PTI) ≥ 32 mmHg·2h (threshold for reversible tissue damage), i.e., vertical pressure ≥ 32 mmHg and duration ≥ 2h; (2) Advanced warning: The warning level is red when any of the following conditions are met, indicating a high risk of pressure ulcers and requiring immediate intervention: ① Identify early signs of deep tissue damage: shear force ≥ 20N and local tissue oxygen saturation < 40% (even if surface pressure < 32mmHg); ② Abnormal DRI first derivative analysis: The first derivative of DRI is calculated every 5 minutes. The slope is ≥0.1 for three consecutive sampling periods (15min), which means that DRI is showing a rapid upward trend and the risk is continuously deteriorating. ③ If the DRI is ≥3.5 and lasts for ≥3 minutes, the risk of pressure ulcers is high. Once an alert is triggered, the system identifies the dominant risk factors for pressure ulcer formation through feature parameter contribution rate analysis. These dominant risk factors are categorized into five types, which can exist individually or in combination: Excessive vertical pressure: peak pressure ≥ 60 mmHg or pressure-time integral ≥ 32 mmHg·2h, contribution rate ≥ 40%; Excessive shear force: Shear force intensity ≥ 20 N or shear force-time integral ≥ 10 N·min, contribution rate ≥ 40%; Tissue ischemia: tissue oxygen saturation <60% or perfusion index <0.5, contribution rate ≥40%; Humid microenvironment: humidity ≥ 70%RH and duration ≥ 30 min, or temperature and humidity stimulation coefficient ≥ 0.5, contribution rate ≥ 30%; High-risk compound: Two or more of the above risk factors exceed the standard at the same time, with a single factor contributing ≥20%.

8. The method for dynamic assessment of pressure ulcer risk and the mattress pressure adjustment system according to claim 7, characterized in that: The closed-loop execution layer includes an airbag array pressure regulation module, an electric backplate angle adjustment module, a microenvironment adjustment module, and a nursing terminal prompt module.

9. The method for dynamic assessment of pressure ulcer risk and the mattress pressure adjustment system according to claim 8, characterized in that: The airbag array pressure adjustment module includes an airbag array with 16 independent zones located inside the mattress. The pressure adjustment range of a single airbag is 5-40 mmHg, with an adjustment accuracy of ±0.5 mmHg. Each zone can be adjusted independently, synchronously, or with gradient pressure to correspond to different pressure points on the patient.

10. The method for dynamic assessment of pressure ulcer risk and the mattress pressure adjustment system according to claim 9, characterized in that: The specific intervention commands, parameters, and execution steps of the closed-loop execution layer are as follows: The specific intervention instructions, parameters, and execution steps are as follows: Intervention instructions for excessive vertical pressure are: control the local decompression of the corresponding area airbag to 15-20 mmHg, or perform alternating inflation and deflation operations of 5 minutes inflation and 3 minutes deflation; The intervention command for excessive shear force is: automatically reduce the headboard angle to ≤30° and adjust the airbag array to form a gradient support structure with a gradient of 5mmHg from the feet to the head to counteract the downward trend of the torso's gravity. The intervention instructions for tissue ischemia are: to perform high-frequency micro-motion intervention with a micro-motion amplitude of ±5mm and a frequency of 10 times / minute on the ischemic area to interrupt the continuous ischemic process and prevent reperfusion injury; The intervention instructions for the humid microenvironment are as follows: activate the miniature fan and semiconductor cooling module to adjust the local temperature and humidity of the pressure area to 28-32℃ and 40%-60%RH, and at the same time push a reminder message to the nursing terminal to change the bed sheet; The intervention instructions for complex high-risk cases are as follows: comprehensively execute all intervention operations corresponding to a single risk factor, and push comprehensive prompt information including risk factor type, intervention measures, and nursing recommendations to the nursing terminal, and simultaneously update it to the hospital's electronic medical record system.

Images