Diabetic foot risk real-time monitoring system and method for wearable device

Abstract

The invention relates to the technical field of diabetic foot monitoring, in particular to a diabetic foot risk real-time monitoring system and method for wearable equipment. Foot microenvironment and whole body physiological data are synchronously collected through a wearable multi-mode sensor group, and a non-linear damage assessment model is constructed in combination with a sensor self-inspection module; quantitative early warning of sensor aging is realized by fusing sensor performance data and a material attenuation index. According to the mechanism, the data distortion risk caused by sensor recessive aging can be recognized in advance, the reliability of diabetic foot risk assessment is remarkably improved, and missing detection of early-stage lesions due to equipment performance degradation is avoided; aiming at the operation pain point of the elderly patient, the system realizes multi-mode alarm through a local intelligent interaction module, and forms a three-level response closed loop of'patient, family member and medical treatment 'through an interaction interface and family member end linkage mechanism, so that the health management efficiency of the elderly living alone is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of diabetic foot monitoring, and specifically to a real-time diabetic foot risk monitoring system and method for wearable devices. Background Art

[0002] Diabetic foot is one of the most serious chronic complications of diabetes. Long-term high blood sugar levels in diabetic patients can lead to lower limb nerve and vascular disease, which can cause foot infections, ulcers, or deep tissue damage. Patients may experience paresthesias in their feet, such as numbness, pain, and burning sensations, as well as dry, cracked, blistered, and ulcerated skin. Severe diabetic foot can lead to foot gangrene and even require amputation, severely impacting patients' quality of life and life expectancy.

[0003] Currently, when conducting real-time monitoring of diabetic foot risks, wearable devices are exposed to the patient's sweat for a long time and are corroded by sweat. This leads to frequent washing and cleaning, which causes sensor aging and performance degradation, resulting in data loss or distortion. In addition, some elderly patients are not familiar with smartphone operations and are unable to respond to sensor aging prompts or replace devices in a timely manner. Summary of the Invention

[0004] The present invention provides a wearable device-based real-time monitoring system and method for diabetic foot risk, which are used to solve the above-mentioned technical problems.

[0005] The first aspect of the present invention provides a real-time monitoring system for diabetic foot risk for wearable devices, including a wearable sensor module, a sensor self-test module, an edge computing evaluation module, and a local intelligent interaction module.

[0006] The wearable sensor module includes a multimodal sensor group deployed in the wearable device according to a pre-designed position. The cloud server controls the multimodal sensor group to collect patient information to obtain patient vital signs data, and controls the sensor to perform self-test to obtain self-test data; and the server transmits the patient vital signs data to the edge computing evaluation module and the self-test data to the sensor self-test module; the patient vital signs data includes the patient's body temperature, pressure, blood oxygen concentration, humidity and motion data at each collection point; the self-test data includes environmental exposure data, sensor performance data and material attenuation data.

[0007] The sensor self-test module is used to receive self-test data and identify the self-test data to obtain environmental exposure data, sensor performance data and material attenuation data, and substitute the environmental exposure data into the exposure damage analysis mechanism to obtain the total exposure damage value; the sensor performance data, material attenuation data and the total exposure damage value are comprehensively evaluated and analyzed to obtain the sensor self-test status, and the cloud server transmits the sensor self-test status to the local intelligent interaction module.

[0008] As a further improvement of the present invention, the specific analysis method of the exposure damage analysis mechanism is as follows:

[0009] The number of machine washing times of each sensor is obtained according to the environmental exposure data and recorded as N. The machine washing damage sensitivity coefficient of each sensor in the database corresponding to the washing resistance experiment is obtained and recorded as α. N ; Substitute the machine wash times and corresponding machine wash damage sensitivity coefficients of each sensor into the preset machine wash damage function Calculate the machine wash damage value S N ; Where β represents a nonlinear index; when the machine wash damage value is greater than a preset threshold, the corresponding machine wash damage value is marked as a machine wash abnormality;

[0010] The sweat immersion time, pH value and chloride concentration corresponding to each sensor are obtained according to the environmental exposure data; the sweat immersion time is divided into multiple immersion time intervals, and each immersion time interval is designed to give a corresponding time influence coefficient. The sweat immersion time of each sensor is matched with each immersion time interval to obtain the corresponding time influence coefficient, and the immersion time and the corresponding time influence coefficient are multiplied to obtain the immersion impact value; the immersion impact value, chloride concentration and pH value are normalized and their values ​​are taken, and the preset chloride concentration damage function is used. The total chloride damage value S was calculated cl ; Where T represents the immersion impact value, cl, cl max Expressed as chloride concentration and upper limit of chloride concentration, α cl It is expressed as the chloride basic damage coefficient; pH is expressed as the pH value, and λ is expressed as the synergistic corrosion coefficient of pH and chloride;

[0011] Based on the database, the acid environment damage coefficient and alkaline environment damage coefficient obtained by fitting the acid and alkali corrosion resistance experiment corresponding to each sensor are obtained; using the formula Calculate the pH damage value S pH ; Wherein, η1 and η2 represent the acid environment damage coefficient and the alkaline environment damage coefficient respectively; m and n are both preset nonlinear exponents;

[0012] Substitute the machine washing damage value, total chloride damage value and pH damage value corresponding to the machine washing abnormality into the preset nonlinear weighted cumulative model Calculate the total exposure damage value S 总 ; where μ i Represents the weight coefficient of each factor, that is, μ N 、μ cl 、μ pH ; Si represents the specific value corresponding to each factor.

[0013] As a further improvement of the present invention, a comprehensive evaluation and analysis of the sensor performance data, material attenuation data, and total exposure damage value is performed. The specific analysis method is as follows:

[0014] Corresponding analysis factors obtained based on sensor performance data include impedance value, signal-to-noise ratio, signal drift, and response time; and corresponding analysis factors obtained based on material attenuation data include substrate bending times and coating thickness;

[0015] Obtain the initial impedance value corresponding to the corresponding impedance value of the sensor, calculate the difference between the current impedance value and the initial impedance value and take the absolute value to obtain the impedance drift value, calculate the ratio of the impedance drift value to the initial impedance value to obtain the impedance drift rate, and when the impedance drift rate is greater than the preset impedance drift threshold, generate the corresponding impedance abnormality signal, and mark the impedance abnormality signal as zk.

[0016] pass The current signal-to-noise ratio damage rate is calculated and the signal-to-noise ratio damage rate is generated. When the signal-to-noise ratio damage rate is greater than the preset signal-to-noise ratio damage threshold, a signal-to-noise ratio abnormality signal is generated and the signal-to-noise ratio abnormality signal is marked as xz. Similarly, the drift damage rate corresponding to the current signal drift of the sensor is obtained according to the signal drift and the corresponding drift tolerance threshold. When the drift damage rate is greater than the preset threshold, a drift abnormality signal is generated and the drift abnormality signal is marked as py.

[0017] Based on the preset information collection cycle, the total number of corresponding sensor responses is obtained, and the corresponding response duration is obtained. When the response duration is greater than the preset upper limit of the response duration, the corresponding number of responses is marked as a timeout response. The number of timeout responses is counted to obtain the number of timeout responses. The timeout response rate is calculated by ratioing the number of timeout responses to the total number of sensor responses corresponding to the corresponding collection cycle. When the timeout response rate is greater than the preset timeout response threshold, a response abnormality signal is generated and marked as yt.

[0018] Obtain the number of substrate bends corresponding to the current sensor and the corresponding maximum bend number. Calculate the difference between the maximum bend number and the current substrate bend number to obtain the remaining bend number. Calculate the ratio of the remaining bend number to the maximum bend number to obtain the remaining bend index. When the remaining bend index is less than the preset remaining bend warning number, a substrate bend abnormality signal is generated and marked as wq.

[0019] The coating thickness corresponding to the current sensor is compared with the initial coating thickness given in advance to obtain the wear thickness, and the wear ratio is calculated by ratioing the wear thickness to the initial coating thickness. When the wear ratio is greater than the preset wear ratio threshold, a thickness abnormality signal is generated and marked as hd.

[0020] The signals corresponding to each analysis factor are aggregated to obtain the total number of factors Y. The signal set corresponding to the analysis factor is (Y zk , Y xz , Y py , Y yt , Y wq , Y hd ), when an impedance abnormal signal, drift abnormal signal, response abnormal signal, substrate bending abnormal signal, or thickness abnormal signal is detected, the corresponding signal set is lit up, and the signal set of the current sensor is identified to obtain the number of lit signals and recorded as the number of abnormal signals Ys; the total number of analysis factors is based on the preset minimum signal number and maximum signal number, and marked as Ymax and Ymin respectively;

[0021] When Ys<Ymin, the corresponding sensing performance state is generated as no performance abnormality;

[0022] When Ymin≤Ys<Ymax, the corresponding sensor performance state is generated as the performance fluctuation stage;

[0023] When Ys≥Ymax, the corresponding sensor performance state is generated as the performance abnormality stage.

[0024] Obtain the exposure damage threshold value given in advance by each sensor. When the total exposure damage value corresponding to the sensor currently exceeds the exposure damage threshold, generate a sensor exposure abnormality signal. At the same time, when the sensor performance status corresponding to the sensor is in the performance abnormality stage, generate a sensor performance abnormality signal. When it is monitored that the sensor has either or both the sensor exposure abnormality signal and the sensor performance abnormality signal, generate a sensor self-test status indicating that the sensor needs to be replaced.

[0025] The edge computing evaluation module is used to receive the patient's vital signs data and perform health assessment and analysis on the vital signs data to obtain a health risk signal. The cloud server transmits the patient's health risk signal to the local intelligent interaction module. Specifically, the patient's vital signs data is identified to obtain the patient's body temperature, pressure, blood oxygen concentration, humidity and motion data at each collection point, and the patient's vital signs data corresponding to each collection point is input into a pre-set diabetic foot risk monitoring model, and the diabetic foot risk monitoring model outputs a diabetic foot risk index. When the diabetic foot risk index is greater than the preset risk threshold, a health risk signal is generated.

[0026] The local intelligent interaction module is used to store the patient's vital signs data and self-test data in the deployed database, and to provide intelligent reminders based on the received sensor self-test status and health risk signals; when the sensor self-test status corresponds to the need to replace the sensor, the corresponding sensor location information and status information will be sent to the patient's mobile terminal for intelligent voice broadcast reminders; when a health risk signal is detected, a multimodal alarm will be triggered to the patient's mobile terminal.

[0027] A second aspect of the present invention provides a real-time monitoring method for diabetic foot risk using a wearable device, comprising the following steps:

[0028] Step 1: Obtain the sensor self-test status corresponding to each sensor and identify it. When it is detected that the sensor self-test status corresponds to a sensor that needs to be replaced, mark the corresponding sensor as a sensor to be replaced and increase the self-test frequency of the sensor to be replaced.

[0029] Step 2: When a sensor to be replaced is detected, a risk warning instruction is generated, and the interaction status of the risk warning instruction is obtained.

[0030] Step 3: Identify the interaction status and obtain the emergency start instruction based on the analysis of the interaction status.

[0031] As a further improvement of the present invention, analysis is performed based on the interaction status, specifically: the interaction status is identified, and when the instruction feedback corresponding to the patient's interaction status is received, the self-inspection frequency of the sensor to be replaced is increased, and the risk prompt frequency is reduced, and the preset prompt time limit is obtained. When the corresponding sensor self-inspection status within the prompt time limit corresponds to no performance abnormality, the risk prompt is lifted, otherwise an emergency start instruction is generated; when the instruction feedback corresponding to the patient's interaction status is no response, the self-inspection frequency of the sensor to be replaced is increased, and the risk prompt frequency is increased, and the preset no-feedback time limit is obtained. When the no-response time exceeds the no-feedback time limit, an emergency start instruction is generated.

[0032] Step 4: When receiving the emergency start command, the cloud server sends the risk warning instruction to the pre-set emergency family contacts and medical care terminals.

[0033] Compared with the prior art, the technical solution provided by the present invention has the following beneficial effects:

[0034] 1. The present invention uses a wearable multimodal sensor group to synchronously collect foot microenvironment and whole-body physiological data, and combines it with a sensor self-test module to construct a nonlinear damage assessment model that includes environmental exposure data such as the number of machine washes, sweat pH, and chloride concentration. Furthermore, by fusing sensor performance data with material attenuation indicators, a quantitative early warning of sensor aging is achieved. This mechanism can identify the risk of data distortion caused by latent sensor aging in advance, significantly improving the reliability of diabetic foot risk assessment and avoiding missed detection of early lesions due to equipment performance degradation.

[0035] 2. The present invention targets the operational pain points of elderly patients. The system implements multimodal alarms through a local intelligent interactive module and a linkage mechanism with family members through an interactive interface. When a sensor needs to be replaced or a high-risk signal is detected, the system first guides the patient to respond through voice. If there is no feedback within the time limit, the system automatically contacts the emergency contact or medical staff, forming a three-level response closed loop of "patient, family, and medical care", thereby improving the health management efficiency of elderly people living alone. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the following is a brief introduction to the drawings required for the description of the embodiments. The following drawings are not intentionally scaled to the actual size, and the focus is on illustrating the main purpose of the present application.

[0037] Figure 1 This is a principle block diagram of the wearable device-oriented diabetic foot risk real-time monitoring system of the present invention;

[0038] Figure 2 This is a flow chart of the method for real-time monitoring of diabetic foot risk based on wearable devices of the present invention. DETAILED DESCRIPTION

[0039] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making any creative efforts shall fall within the scope of protection of the present invention.

[0040] For ease of understanding, the specific process of the embodiment of the present invention is described below. Figure 1-2 In an embodiment of the present invention, an embodiment of a real-time monitoring system for diabetic foot risk for wearable devices includes:

[0041] The wearable sensor module includes a multimodal sensor group deployed in a wearable device according to a pre-designed position. The cloud server controls the multimodal sensor group to collect patient information to obtain patient vital signs data, and controls the sensor to perform self-inspection to obtain self-inspection data; and the server transmits the patient vital signs data to the edge computing evaluation module and transmits the self-inspection data to the sensor self-inspection module.

[0042] Patient vital sign data includes body temperature, pressure, blood oxygen concentration, humidity and movement data at each collection point of the patient; self-test data includes environmental exposure data, sensor performance data and material attenuation data.

[0043] The sensor self-test module receives the self-test data and identifies the self-test data to obtain environmental exposure data, sensor performance data and material attenuation data, and substitutes the environmental exposure data into the exposure damage analysis mechanism to obtain the total exposure damage value; the sensor performance data, material attenuation data and the total exposure damage value are comprehensively evaluated and analyzed to obtain the sensor self-test status, and the cloud server transmits the sensor self-test status to the local intelligent interaction module.

[0044] The specific analysis method of exposure damage analysis mechanism is as follows:

[0045] The number of machine washing times of each sensor is obtained according to the environmental exposure data and recorded as N. The machine washing damage sensitivity coefficient of each sensor in the database corresponding to the washing resistance experiment is obtained and recorded as α. N ; Substitute the machine wash times and corresponding machine wash damage sensitivity coefficients of each sensor into the preset machine wash damage function Calculate the machine wash damage value S N ; Where β>1, which is expressed as a nonlinear index, reflecting the cumulative acceleration effect of multiple machine washings; when the machine washing damage value is greater than the preset threshold, the corresponding machine washing damage value is marked as a machine washing abnormality; for example, the design life of a certain sensor can withstand 50 machine washings (Nmax=50). When the patient has been machine washed 30 times in total (N=30), the formula is substituted to calculate SN=0.45, which exceeds the threshold of 0.4 and is marked as "machine washing abnormality".

[0046] The sweat immersion time, pH value and chloride concentration corresponding to each sensor are obtained according to the environmental exposure data; the sweat immersion time is divided into multiple immersion time intervals, and each immersion time interval is designed to give a corresponding time influence coefficient. The sweat immersion time of each sensor is matched with each immersion time interval to obtain the corresponding time influence coefficient, and the immersion time and the corresponding time influence coefficient are multiplied to obtain the immersion impact value; the immersion impact value, chloride concentration and pH value are normalized and their values ​​are taken, and the preset chloride concentration damage function is used. The total chloride damage value S was calculated cl ; Where T represents the immersion impact value, cl, cl max Expressed as chloride concentration and upper limit of chloride concentration, α cl It is expressed as the chloride-based damage coefficient; pH is expressed as the pH value, and λ is expressed as the synergistic corrosion coefficient of pH and chloride. Under acidic conditions, λ>0, and under alkaline conditions, λ≈0.

[0047] Based on the database, the acid environment damage coefficient and alkaline environment damage coefficient obtained by fitting the acid and alkali corrosion resistance experiment corresponding to each sensor are obtained; using the formula Calculate the pH damage value S pH; Among them, η1 and η2 represent the acid environment damage coefficient and alkaline environment damage coefficient respectively. Acidity is usually more corrosive than alkalinity, η1>η2; m and n are both preset nonlinear exponents, and their values ​​are both greater than 1, reflecting the accelerated damage of extreme pH.

[0048] Substitute the machine washing damage value, total chloride damage value and pH damage value corresponding to the machine washing abnormality into the preset nonlinear weighted cumulative model Calculate the total exposure damage value S 总 ; where μ i Represents the weight coefficient of each factor, that is, μ N 、μ cl 、μ pH ; Si represents the specific value corresponding to each factor;

[0049] For example, the chloride concentration in a patient's foot sweat, C cl =130, soaking time 8 hours (corresponding to time influence coefficient 0.9), pH = 5.5, substitute into the formula, and α cl =0.8, λ=0.2, and S cl =0.62;

[0050] In an acidic environment, pH = 5.5, using the formula S pH =η1(7-pH) m , η1=0.5,m=2,calculated S pH =0.5×(1.5)2=1.125, which is 0.75 after normalization.

[0051] Machine wash damage N =0.45, S cl =0.62, S pH =0.75, weight μ N =0.3, μ cl =0.4, μ pH =0.3, substitute S 总 =1-(1-0.3×0.45)(1-0.4×0.62)(1-0.3×0.75)=0.68, which exceeds the threshold of 0.6 and triggers the "sensor replacement required" alarm.

[0052] Taking into account the synergistic corrosion of acidity, alkalinity and chlorides, and the cumulative acceleration of machine washing times, the degradation path of sensors in complex environments is reflected; the corrosion resistance parameters of different sensors are stored in the database to achieve personalized damage assessment for one device and one model.

[0053] A comprehensive evaluation and analysis of sensor performance data, material attenuation data, and total exposure damage is performed. The specific analysis method is as follows:

[0054] Corresponding analysis factors obtained based on sensor performance data include impedance value, signal-to-noise ratio, signal drift, and response time; and corresponding analysis factors obtained based on material attenuation data include substrate bending times and coating thickness;

[0055] Obtain the initial impedance value corresponding to the corresponding impedance value of the sensor, calculate the difference between the current impedance value and the initial impedance value and take the absolute value to obtain the impedance drift value, calculate the ratio of the impedance drift value to the initial impedance value to obtain the impedance drift rate, and when the impedance drift rate is greater than the preset impedance drift threshold, generate the corresponding impedance abnormality signal, and mark the impedance abnormality signal as zk.

[0056] pass The current signal-to-noise ratio damage rate is calculated and the signal-to-noise ratio damage rate is generated. When the signal-to-noise ratio damage rate is greater than the preset signal-to-noise ratio damage threshold, a signal-to-noise ratio abnormality signal is generated and the signal-to-noise ratio abnormality signal is marked as xz. Similarly, the drift damage rate corresponding to the current signal drift of the sensor is obtained according to the signal drift and the corresponding drift tolerance threshold. When the drift damage rate is greater than the preset threshold, a drift abnormality signal is generated and the drift abnormality signal is marked as py.

[0057] Based on the preset information collection cycle, the total number of corresponding sensor responses is obtained, and the corresponding response duration is obtained. When the response duration is greater than the preset upper limit of the response duration, the corresponding number of responses is marked as a timeout response. The number of timeout responses is counted to obtain the number of timeout responses. The timeout response rate is calculated by ratioing the number of timeout responses to the total number of sensor responses corresponding to the corresponding collection cycle. When the timeout response rate is greater than the preset timeout response threshold, a response abnormality signal is generated and marked as yt.

[0058] Obtain the number of substrate bends corresponding to the current sensor and the corresponding maximum bend number. Calculate the difference between the maximum bend number and the current substrate bend number to obtain the remaining bend number. Calculate the ratio of the remaining bend number to the maximum bend number to obtain the remaining bend index. When the remaining bend index is less than the preset remaining bend warning number, a substrate bend abnormality signal is generated and marked as wq.

[0059] The coating thickness corresponding to the current sensor is compared with the initial coating thickness given in advance to obtain the wear thickness, and the wear ratio is calculated by ratioing the wear thickness to the initial coating thickness. When the wear ratio is greater than the preset wear ratio threshold, a thickness abnormality signal is generated and marked as hd.

[0060] The signals corresponding to each analysis factor are aggregated to obtain the total number of factors Y. The signal set corresponding to the analysis factor is (Y zk , Y xz , Y py , Yyt , Y wq , Y hd ), when an impedance abnormal signal, a drift abnormal signal, a response abnormal signal, a substrate bending abnormal signal, or a thickness abnormal signal is detected, the corresponding signal set is lit up, and the signal set of the current sensor is identified to obtain the number of lit signals and recorded as the number of abnormal signals Ys; the total number of analysis factors is based on the preset minimum signal number and maximum signal number, and marked as Ymax and Ymin respectively; when Ys<Ymin, the corresponding sensing performance state is generated as no performance abnormality; when Ymin≤Ys<Ymax, the corresponding sensor performance state is generated as the performance fluctuation stage; when Ys≥Ymax, the corresponding sensor performance state is generated as the performance abnormality stage.

[0061] Obtain the exposure damage threshold value given in advance by each sensor. When the total exposure damage value corresponding to the sensor currently exceeds the exposure damage threshold, generate a sensor exposure abnormality signal. At the same time, when the sensor performance status corresponding to the sensor is in the performance abnormality stage, generate a sensor performance abnormality signal. When it is monitored that the sensor has either or both the sensor exposure abnormality signal and the sensor performance abnormality signal, generate a sensor self-test status indicating that the sensor needs to be replaced.

[0062] The edge computing evaluation module receives the patient's vital signs data and performs health assessment and analysis on the vital signs data to obtain a health risk signal. The cloud server transmits the patient's health risk signal to the local intelligent interaction module. Specifically, the patient's vital signs data is identified to obtain the patient's body temperature, pressure, blood oxygen concentration, humidity and motion data at each collection point, and the patient's vital signs data corresponding to each collection point is input into a pre-set diabetic foot risk monitoring model, and the diabetic foot risk monitoring model outputs a diabetic foot risk index. When the diabetic foot risk index is greater than the preset risk threshold, a health risk signal is generated.

[0063] The local intelligent interaction module stores the patient's vital signs data and self-examination data in the deployed database, and provides intelligent reminders based on the received sensor self-examination status and health risk signals; when the sensor self-examination status corresponds to the need to replace the sensor, the corresponding sensor location information and status information are sent to the patient's mobile terminal for intelligent voice broadcast reminders; when a health risk signal is detected, a multimodal alarm is triggered to the patient's mobile terminal.

[0064] The present invention also provides a real-time monitoring method for diabetic foot risk based on a wearable device, comprising the following steps:

[0065] Step 1: Identify the sensor self-test status: Obtain and identify the sensor self-test status corresponding to each sensor. When it is detected that the sensor self-test status corresponds to a sensor that needs to be replaced, mark the corresponding sensor as a sensor to be replaced, and increase the self-test frequency of the sensor to be replaced.

[0066] Step 2: Risk Warning Guidance: When a sensor to be replaced is detected, a risk warning instruction is generated and the interaction status of the risk warning instruction is obtained. The risk warning instruction includes but is not limited to information pop-up windows, voice broadcasts, and mobile terminal vibrations. The interaction status is the patient's instruction feedback, such as clicking a physical button to confirm receipt of the alarm.

[0067] Step 3. Interaction status identification: Identify the interaction status. When the instruction feedback corresponding to the patient's interaction status is received, increase the self-check frequency of the sensor to be replaced, reduce the risk prompt frequency, and obtain the preset upper limit of the prompt time. When the corresponding sensor self-check status within the upper limit of the prompt time is no performance abnormality, cancel the risk prompt. Otherwise, generate an emergency start instruction. When the instruction feedback corresponding to the patient's interaction status is no response, increase the self-check frequency of the sensor to be replaced, increase the risk prompt frequency, and obtain the preset upper limit of the no-feedback time. When the no-response time exceeds the upper limit of the no-feedback time, generate an emergency start instruction.

[0068] Step 4: Emergency Startup Execution: Upon receiving the emergency start command, the cloud server sends a risk warning command to pre-set emergency family contacts and medical personnel. For example, after the patient receives a voice reminder that the "sensor needs to be replaced," they press the device's physical button to confirm "receipt." The system reduces the reminder frequency to once a day and increases the self-check frequency to twice an hour. If the damage value does not decrease within 72 hours, it is determined to be "required for mandatory replacement." A work order containing the sensor location (e.g., "left foot insole pressure module") and status is automatically sent to the child, who can then place an online order for replacement service.

[0069] As described above, the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that the technical solutions described in the above embodiments can still be modified, or some of the technical features thereof can be replaced by equivalents. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A real-time monitoring system for diabetic foot risk based on wearable devices, characterized by: include: The sensor self-test module is used to receive and identify the self-test data to obtain environmental exposure data, sensor performance data, and material attenuation data, and substitute the environmental exposure data into the exposure damage analysis mechanism to obtain the total exposure damage value; A comprehensive evaluation and analysis of sensor performance data, material attenuation data, and total exposure damage value is performed to obtain the sensor self-test status. The cloud server transmits the sensor self-test status to the local intelligent interaction module. The edge computing assessment module is used to receive patient vital sign data and perform health assessment analysis on the data to obtain health risk signals. The cloud server transmits the patient's health risk signals to the local intelligent interaction module; A local intelligent interaction module is used to store patient vital sign data and self-examination data in the deployed database and provide intelligent reminders based on the received sensor self-examination status and health risk signals; When the sensor self-test status corresponds to the need to replace the sensor, the corresponding sensor location information and status information will be sent to the patient's mobile terminal for intelligent voice broadcast reminder; When health risk signals are detected, a multimodal alarm is triggered to the patient's mobile terminal.

2. The real-time monitoring system for diabetic foot risk based on wearable devices according to claim 1 is characterized in that: The specific analysis method of the exposure damage analysis mechanism is as follows: The number of machine washing times of each sensor is obtained according to the environmental exposure data and recorded as N. The machine washing damage sensitivity coefficient of each sensor in the database corresponding to the washing resistance experiment is obtained and recorded as α. N ; Substitute the machine wash times and corresponding machine wash damage sensitivity coefficients of each sensor into the preset machine wash damage function S N =1-e -α N Nβ Calculate the machine wash damage value S N ; Where β represents a nonlinear index; when the machine wash damage value is greater than a preset threshold, the corresponding machine wash damage value is marked as a machine wash abnormality; The sweat immersion time, pH value and chloride concentration corresponding to each sensor are obtained according to the environmental exposure data; the sweat immersion time is divided into multiple immersion time intervals, and each immersion time interval is designed to give a corresponding time influence coefficient. The sweat immersion time of each sensor is matched with each immersion time interval to obtain the corresponding time influence coefficient, and the immersion time and the corresponding time influence coefficient are multiplied to obtain the immersion impact value; the immersion impact value, chloride concentration and pH value are normalized and their values ​​are taken, and the preset chloride concentration damage function is used. The total chloride damage value S was calculated cl ; Where T represents the immersion impact value, cl, cl max Expressed as chloride concentration and upper limit of chloride concentration, α cl It is expressed as the chloride basic damage coefficient; pH is expressed as the pH value, and λ is expressed as the synergistic corrosion coefficient of pH and chloride; Based on the database, the acid environment damage coefficient and alkaline environment damage coefficient obtained by fitting the acid and alkali corrosion resistance experiment corresponding to each sensor are obtained; using the formula Calculate the pH damage value S pH ; Wherein, η1 and η2 represent the acid environment damage coefficient and the alkaline environment damage coefficient respectively; m and n are both preset nonlinear exponents; Substitute the machine washing damage value, total chloride damage value and pH damage value corresponding to the machine washing abnormality into the preset nonlinear weighted cumulative model Calculate the total exposure damage value S 总 ; where μ i Represents the weight coefficient of each factor, that is, μ N 、μ cl 、μ pH ; Si represents the specific value corresponding to each factor.

3. The real-time monitoring system for diabetic foot risk based on wearable devices according to claim 1 is characterized in that: The comprehensive evaluation and analysis of sensor performance data, material attenuation data, and total exposure damage value is performed as follows: Corresponding analysis factors obtained based on sensor performance data include impedance value, signal-to-noise ratio, signal drift, and response time; and corresponding analysis factors obtained based on material attenuation data include substrate bending times and coating thickness; Obtain the initial impedance value corresponding to the corresponding impedance value of the sensor, calculate the difference between the current impedance value and the initial impedance value and take the absolute value to obtain the impedance drift value, calculate the ratio of the impedance drift value to the initial impedance value to obtain the impedance drift rate, and when the impedance drift rate is greater than the preset impedance drift threshold, generate the corresponding impedance abnormality signal and mark the impedance abnormality signal as zk; pass Calculate the current signal-to-noise ratio damage rate. When the signal-to-noise ratio damage rate is greater than a preset signal-to-noise ratio damage threshold, generate a signal-to-noise ratio abnormal signal, and mark the signal-to-noise ratio abnormal signal as xz. Similarly, obtain the drift damage rate corresponding to the current signal drift of the sensor based on the signal drift and the corresponding drift tolerance threshold. When the drift damage rate is greater than the preset threshold, generate a drift abnormal signal, and mark the drift abnormal signal as py. Based on a preset information collection period, the total number of corresponding sensor responses is obtained, and the corresponding response duration is obtained. When the response duration is greater than a preset upper limit of the response duration, the corresponding number of responses is marked as a timeout response. The number of timeout responses is counted to obtain the number of timeout responses. The timeout response number is calculated by ratioing the number of timeout responses to the total number of sensor responses corresponding to the corresponding collection period to obtain a timeout response rate. When the timeout response rate is greater than a preset timeout response threshold, a response abnormality signal is generated, and the response abnormality signal is marked as yt. Obtain the number of substrate bends corresponding to the current sensor and the corresponding maximum bend number. Calculate the difference between the maximum bend number and the current substrate bend number to obtain the remaining bend number. Calculate the ratio of the remaining bend number to the maximum bend number to obtain the remaining bend index. When the remaining bend index is less than the preset remaining bend warning number, generate a substrate bend abnormality signal, and mark the substrate bend abnormality signal as wq. The coating thickness corresponding to the current sensor is compared with the initial coating thickness given in advance to obtain the wear thickness, and the wear ratio is calculated by comparing the wear thickness to the initial coating thickness. When the wear ratio is greater than the preset wear ratio threshold, a thickness abnormality signal is generated and marked as hd; The signals corresponding to each analysis factor are aggregated to obtain the total number of factors Y. The signal set corresponding to the analysis factor is (Y zk , Y xz , Y py , Y yt , Y wq , Y hd ), when an impedance abnormal signal, a drift abnormal signal, a response abnormal signal, a substrate bending abnormal signal, or a thickness abnormal signal is detected, the corresponding signal set is lit up, the signal set of the current sensor is identified to obtain the number of lit signals and recorded as the number of abnormal signals Ys, and the sensor performance status is obtained by analysis based on the number of abnormal signals; and the total value of sensor exposure damage and the sensor performance status are diagnosed and analyzed to obtain the sensor self-test status.

4. The real-time monitoring system for diabetic foot risk based on wearable devices according to claim 3 is characterized in that: The analysis is performed based on the number of abnormal signals, which is specifically as follows: Get the preset minimum and maximum signal numbers corresponding to the total number of analysis factors, and mark them as Ymax and Ymin respectively; When Ys<Ymin, the corresponding sensing performance state is generated as no performance abnormality; When Ymin≤Ys<Ymax, the corresponding sensor performance state is generated as the performance fluctuation stage; When Ys≥Ymax, the corresponding sensor performance state is generated as the performance abnormality stage.

5. The real-time monitoring system for diabetic foot risk based on wearable devices according to claim 3 is characterized in that: The diagnostic analysis of the total sensor exposure damage value and the sensor performance status is specifically as follows: obtaining a pre-designed exposure damage threshold corresponding to each sensor, and generating a sensor exposure abnormality signal when the current total exposure damage value corresponding to the sensor exceeds the exposure damage threshold; simultaneously obtaining a sensor performance abnormality signal when the corresponding sensor performance status is in the performance abnormality stage; and generating a sensor self-test status indicating that the sensor needs to be replaced when either or both the sensor exposure abnormality signal and the sensor performance abnormality signal are detected in the sensor.

6. The real-time monitoring system for diabetic foot risk based on wearable devices according to claim 1, characterized in that: The edge computing evaluation module specifically identifies the patient's vital signs data to obtain the patient's body temperature, pressure, blood oxygen concentration, humidity and movement data at each collection point, inputs the patient's vital signs data corresponding to each collection point into a pre-set diabetic foot risk monitoring model, and the diabetic foot risk monitoring model outputs a diabetic foot risk index. When the diabetic foot risk index is greater than a preset risk threshold, a health risk signal is generated.

7. The real-time monitoring system for diabetic foot risk based on wearable devices according to claim 1, characterized in that: It also includes a wearable sensor module, which includes a multimodal sensor group deployed in a wearable device according to a pre-designed position. The cloud server controls the multimodal sensor group to collect patient information to obtain patient vital signs data, and controls the sensor to perform self-test to obtain self-test data; and the server transmits the patient vital signs data to the edge computing evaluation module and transmits the self-test data to the sensor self-test module; the patient vital signs data includes the patient's body temperature, pressure, blood oxygen concentration, humidity and motion data at each collection point; the self-test data includes environmental exposure data, sensor performance data and material attenuation data.

8. A real-time monitoring method for diabetic foot risk based on wearable devices, characterized in that: The system for real-time monitoring of diabetic foot risk for wearable devices according to any one of claims 1 to 7 comprises the following steps: Step 1: Obtain the sensor self-test status corresponding to each sensor and identify it. When it is detected that the sensor self-test status corresponds to a sensor that needs to be replaced, mark the corresponding sensor as a sensor to be replaced and increase the self-test frequency of the sensor to be replaced. Step 2: When a sensor to be replaced is detected, a risk warning instruction is generated, and the interaction status of the risk warning instruction is obtained; Step 3: Identify the interaction status and obtain the emergency start instruction based on the analysis of the interaction status; Step 4: When receiving the emergency start command, the cloud server sends the risk warning instruction to the pre-set emergency family contacts and medical care terminals.

9. The real-time monitoring method for diabetic foot risk based on wearable devices according to claim 8, characterized in that: The third step specifically includes: identifying the interaction status; when the instruction feedback corresponding to the patient's interaction status is received, increasing the self-check frequency of the sensor to be replaced and reducing the risk prompt frequency; obtaining a preset upper limit on the prompt duration; and when the corresponding sensor self-check status within the upper limit indicates no performance abnormality, releasing the risk prompt; otherwise, generating an emergency start instruction; When the command feedback corresponding to the patient's interaction status is no response, the self-inspection frequency of the sensor to be replaced is increased, and the risk prompt frequency is increased to obtain the preset upper limit of the no-feedback time. When the no-response time exceeds the upper limit of the no-feedback time, an emergency start command is generated.

Images