Pressure injury prediction method and system based on multi-source detection

Abstract

The application discloses a pressure injury prediction method and system based on multi-source detection, and relates to the field of data collection and analysis, which comprises the following steps: collecting multi-source data, and performing data preprocessing and space-time alignment; performing organizational biomechanics modeling based on the preprocessed data to obtain a cell metabolic heat production rate model; solving the cell metabolic heat production rate based on the cell metabolic heat production rate model, and extracting cell metabolic characteristics; and obtaining a pressure injury state based on the cell metabolic characteristics; the application is based on physical and physiological principles, directly monitors the process in which cells are changing from aerobic metabolism to anaerobic metabolism by monitoring the change of metabolic heat production rate, has extremely high individualization degree and specific adaptability compared with the scheme of predicting pressure injury only by pressure distribution, and realizes real early warning of pressure injury.

Description

Technical Field

[0001] This application relates to the field of data acquisition and analysis, and in particular to a method and system for predicting pressure damage based on multi-source detection. Background Technology

[0002] Pressure injuries, also known as pressure ulcers, are injuries, ulcers, or necrosis caused by prolonged pressure on local tissues, leading to impaired blood circulation, ischemia, hypoxia, and malnutrition of the skin and subcutaneous tissues. They primarily occur in people with limited mobility, those who are bedridden for extended periods, or those who use wheelchairs. The pathogenesis involves the combined effects of pressure, shear force, and friction.

[0003] Pressure injuries are difficult to treat once they occur, so the core objective is "prevention first, early detection, and early intervention." While traditional manual examinations are necessary, they are prone to omissions, especially in high-risk patients. Existing non-manual pressure injury detection methods often use mattresses equipped with pressure sensor arrays to detect the patient's pressure distribution and time, thereby determining whether the patient is at risk of pressure injury. Predicting pressure injury solely through pressure distribution, while an important starting point, has inherent and fundamental limitations that restrict its accuracy and clinical value, such as: It fails to reflect the individual's physiological vulnerability: the pressure distribution only measures the external load, but completely ignores the patient's internal tolerance capacity. The exact same pressure distribution and time applied to different patients will result in vastly different results. 2. Insensitive to shear force: Shear force is one of the main culprits of tissue damage. It causes tissue ischemia more easily than vertical pressure by twisting and stretching blood vessels. Most pressure distribution pads are primarily designed to measure vertical interface pressure. Direct measurement of shear force is very difficult and inaccurate. When a patient slides down in a semi-recumbent position, the sacrum and coccyx may experience enormous shear force, but the vertical pressure may not necessarily be high. The pressure distribution system may mistakenly classify this as a low-risk situation.

[0004] 3. Inability to detect deep tissue damage: Pressure injuries, especially the most dangerous types, often occur from the inside out. For example, deep muscle tissue may become ischemic and necrotic first due to greater stress (especially shear force), but the surface skin may appear completely normal or only show purple / maroon discoloration. Surface pressure sensors mainly measure vertical pressure and are inaccurate in estimating shear stress in deep tissues, let alone directly measuring the physiological state of deep tissues. Summary of the Invention

[0005] To address the aforementioned shortcomings in existing technologies, this invention provides a method and system for predicting pressure injuries based on multi-source detection, which solves the problems that pressure injury detection methods based on pressure distribution cannot reflect individual physiological vulnerability, are insensitive to shear force, and cannot detect deep tissue damage.

[0006] In order to achieve the above-mentioned objective, in a first aspect, the present invention provides a method for predicting pressure injury based on multi-source detection, comprising: collecting multi-source data and performing data preprocessing and spatiotemporal alignment, wherein the multi-source data includes pressure distribution data, temperature data and blood flow data; Based on the preprocessed data, tissue biomechanical modeling was performed to obtain a cell metabolic heat production rate model. Based on the cell metabolic heat production rate model, the cell metabolic heat production rate is solved, and cell metabolic features are extracted; the cell metabolic features include metabolic hotspot features, metabolic-blood flow coupling features, tissue vitality features, and dynamic recovery capacity features. Based on cellular metabolic characteristics, the state of stress injury is obtained.

[0007] Preferably, the method for data preprocessing includes: Use a low-pass filter to remove high-frequency noise from the pressure distribution data; Temperature data is processed using a moving average filter. Blood flow data is processed using a bandpass filter to preserve physiologically relevant frequency components. Identify and remove abnormal data points caused by sensor malfunctions or motion artifacts.

[0008] Preferably, the method for spatiotemporal alignment includes adding a unified timestamp to the pressure distribution data, temperature data, and blood flow data; Establish a coordinate system mapping relationship between pressure distribution data and corresponding human body positions.

[0009] Preferably, the process of constructing the cellular metabolic heat production rate model includes: Based on the preprocessed pressure distribution data and the tissue layering data of human skin, fat and muscle, the stress distribution inside human tissue is obtained by finite element analysis. By correcting the stress distribution using preprocessed blood flow data, a blood perfusion rate model is obtained. Individualized calibration of the thermophysical parameters of blood perfusion rate yields a cell metabolic heat production rate model.

[0010] Preferably, the blood perfusion rate model is expressed as: in, For the corrected blood perfusion rate, Baseline blood perfusion rate, The critical normal stress, This is the influence coefficient of normal stress. This is the influence coefficient of shear stress. The critical shear stress. It is normal stress. This is shear stress.

[0011] Preferably, the cell metabolic heat production rate model is expressed as follows: in, For tissue density, To organize specific heat capacity, Thermal conductivity, Blood density, For the specific heat capacity of blood, Arterial blood temperature, For the rate of heat production from cell metabolism, Indicates the temperature distribution within the tissue. For time parameters, This is the gradient operator.

[0012] Preferably, the boundary conditions of the cell metabolic heat production rate model are set based on the corrected temperature data, and the cell metabolic heat production rate is solved inversely, as follows: in, Indicates minimization. Let be the objective function. This indicates the corrected temperature data. For regularization parameters, For regularization terms, For the internal temperature distribution of the tissue, Let be the geometric space of the tissue region to be analyzed. Preferably, the final cellular metabolic heat production rate is obtained by solving for the cellular metabolic heat production rate, and the cellular metabolic features include: metabolic hotspot features. : Metabolic-blood flow coupling characteristics : Organizational vitality characteristics : Dynamic recovery capability characteristics : in, Metabolic heat production rate The mean, Metabolic heat production rate standard deviation For time step, It is the rate of heat production from basal metabolism.

[0013] Preferably, the stress injury state is obtained by weighted calculation of metabolic hotspot characteristics, metabolic-blood flow coupling characteristics, tissue vitality characteristics, and dynamic recovery capacity characteristics.

[0014] In a second aspect, the present invention also provides a pressure injury prediction system based on multi-source detection for performing the method of the first aspect, comprising: The data acquisition unit is used to collect multi-source data, including pressure distribution data, temperature data, and blood flow data. The data preprocessing unit is used to preprocess and spatiotemporally align the collected multi-source data; The cell metabolic heat production rate model is constructed by performing tissue biomechanical modeling on preprocessed data, and is used to extract and obtain cell metabolic characteristics. A pressure injury state prediction unit is used to obtain the pressure injury state based on cellular metabolic characteristics.

[0015] The beneficial effects of this invention are as follows: Based on physical and physiological principles, this method directly monitors the process of cells transitioning from aerobic to anaerobic metabolism by monitoring changes in metabolic heat production rate. Compared to methods that predict stress injury solely based on pressure distribution, this method offers a high degree of individualization and specificity. Attached Figure Description

[0016] Figure 1 This is a flowchart of a pressure damage prediction method based on multi-source detection provided in one or more embodiments of the present invention; Figure 2 This is a diagram illustrating the process of cells undergoing continuous stress from aerobic metabolism to death, as provided in one or more embodiments of the present invention. Detailed Implementation

[0017] The specific embodiments of the present invention are described below to enable those skilled in the art to understand the present invention. However, it should be understood that the present invention is not limited to the scope of the specific embodiments. For those skilled in the art, various changes are obvious as long as they are within the spirit and scope of the present invention as defined and determined by the appended claims. All inventions utilizing the concept of the present invention are protected.

[0018] like Figure 1 As shown, this embodiment provides a method for predicting pressure injuries based on multi-source detection, including: collecting multi-source data and performing data preprocessing and spatiotemporal alignment, wherein the multi-source data includes pressure distribution data, temperature data, and blood flow data; Based on the preprocessed data, tissue biomechanical modeling was performed to obtain a cell metabolic heat production rate model. Based on the cell metabolic heat production rate model, the cell metabolic heat production rate is solved, and cell metabolic features are extracted; the cell metabolic features include metabolic hotspot features, metabolic-blood flow coupling features, tissue vitality features, and dynamic recovery capacity features. Based on cellular metabolic characteristics, the state of stress injury is obtained.

[0019] In some embodiments of the present invention, pressure distribution data can be obtained through a mattress based on a pressure sensor array; temperature data can be obtained by deploying digital temperature sensors at key anatomical locations (e.g., sacrum, ischium, heel, etc.); and blood flow data can be obtained through a laser Doppler blood flow monitor or through a blood oxygen saturation probe.

[0020] In the description of the embodiments of this application, "data preprocessing" means applying a series of operations or transformations to the original data to make it meet the requirements of subsequent analysis, modeling or application; for example, such processing may include, but is not limited to, data optimization through data cleaning, normalization and feature extraction, data simplification through data reduction and sampling, or data compatibility achieved through format conversion and encoding.

[0021] like Figure 2 As shown, this illustrates the entire process of cell metabolism from aerobic metabolism to cell death. Under aerobic metabolism, cells are in a normal state, and the metabolic products are water and carbon dioxide. Under anaerobic metabolism, cells are in a state of ischemia and hypoxia, and the metabolic products also include lactic acid. When anaerobic metabolism continues for a certain period of time, the end product of anaerobic metabolism, lactic acid, cannot be transported away in time, and its concentration in the tissue increases sharply, leading to acidosis. Due to the sharp decrease in energy production, in order to maintain basic life activities, cells will: significantly increase the rate of glucose consumption and rapidly deplete intracellular energy reserves, resulting in a decrease in metabolic heat production rate; at the same time, hypoxia and the accumulation of metabolites will activate inflammasomes, attracting inflammatory cell infiltration, leading to vasodilation and hypermetabolism, and an increase in surface temperature; finally, cell death may lead to cell swelling, disintegration, leakage of contents, and triggering more severe inflammation. Therefore, the cellular changes we can detect throughout the process are shown in Table 1. Anaerobic metabolism Increased heat production acidosis tissue pH decrease Energy depletion Decreased metabolic heat production rate Inflammatory response Local temperature rise Cell death The metabolic heat production rate decreased, and the local temperature increased. Therefore, by monitoring tissue thermal changes (i.e. metabolic heat production rate) and blood flow, this application is actually indirectly monitoring its deep metabolic state, thereby achieving a very early warning of stress injury before irreversible cell damage occurs.

[0022] As a preferred embodiment of the present invention, the method for performing data preprocessing includes: Use a low-pass filter to remove high-frequency noise from the pressure distribution data; Temperature data is processed using a moving average filter. Blood flow data is processed using a bandpass filter to preserve physiologically relevant frequency components. Abnormal data points caused by sensor malfunctions or motion artifacts can be identified and removed using the Z-Score method or time series processing.

[0023] In the description of the embodiments of this application, "spatiotemporal alignment" means establishing a consistent correspondence between data from different sources in terms of timestamps and spatial coordinates to eliminate inconsistencies caused by differences in collection time or location. For example, such alignment may include, but is not limited to, time axis unification through timestamp synchronization and interpolation compensation, spatial location mapping achieved through coordinate system transformation and projection matching, or joint calibration achieved through a spatiotemporal correlation model.

[0024] As a preferred approach, in some embodiments of the present invention, the method for performing spatiotemporal alignment includes: Add a unified timestamp to pressure distribution data, temperature data, and blood flow data; Establish a coordinate system mapping relationship between pressure distribution data and corresponding human body positions.

[0025] As one possible implementation, the corresponding human body location can be determined by embedding a standard human anatomical structure model; preferably, it can also be corrected by factors such as the patient's age and body mass index.

[0026] After obtaining the necessary information mentioned above, a cell metabolic heat production rate model can be constructed. In some embodiments of the present invention, the process of constructing the cell metabolic heat production rate model includes: Based on the preprocessed pressure distribution data and the tissue layering data of human skin, fat and muscle, the stress distribution inside human tissue is obtained by finite element analysis. By correcting the stress distribution using preprocessed blood flow data, a blood perfusion rate model is obtained. Individualized calibration of the thermophysical parameters of blood perfusion rate yields a cell metabolic heat production rate model.

[0027] In some embodiments of this application, tissue layering data of human skin, fat and muscle can be obtained by using a portable high-frequency ultrasound device to scan key high-risk areas such as the sacrum, ischial tuberosity and heel of the patient, and to measure and record the thickness of the skin, fat and muscle at each point.

[0028] In the description of the embodiments of this application, "individualized calibration of thermophysical parameters" means that for a specific biological individual, key attribute parameters in the heat transfer process are determined and corrected in a personalized manner through a series of specialized measurement and estimation methods. For example, such calibration may include, but is not limited to: volumetric measurements using the transient planar heat source method to determine the thermal conductivity and volumetric heat capacity of tissues; measurements under no-pressure conditions to determine the basal metabolic heat production rate; measurements in unpressurized areas to calibrate the basal blood perfusion rate; and estimations based on environmental conditions to determine the surface convective heat transfer coefficient. Whether the calibration is completed directly through the above measurement methods in one go, or indirectly iteratively approximated through intermediate models or historical data, it falls within the scope of "individualized calibration of thermophysical parameters" as defined in this application, but it does not necessarily mean that all listed parameters must be calibrated simultaneously or in full.

[0029] The blood perfusion rate model and the cell metabolic heat production rate model obtained through the above processing are respectively expressed as follows: in, For the corrected blood perfusion rate, Baseline blood perfusion rate, The critical normal stress, This is the influence coefficient of normal stress. This is the influence coefficient of shear stress. The critical shear stress. It is normal stress. For shear stress; and: in, For tissue density, To organize specific heat capacity, Thermal conductivity, Blood density, For the specific heat capacity of blood, Arterial blood temperature, For the rate of heat production from cell metabolism, Indicates the temperature distribution within the tissue. For time parameters, This is the gradient operator. Boundary conditions for the cell metabolic heat production rate model are set based on the corrected temperature data, and the cell metabolic heat production rate is solved inversely, expressed as: in, Indicates minimization. Let be the objective function. This indicates the corrected temperature data. For regularization parameters, For regularization terms, For the internal temperature distribution of the tissue, The geometric space of the tissue region to be analyzed is defined. Next, the final cellular metabolic heat production rate is obtained by solving for the cellular metabolic heat production rate, and the cellular metabolic features include: metabolic hotspot features. : Metabolic-blood flow coupling characteristics : Organizational vitality characteristics : Dynamic recovery capability characteristics : in, Metabolic heat production rate The mean, Metabolic heat production rate standard deviation For time step, It is the rate of heat production from basal metabolism.

[0030] Preferably, in some embodiments of the present invention, in order to match with cellular metabolic processes, the risk of stress injury is first classified according to the cellular metabolic state before judging stress injury, for example: metabolic warning period (corresponding to clinical phase 0): An increase of 50-100%, MBC > 1.5, TVI = 0.5-0.8; at this point, the monitoring frequency can be increased; hypermetabolic phase (corresponding to clinical phase 1): An increase of 100-200% indicates a metabolic hotspot, with TVI = 0.3-0.5; at this point, targeted decompression can be implemented; metabolic transition period (corresponding to phase 2 clinical trials): A sharp decline from the peak, with MBC < 0.5 and TVI = 0.1-0.3, requires emergency medical intervention; metabolic failure stage (corresponding to clinical phase 3 or above): < 30% of baseline, MRR ≈ 0, TVI < 0.1; at this point, tissue necrosis can be confirmed, and a corresponding treatment plan needs to be formulated. In other embodiments of the present invention, the stress injury state is obtained by weighted calculation of metabolic hotspot characteristics, metabolic-blood flow coupling characteristics, tissue viability characteristics, and dynamic recovery capacity characteristics, which can be specifically represented as: in, This is a state of pressure injury. It is used to quantify the non-uniformity or dispersion of metabolic heat production rate in spatial distribution; This represents the standard deviation of metabolic heat production rate at all points within the entire region of interest at a specific time point. This represents the average metabolic heat production rate of all points within the same region.

[0031] In other embodiments of the present invention, the process of obtaining the pressure injury state can be represented as follows: displaying the location of high-risk areas through metabolic hotspot features; performing pathophysiological assessment through metabolic-blood flow coupling features, tissue viability features, and dynamic recovery capacity features; and calculating the pressure injury state by weighting the metabolic-blood flow coupling features and dynamic recovery capacity features. in, This is a state of pressure injury. It is used to quantify the non-uniformity or dispersion of metabolic heat production rate in spatial distribution; This represents the standard deviation of metabolic heat production rate at all points within the entire region of interest at a specific time point. This represents the average metabolic heat production rate of all points within the same region. Preferably, in some embodiments of the present invention, the process of obtaining the stress injury status can be expressed as follows: if metabolic hotspot characteristics are found and tissue viability is <0.5: then the risk baseline score is significantly amplified, for example, by multiplying the calculated value of the stress injury status by a correction factor greater than 1; if tissue viability is <0.3: then the highest level alarm is directly triggered; if none of the above conditions are met, then the stress injury status is calculated using the following formula: in, This is a state of pressure injury. It is used to quantify the non-uniformity or dispersion of metabolic heat production rate in spatial distribution; This represents the standard deviation of metabolic heat production rate at all points within the entire region of interest at a specific time point. This represents the average metabolic heat production rate of all points within the same region.

[0032] After obtaining the pressure injury status value, the current pressure injury risk can be determined by combining it with the pressure injury risk classification.

[0033] To implement the pressure injury prediction method based on multi-source detection provided by the present invention, an embodiment of the present invention also provides a system for performing the method, including: a data acquisition unit for acquiring multi-source data including pressure distribution data, temperature data, and blood flow data; a data preprocessing unit for preprocessing and spatiotemporally aligning the acquired multi-source data; and a cell metabolic heat production rate model for constructing tissue biomechanical modeling based on the preprocessed data to extract and obtain cell metabolic characteristics. A pressure injury state prediction unit is used to obtain the pressure injury state based on cellular metabolic characteristics.

[0034] The above embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application.

Claims

1. A method for predicting pressure damage based on multi-source detection, characterized in that, include: Collect multi-source data and perform data preprocessing and spatiotemporal alignment. The multi-source data includes pressure distribution data, temperature data, and blood flow data. Based on the preprocessed data, tissue biomechanical modeling was performed to obtain a cell metabolic heat production rate model. Based on the cell metabolic heat production rate model, the cell metabolic heat production rate is solved, and cell metabolic features are extracted; the cell metabolic features include metabolic hotspot features, metabolic-blood flow coupling features, tissue vitality features, and dynamic recovery capacity features. Based on cellular metabolic characteristics, the state of stress injury is obtained.

2. The method for predicting pressure damage based on multi-source detection according to claim 1, characterized in that, Data preprocessing methods include: Use a low-pass filter to remove high-frequency noise from the pressure distribution data; Temperature data is processed using a moving average filter. Blood flow data is processed using a bandpass filter to preserve physiologically relevant frequency components. Identify and remove abnormal data points caused by sensor malfunctions or motion artifacts.

3. The method for predicting pressure damage based on multi-source detection according to claim 1, characterized in that, Methods for spatiotemporal alignment include: Add a unified timestamp to pressure distribution data, temperature data, and blood flow data; Establish a coordinate system mapping relationship between pressure distribution data and corresponding human body positions.

4. The method for predicting pressure damage based on multi-source detection according to claim 1, characterized in that, The process of constructing a cellular metabolic heat production rate model includes: Based on the preprocessed pressure distribution data and the tissue layering data of human skin, fat and muscle, the stress distribution inside human tissue is obtained by finite element analysis. By correcting the stress distribution using preprocessed blood flow data, a blood perfusion rate model is obtained. Individualized calibration of the thermophysical parameters of blood perfusion rate yields a cell metabolic heat production rate model.

5. The method for predicting pressure damage based on multi-source detection according to claim 4, characterized in that, The blood perfusion rate model is expressed as follows: in, For the corrected blood perfusion rate, Baseline blood perfusion rate, The critical normal stress, This is the influence coefficient of normal stress. This is the influence coefficient of shear stress. The critical shear stress. It is normal stress. This is shear stress.

6. The method for predicting pressure damage based on multi-source detection according to claim 5, characterized in that, The cell metabolic heat production rate model is expressed as follows: in, For tissue density, To organize specific heat capacity, Thermal conductivity, Blood density, For the specific heat capacity of blood, Arterial blood temperature, For the rate of heat production from cell metabolism, Indicates the temperature distribution within the tissue. For time parameters, This is the gradient operator.

7. The pressure damage prediction method based on multi-source detection according to claim 6, characterized in that, The boundary conditions for the cell metabolic heat production rate model are set based on the corrected temperature data, and the cell metabolic heat production rate is solved inversely, as follows: in, Indicates minimization. Let be the objective function. This indicates the corrected temperature data. For regularization parameters, For regularization terms, For the internal temperature distribution of the tissue, The geometric space of the tissue region to be analyzed.

8. The method for predicting pressure damage based on multi-source detection according to claim 7, characterized in that, The final cellular metabolic heat production rate obtained by solving the problem is used to extract cellular metabolic features, which include: metabolic hotspot features. : Metabolic-blood flow coupling characteristics : Organizational vitality characteristics : Dynamic recovery capability characteristics : in, Metabolic heat production rate The mean, Metabolic heat production rate standard deviation For time step, It is the rate of heat production from basal metabolism.

9. The method for predicting pressure damage based on multi-source detection according to claim 1, characterized in that, The stress injury state is obtained by weighted calculation of metabolic hotspot characteristics, metabolic-blood flow coupling characteristics, tissue vitality characteristics, and dynamic recovery capacity characteristics.

10. A pressure damage prediction system based on multi-source detection, characterized in that, For performing the method according to any one of claims 1-9, comprising: The data acquisition unit is used to collect multi-source data, including pressure distribution data, temperature data, and blood flow data. The data preprocessing unit is used to preprocess and spatiotemporally align the collected multi-source data; The cell metabolic heat production rate model is constructed by performing tissue biomechanical modeling on preprocessed data, and is used to extract and obtain cell metabolic characteristics. A pressure injury state prediction unit is used to obtain the pressure injury state based on cellular metabolic characteristics.

Images