Noninvasive skin acute failure detector, system and method

Abstract

The invention relates to a non-invasive acute skin failure detector, system and method, and provides a silica gel wrist strap type non-invasive acute skin failure detector which can detect acute skin failure through a temperature sensor, a photoplethysmography sensor and a capacitance sensor which are integrated on the silica gel wrist strap type non-invasive acute skin failure detector. The skin temperature value, the capillary artery perfusion pressure and the skin water content are measured and calculated respectively, so that skin data of a patient are noninvasively detected and reported to a background, whether the patient has an acute skin failure symptom or not is judged through background prediction analysis, intervention measures are quickly given, clinical diagnosis and treatment efficiency can be greatly improved, and the clinical diagnosis and treatment cost is reduced. And intelligent auxiliary diagnosis and treatment are realized.

Description

Technical Field

[0001] The present disclosure relates to the field of clinical detection technologies, and particularly to a non-invasive acute skin failure detector, a system and a method, and an electronic device. Background Art

[0002] Heart failure will activate the body's renin-angiotensin-aldosterone system and promote the secretion of antidiuretic hormone, thereby causing water and sodium retention and resulting in fluid accumulation. Patients usually present with lower extremity or systemic edema. In addition, the skin of elderly patients is thin and their resistance is reduced, making them prone to acute skin failure.

[0003] Clinically, the diagnosis of the definition of acute skin failure and various clinical manifestations of skin perfusion insufficiency is mainly based on the following indicators: ① Skin cold and wet: at room temperature, the skin temperature is 18-22°C, the humidity is 55%-65%, and the body surface temperature measured by an infrared thermometer gun is <33.5°C without physical cooling, and phenomena such as skin moisture appear; ② Skin Mottling Score (SMS). When the body shows acute hypoperfusion, the skin blood vessels constrict to meet the perfusion of vital organs. When the perfusion pressure at the arterial end of the skin capillaries is lower than 25-30 mmHg (1 mmHg = 0.133 kPa), the arterial end of the skin capillaries will close, resulting in the inability of the arterial end of the skin capillaries to maintain an effective perfusion pressure, thereby causing local skin ischemia and necrosis and resulting in acute skin failure.

[0004] Currently, there are three techniques available for measuring skin perfusion pressure: radionuclide clearance technique, photoplethysmography, and laser Doppler technique. The principles of the three techniques are the same, that is, during the slow release of the pressure of the inflatable cuff at the measurement site, the flushing of the radionuclide, the reappearance of the pulsatile flow, and the movement of red blood cells are detected. At this moment, the minimum pressure generated by the cuff and acting on the skin is the skin perfusion pressure. When the pressure is higher than this value, the skin blood flow stops.

[0005] Currently, laser Doppler technique is mostly used in clinical practice for measurement because it can more simply measure lower levels of skin perfusion pressure, and has less trauma, high repeatability, and short time consumption. It is the most widely used measurement technique and has been widely used in evaluating the prognosis of limb ischemic ulcers, the severity of ischemia, and predicting wound outcomes. Its application in critically ill patients needs to be gradually promoted. However, laser Doppler technique also has the following inconveniences in use:

[0006] First of all, laser Doppler technique is only used to measure skin perfusion pressure, which requires one-to-one operation by nurses, and only one indicator of the patient can be measured each time, and it is impossible to simultaneously measure the skin temperature, humidity and skin water content of the patient, because these are all characteristic data reflecting the changes in acute skin failure of the patient, so it cannot measure them simultaneously;

[0007] Secondly, for patients with acute skin failure, the above-mentioned indicators are currently used for manual assessment, which takes a long time and occupies nursing resources. There is a lack of intelligent and automated assessment technology, resulting in low efficiency. Summary of the Invention

[0008] On the one hand, the present application provides a non-invasive acute skin failure detector, including a silicone wristband, on which a non-invasive acute skin failure detection system is provided. The non-invasive acute skin failure detector includes:

[0009] A temperature sensor for collecting skin temperature signals;

[0010] A photoplethysmography sensor for collecting skin pulse wave signals;

[0011] A capacitance sensor for measuring skin dielectric constant;

[0012] A DSP processor for signal digital-to-analog conversion processing and sending the converted value to the MCU;

[0013] An MCU for controlling the sensor to collect the corresponding skin temperature value, skin pulse signal value and skin dielectric constant, performing pulse waveform analysis based on the skin pulse signal value, calculating the capillary arterial perfusion pressure at the arterial skin, and estimating the skin water content in combination with the skin dielectric constant; and, forwarding the skin temperature value, the capillary arterial perfusion pressure and the skin water content to the communication module in real time;

[0014] A communication module for providing data communication between the non-invasive acute skin failure detector and the background server, including: reporting the skin temperature value, the capillary arterial perfusion pressure and the skin water content to the background server in real time through the communication module; after the background server records and analyzes the skin temperature value, the capillary arterial perfusion pressure and the skin water content, obtaining the skin acute failure detection result of the patient and sending it to the non-invasive acute skin failure detector, which is received by the communication module and forwarded to the MCU;

[0015] An LED for real-time displaying the skin temperature value, the capillary arterial perfusion pressure and the skin water content of the patient and the corresponding skin acute failure detection result;

[0016] A battery for power supply;

[0017] The temperature sensor, the photoplethysmography sensor and the capacitance sensor are respectively electrically connected to the DSP processor;

[0018] The DSP processor, the communication module, the LED and the battery are respectively electrically connected to the MCU.

[0019] As an alternative embodiment of the present application, optionally, the capillary arterial perfusion pressure P perf is calculated as follows:

[0020]

[0021] Where:

[0022] F is the amplitude of the pulse pressure signal detected by the sensor;

[0023] E is the vascular elasticity coefficient (related to age and disease);

[0024] R is the vascular radius (inverted by waveform analysis);

[0025] η is the blood viscosity correction factor (default value is 0.15).

[0026] As an alternative embodiment of the present application, optionally, the calculation method of the skin water content W is as follows:

[0027]

[0028] ε S is the measured skin dielectric constant;

[0029] ε d is the reference dielectric constant of dry skin (default value is 3.5);

[0030] ε W is the dielectric constant of pure water (default value is 80);

[0031] ρ is the tissue density correction factor (empirical value is 0.93).

[0032] On the other hand, the present application proposes a non-invasive skin acute failure detection system, which optionally includes:

[0033] A non-invasive skin acute failure detector;

[0034] A background server for recording and saving the skin temperature value, capillary arterial perfusion pressure and skin water content of the patient uploaded by the non-invasive skin acute failure detector; and, through a preset acute skin failure AI recognition model, performing data feature analysis on the skin temperature value, the capillary arterial perfusion pressure and the skin water content of the patient to identify whether the patient has acute skin failure:

[0035] If so, output the corresponding acute skin failure symptom stage and corresponding intervention strategy, and send them to the nursing terminal; at the same time, send them to the non-invasive skin acute failure detector;

[0036] Otherwise, generate a skin normal notice and send it to the nursing terminal;

[0037] A nursing terminal, which is used to receive and display the acute skin failure symptom stage and the corresponding intervention strategies, or the skin normal notification;

[0038] The non-invasive acute skin failure detector and the nursing terminal are respectively communicatively connected to the background server.

[0039] As an optional implementation scheme of the present application, optionally, the method for generating the acute skin failure AI recognition model includes:

[0040] Collect the diagnostic data of a number of acute skin failure patients;

[0041] Conduct data feature analysis, and count the skin temperature value, capillary arterial perfusion pressure and skin water content in the diagnostic data;

[0042] Conduct feature annotation, and label the corresponding acute skin failure symptom stage and the corresponding intervention strategies for the data features;

[0043] Count the data features of each acute skin failure patient, construct a feature set and divide it into a training set and a validation set according to a preset ratio;

[0044] Input the training set into the RF model for feature classification learning to generate the acute skin failure AI recognition model;

[0045] Use the validation set to verify the classification and recognition performance of the acute skin failure AI recognition model:

[0046] If the verification is passed, deploy and apply the acute skin failure AI recognition model to the background server;

[0047] Otherwise, retrain.

[0048] On the other hand, the present application proposes a non-invasive acute skin failure detection method, including the following steps:

[0049] Wear and activate the non-invasive acute skin failure detector, and establish data communication with the background server;

[0050] The non-invasive acute skin failure detector continuously collects the skin temperature value, capillary arterial perfusion pressure and skin water content of the patient, and continuously reports them to the background server;

[0051] The background server records and saves the skin temperature value, the capillary arterial perfusion pressure and the skin water content of the patient, and analyzes the data features of the skin temperature value, the capillary arterial perfusion pressure and the skin water content of the patient through a preset acute skin failure AI recognition model to identify whether the patient has acute skin failure:

[0052] If so, output the corresponding acute skin failure symptom stage and the corresponding intervention strategy, and send them to the nursing terminal; at the same time, send them to the non-invasive skin acute failure detector;

[0053] Otherwise, generate a skin normal notice and send it to the nursing terminal;

[0054] The nursing terminal is used to receive and display the acute skin failure symptom stage and the corresponding intervention strategy, or the skin normal notice.

[0055] On the other hand, the present application also proposes an electronic device, including:

[0056] A processor;

[0057] A memory for storing instructions executable by the processor;

[0058] Wherein, when the processor is configured to execute the executable instructions, the method described above is implemented.

[0059] Technical effects of the present invention:

[0060] The present invention proposes a non-invasive skin acute failure detector in the form of a silicone wristband, which can respectively measure and calculate the skin temperature value, capillary arterial perfusion pressure and skin water content through the integrated temperature sensor, photoplethysmography sensor and capacitance sensor thereon, so as to non-invasively detect the skin data of patients and report it to the background. Through the prediction and analysis of the background, it can judge whether the patient has acute skin failure symptoms and quickly give intervention measures, which can greatly improve the clinical diagnosis and treatment efficiency and realize intelligent assisted diagnosis and treatment.

[0061] According to the following detailed description of the exemplary embodiments with reference to the accompanying drawings, other features and aspects of the present disclosure will become clear. Brief Description of the Drawings

[0062] The drawings included in the specification and constituting a part of the specification, together with the specification, illustrate the exemplary embodiments, features and aspects of the present disclosure, and are used to explain the principles of the present disclosure.

[0063] Figure 1 It shows the external structure schematic diagram of the detector of the present invention;

[0064] Figure 2 It shows the control system structure schematic diagram of the detector of the present invention;

[0065] Figure 3 It shows the composition structure schematic diagram of the system of the present invention. Detailed Embodiments

[0066] Various exemplary embodiments, features, and aspects of the present disclosure will be described in detail below with reference to the accompanying drawings. Like reference numerals in the drawings denote functionally identical or similar elements. Although various aspects of the embodiments are shown in the drawings, the drawings are not necessarily drawn to scale unless otherwise specified.

[0067] As used herein, the term "exemplary" means "serving as an example, embodiment, or illustration." Any embodiment described herein as "exemplary" is not necessarily to be construed as superior to or better than other embodiments.

[0068] In addition, for a better description of the present disclosure, numerous specific details are given in the following detailed description. Those skilled in the art should understand that the present disclosure can be implemented without some of these specific details. In some instances, well-known means, elements, and circuits have not been described in detail so as to highlight the gist of the present disclosure.

[0069] Embodiment 1

[0070] As Figure 1 shown, in one aspect of the present application, a non-invasive skin acute failure detector is proposed, including a silicone wristband 100, on which a non-invasive skin acute failure detection system 200 is provided. The non-invasive skin acute failure detector 200 includes:

[0071] A temperature sensor for collecting skin temperature signals;

[0072] A photoplethysmography sensor for collecting skin pulse wave signals;

[0073] A capacitance sensor for measuring skin dielectric constant;

[0074] A DSP processor for signal digital-to-analog conversion processing and sending the converted value to the MCU;

[0075] An MCU for controlling the sensor to collect the corresponding skin temperature value, skin pulse signal value, and skin dielectric constant, performing pulse waveform analysis based on the skin pulse signal value, calculating the capillary arterial perfusion pressure at the arterial skin, and estimating the skin water content in combination with the skin dielectric constant; and, forwarding the skin temperature value, the capillary arterial perfusion pressure, and the skin water content to the communication module in real time;

[0076] A communication module is used to provide data communication between the non-invasive skin acute failure detector and the background server, including: reporting the skin temperature value, the capillary arterial perfusion pressure, and the skin water content to the background server in real time through the communication module; after the background server records and analyzes the skin temperature value, the capillary arterial perfusion pressure, and the skin water content, obtaining the skin acute failure detection result of the patient and sending it to the non-invasive skin acute failure detector, which is received by the communication module and forwarded to the MCU;

[0077] An LED is used to display the skin temperature value, the capillary arterial perfusion pressure, and the skin water content of the patient in real time, as well as the corresponding skin acute failure detection result;

[0078] A battery is used for power supply;

[0079] The temperature sensor, the photoplethysmography sensor, and the capacitance sensor are respectively electrically connected to the DSP processor;

[0080] The DSP processor, the communication module, the LED, and the battery are respectively electrically connected to the MCU.

[0081] The invention combines an instrument for measuring body surface temperature, capillary arterial perfusion pressure, and skin water content, which can be in the mode of an infrared gun or a wearable instrument, and measures these values by attaching to the skin surface. The detector has an IoT communication function and can report data to the background. Users such as nurses can record these values and perform model prediction through a small program or APP connected to the instrument to predict the progression of the patient's skin failure.

[0082] In this embodiment, a silicone wristband is used to integrate the non-invasive skin acute failure detector 200 to provide multi-dimensional skin data collection and report it to the background. It can be understood in combination with the data communication principle between the existing smart bracelet and the background.

[0083] As Figure 2 shown, it is the control composition of the system.

[0084] I. Design and Implementation of the Sensor Module

[0085] Temperature Sensor

[0086] Model Selection and Deployment: An NTC thermistor (negative temperature coefficient) is used as the core component, with a response time ≤ 50 ms, a measurement range of 25°C to 45°C, and an accuracy of ±0.1°C.

[0087] Signal Acquisition: The sensor is driven by a constant current source circuit, and the environmental interference is eliminated by combining a temperature compensation algorithm.

[0088] Photoplethysmography Sensor

[0089] Optical path design: Use green light (wavelength 530nm) LED and photodiode combination, with a penetration depth of 0.5-2mm to detect changes in subcutaneous blood flow.

[0090] Motion Artifact Suppression: Integrated acceleration sensor, using adaptive filtering algorithm to eliminate signal distortion caused by limb movement.

[0091] Capacitive Sensors

[0092] Electrode structure: Design interdigital flexible electrodes (spacing 0.5mm), detection frequency 100kHz-1MHz, measurement error of skin dielectric constant ≤3%1.

[0093] Temperature Compensation: Dynamically corrects dielectric constant measurements based on temperature sensor data.

[0094] 2. Signal Processing and Control Logic

[0095] DSP Processor

[0096] Digital-to-analog conversion: Equipped with 24-bit high-precision ADC, sampling rate ≥ 1kHz, supporting synchronous acquisition of three-way sensor signals.

[0097] Preprocessing process:

[0098] Perform sliding average filtering on the temperature signal;

[0099] Implement bandpass filtering (0.5-10Hz) on the pulse wave signal;

[0100] Perform a baseline calibration on the capacitance signal.

[0101] MCU control and algorithm

[0102] Sensor scheduling: A time slice polling mechanism is used to control the three sensors to work in time-sharing mode to reduce power consumption.

[0103] Key algorithm implementation:

[0104]

[0105] 3. Communication and Data Management

[0106] Communication module selection

[0107] Adopt low-power Bluetooth (BLE 5.2) + Wi-Fi dual-mode chip to achieve the following functions:

[0108] Local data cache (stores 24 hours of historical data);

[0109] Upload data packets in real time (transmission interval is adjustable: 1s-60s);

[0110] Background server interaction protocol:

[0111] {

[0112] "timestamp":"20250303143000",

[0113] "temp":36.5, / / Unit: °C

[0114] "pressure":82.3, / / Unit: mmHg

[0115] "hydration":68.4 / / Unit: %

[0116] }

[0117] 。

[0118] Other communication protocols can also be used, such as hqtt, etc.

[0119] Analysis logic:

[0120] Determination conditions for acute failure:

[0121] (The temperature remains > 38°C and the pressure < 60 mmHg) or (the water content > 75% and the pressure drop rate > 5% / min).

[0122] IV. Power supply and display system

[0123] Battery management

[0124] Configure a 3.7V / 1200mAh lithium polymer battery, supporting the following power consumption optimizations:

[0125] Dynamic voltage regulation: The main frequency of the MCU drops to 32 MHz in the idle state;

[0126] Sensor sleep mode: Automatically enters after 10 minutes of no operation.

[0127] LED display driver

[0128] Adopt a segment code type LED screen, display logic:

[0129] Temperature value (green always on);

[0130] Perfusion pressure (yellow: 60 - 80 mmHg; red: < 60 mmHg);

[0131] Water content (blue: < 70%; purple: ≥ 70%).

[0132] V. System integration and verification

[0133] EMC design

[0134] The sensor is physically isolated from the communication module, and the PCB layout follows the following principles:

[0135] The analog signal traces are wrapped around the ground plane;

[0136] The digital power supply and the analog power supply are independently powered.

[0137] Clinical verification process

[0138] Accuracy test: Compared with the gold standard device (such as a laser Doppler), the correlation coefficient R is required to be 2 ≥0.95.

[0139] Response time: From the sensor trigger to the completion of the LED display update ≤ 300 ms.

[0140] Implementation effect: This solution realizes a skin temperature detection error of ±0.2 °C, a perfusion pressure error of ±3 mmHg, and a water content error of ±2% in a laboratory environment, meeting the early warning requirements for acute failure.

[0141] As an alternative implementation of this application, optionally, the capillary arterial perfusion pressure P perf is calculated as follows:

[0142]

[0143] Where:

[0144] F is the amplitude of the pulse pressure signal detected by the sensor;

[0145] E is the vascular elasticity coefficient (related to age and disease, determined by the administrator according to the patient's condition);

[0146] R is the vascular radius (inverted through waveform analysis or obtained based on empirical values);

[0147] η is the blood viscosity correction factor (default value is 0.15).

[0148] Function of this formula: By correlating the pressure signal with the vascular characteristics, the capillary perfusion pressure is deduced. Combining the characteristics of the patient's blood and blood vessels, the capillary arterial perfusion pressure is comprehensively calculated.

[0149] As an alternative implementation of this application, optionally, the skin water content W is calculated as follows:

[0150]

[0151] ε S is the measured skin dielectric constant;

[0152] ε d is the dry skin reference dielectric constant (default value is 3.5);

[0153] ε W is the dielectric constant of pure water (default 80);

[0154] ρ is the tissue density correction coefficient (empirical value 0.93).

[0155] The function of this formula: Based on the dielectric constant difference, quantify the proportion of skin water content. It can be understood and implemented in combination with the existing water content calculation method.

[0156] As Figure 3 shown, on the other hand, this application proposes a non-invasive skin acute failure detection system, including:

[0157] The above-mentioned non-invasive skin acute failure detector;

[0158] A background server, used to record and save the skin temperature value, capillary arterial perfusion pressure and skin water content of the patient uploaded by the non-invasive skin acute failure detector; and, through a preset acute skin failure AI recognition model, perform data feature analysis on the skin temperature value, the capillary arterial perfusion pressure and the skin water content of the patient to identify whether the patient has acute skin failure:

[0159] If so, output the corresponding acute skin failure symptom stage and the corresponding intervention strategy, and send them to the nursing terminal; at the same time, send them to the non-invasive skin acute failure detector;

[0160] Otherwise, generate a skin normal notification and send it to the nursing terminal;

[0161] A nursing terminal, used to receive and display the acute skin failure symptom stage and the corresponding intervention strategy, or the skin normal notification;

[0162] The non-invasive skin acute failure detector and the nursing terminal are respectively communicatively connected to the background server.

[0163] Among them, the generation method of the acute skin failure AI recognition model includes:

[0164] Collect the diagnostic data of several acute skin failure patients;

[0165] Perform data feature analysis, and count the skin temperature value, capillary arterial perfusion pressure and skin water content in the diagnostic data;

[0166] Perform feature annotation, and label the corresponding acute skin failure symptom stage and the corresponding intervention strategy for the data features;

[0167] Count the data features of each acute skin failure patient, construct a feature set and divide it into a training set and a validation set according to a preset ratio;

[0168] Input the training set into the RF model for feature classification learning to generate the acute skin failure AI recognition model;

[0169] Use the validation set to verify the classification and recognition performance of the acute skin failure AI recognition model:

[0170] If the verification passes, deploy and apply the acute skin failure AI recognition model to the background server;

[0171] Otherwise, retrain.

[0172] The HIS system can be deployed on the background to create the medical records of patients and record the skin temperature value, capillary arterial perfusion pressure, skin water content, recognition structure, and / or intervention measures of patients.

[0173] An acute skin failure AI recognition model is deployed on the background, which can perform acute skin failure AI diagnosis based on a decision tree such as the RF model, assist in recognition, and output corresponding strategies.

[0174] The specific implementation plan is as follows:

[0175] I. System Architecture Design

[0176] 1. Overall Architecture

[0177] Device layer: The non-invasive skin detector collects patient data (temperature / pressure / water content, specifically understood in combination with the above detector) in real time and uploads it to the background server through an encryption protocol (Bluetooth).

[0178] Service layer:

[0179] Data receiving module: Processes the real-time data stream uploaded by the device.

[0180] Database module: Uses a time-series database (such as InfluxDB) to store dynamic monitoring data and a relational database (such as MySQL) to store patient files.

[0181] AI inference module: Loads the trained RF model and analyzes data features in real time.

[0182] Application layer: The nursing terminal subscribes to the results pushed by the server through WebSocket or API and displays them on the Web / App interface.

[0183] 2. Technology Stack Selection

[0184] Backend: Python (Flask / Django) or Java (Spring Boot) to handle business logic.

[0185] Database: InfluxDB + MySQL + Redis (cache).

[0186] AI frameworks: Scikit-learn (RF model training), ONNX (model deployment optimization).

[0187] Communication protocols: HTTPS (secure transmission), MQTT (low-latency device communication).

[0188] II. Development process of the acute skin failure AI model

[0189] 1. Data preparation phase

[0190] Data collection:

[0191] Sources: Hospital historical cases, data shared by collaborating institutions, simulated data generation.

[0192] Types: Structured data (temperature, pressure, water content values) + labels (symptom stage, intervention strategy).

[0193] Data cleaning:

[0194] Handling missing values: Filling with interpolation method or removing invalid records.

[0195] Outlier detection: 3σ principle or Isolation Forest algorithm.

[0196] Standardization: Z-Score normalization to eliminate dimensional differences.

[0197] 2. Feature engineering

[0198] Static features: Patient age, gender, underlying diseases.

[0199] Dynamic features include:

[0200] Time series features: Sliding window statistics (mean, variance, trend slope).

[0201] Cross features: Product term of temperature and water content (may reflect metabolic status).

[0202] Feature selection: Screening Top-K features through Feature Importance of Random Forest.

[0203] 3. Model training and validation

[0204] Training set / validation set division: Divided by time series (to avoid future information leakage), ratio 8:2.

[0205] Hyperparameter tuning: Grid search (n_estimators, max_depth, min_samples_split).

[0206] Verification metrics:

[0207] Classification accuracy, F1-Score (handling class imbalance).

[0208] ROC-AUC is used to evaluate the discriminability of the model.

[0209] Failure handling: If AUC < 0.85, it is necessary to check the data quality or introduce ensemble learning (such as XGBoost).

[0210] 4. Model Deployment

[0211] Export it in PMML or ONNX format, and use TensorFlow Serving or RedisAI to achieve low-latency inference.

[0212] Service encapsulation: Provide a REST API interface, and the input is a JSON-format data packet.

[0213] III. Implementation of Key Modules of the Background Server

[0214] 1. Data Receiving API

[0215] ```python

[0216] Flask example: Receive device data

[0217] from flask import request,jsonify

[0218] @app.route(' / upload',methods=['POST'])

[0219] def handle_upload():

[0220] data = request.json

[0221] if validate(data): Verify device ID and data integrity

[0222] save_to_influxdb(data) Store in the time series database

[0223] result = ai_model.predict(data) Call the AI model

[0224] push_to_terminal(result) Push to the nursing terminal

[0225] return jsonify({"status":"success"})

[0226] else:

[0227] return jsonify({"error":"Invalid data"}),400

[0228] ```

[0229] 2. Real - time decision - making logic

[0230] AI recognition trigger: After the device uploads data, the model inference is automatically triggered.

[0231] Result distribution:

[0232] Positive cases: Asynchronously pushed to the nursing terminal and the device through a message queue (such as Kafka).

[0233] Negative cases: Generate notification logs and mark the Web interface as green status.

[0234] IV. Nursing terminal function development

[0235] 1. Interaction design

[0236] Alarm panel: Highlight acute failure cases in red and click to view details (symptom stage, recommended medications, nursing steps).

[0237] Historical records: The data change curve of the patient within 72 hours can be traced back.

[0238] Intervention feedback: After the nurse executes the measures, the execution record (photo / text) can be uploaded.

[0239] 2. Technical implementation

[0240] ```javascript

[0241] / / WebSocket real - time listening example

[0242] const socket = new WebSocket('wss: / / server / updates');

[0243] socket.onmessage = (event) => {

[0244] const data = JSON.parse(event.data);

[0245] if(data.type === 'emergency'){

[0246] showAlert(`Patient ${data.patientID} has stage ${data.stage} failure and requires ${data.action}`);

[0247] } else {

[0248] updateStatusTable(data.patientID, 'Normal');

[0249] }

[0250] }。

[0251] V. Operation and Maintenance and Security Policies

[0252] 1. Data Security

[0253] Transport Encryption: TLS1.3 ensures communication security.

[0254] Storage Encryption: AES-256 encryption for database fields.

[0255] Access Control: The RBAC model restricts the access rights of nursing staff.

[0256] 2. Model Monitoring

[0257] Drift Detection: Statistically analyze the input data distribution monthly. If the KL divergence > threshold, trigger retraining.

[0258] A / B Testing: Run the new model and the old version in parallel and compare the decision consistency.

[0259] VI. Verification and Testing Plan

[0260] 1. Unit Testing

[0261] Simulate device data upload to verify whether the server can correctly store and respond.

[0262] Inject abnormal data (such as negative temperature) to test the robustness of the system.

[0263] 2. Clinical Verification

[0264] Cooperate with top-three hospitals for a double-blind trial to compare the consistency between AI decisions and expert diagnoses.

[0265] Through the above phased implementation, the system development can be completed within 6 months and pass the medical certification, ultimately reducing the missed diagnosis rate of acute skin failure by more than 30%.

[0266] Obviously, those skilled in the art should understand that to implement all or part of the processes in the above embodiments, it can be completed by instructing relevant hardware through a computer program. The program can be stored in a computer-readable storage medium. When the program is executed, it can include the processes of the above control embodiments. Those skilled in the art can understand that to implement all or part of the processes in the above embodiments, it can be completed by instructing relevant hardware through a computer program. The program can be stored in a computer-readable storage medium. When the program is executed, it can include the processes of the above control embodiments. Among them, the storage medium can be a magnetic disk, an optical disk, a read-only memory (ROM), a random access memory (RAM), a flash memory, a hard disk drive (HDD), or a solid-state drive (SSD), etc.; the storage medium can also include a combination of the above types of memories.

[0267] Embodiment 2

[0268] Based on the implementation principle of Embodiment 1, on the other hand, this application proposes a non-invasive skin acute failure detection method, including the following steps:

[0269] Wear and activate the non-invasive skin acute failure detector, and establish data communication with the background server;

[0270] The non-invasive skin acute failure detector continuously collects the patient's skin temperature value, capillary arterial perfusion pressure, and skin water content, and reports them to the background server in real time;

[0271] The background server records and saves the patient's skin temperature value, capillary arterial perfusion pressure, and skin water content, and through a preset acute skin failure AI recognition model, conducts data feature analysis on the patient's skin temperature value, capillary arterial perfusion pressure, and skin water content to identify whether the patient has acute skin failure:

[0272] If so, output the corresponding acute skin failure symptom stage and corresponding intervention strategy, and send them to the nursing terminal; at the same time, send them to the non-invasive skin acute failure detector;

[0273] Otherwise, generate a skin normal notification and send it to the nursing terminal;

[0274] The nursing terminal is used to receive and display the acute skin failure symptom stage and corresponding intervention strategy, or the skin normal notification.

[0275] Please understand and implement the above steps in combination with Embodiment 1 specifically.

[0276] Each module or step of the present invention described above can be implemented by a general-purpose computing system. They can be centralized on a single computing system or distributed over a network composed of multiple computing systems. Optionally, they can be implemented by program code executable by the computing system. Thus, they can be stored in a storage system and executed by the computing system, or they can be separately fabricated into individual integrated circuit modules, or multiple modules or steps among them can be fabricated into a single integrated circuit module for implementation. In this way, the present invention is not limited to any specific combination of hardware and software.

[0277] Embodiment 3

[0278] Furthermore, on the other hand, the present application also proposes a non-invasive skin acute failure detector, system and method electronic device, including:

[0279] A processor;

[0280] A memory for storing instructions executable by the processor;

[0281] Wherein, when the processor is configured to execute the executable instructions, it implements a non-invasive skin acute failure detector, system and method described in Embodiment 2.

[0282] The electronic device of the embodiment of the present disclosure includes a processor and a memory for storing instructions executable by the processor. Wherein, when the processor is configured to execute the executable instructions, it implements a non-invasive skin acute failure detector, system and method described in the previous Embodiment 2.

[0283] Here, it should be noted that the number of processors can be one or more. At the same time, in the electronic device of the embodiment of the present disclosure, an input system and an output system can also be included. Wherein, the processor, the memory, the input system and the output system can be connected through a bus or in other ways, which is not specifically limited here.

[0284] As a computer-readable storage medium, the memory can be used to store software programs, computer-executable programs and various modules, such as: programs or modules corresponding to a non-invasive skin acute failure detector, system and method of the embodiment of the present disclosure. The processor executes various functional applications and data processing of the electronic device by running the software programs or modules stored in the memory.

[0285] The input system can be used to receive input numbers or signals. Wherein, the signal can be a key signal related to user settings and function control of the device / terminal / server. The output system can include display devices such as a display screen.

[0286] The embodiments of the present disclosure have been described above. The above description is exemplary, not exhaustive, and is not limited to the disclosed embodiments. Many modifications and variations are obvious to those of ordinary skill in the art without departing from the scope and spirit of the described embodiments. The choice of terms used herein is intended to best explain the principles of the embodiments, practical applications, or improvements to technologies in the market, or to enable other ordinary skilled persons in the art to understand the embodiments disclosed herein.

Claims

1. A non-invasive skin acute failure detector, comprising a silicone wristband, characterized in that, The non-invasive skin acute failure detection system is provided on the silicone wristband, and the non-invasive skin acute failure detector includes: A temperature sensor for collecting skin temperature signals; A photoplethysmography sensor for collecting skin pulse wave signals; A capacitance sensor for measuring skin dielectric constant; A DSP processor for signal digital-to-analog conversion processing and sending the converted value to the MCU; The MCU is used to control the sensors to collect the corresponding skin temperature value, skin pulse signal value and skin dielectric constant, perform pulse waveform analysis according to the skin pulse signal value, calculate the capillary arterial perfusion pressure at the arterial skin, and estimate the skin water content in combination with the skin dielectric constant; and, forward the skin temperature value, the capillary arterial perfusion pressure and the skin water content to the communication module in real time; The communication module is used to provide data communication between the non-invasive skin acute failure detector and the background server, including: reporting the skin temperature value, the capillary arterial perfusion pressure and the skin water content to the background server in real time through the communication module; after the background server records and analyzes the skin temperature value, the capillary arterial perfusion pressure and the skin water content, obtaining the skin acute failure detection result of the patient and sending it to the non-invasive skin acute failure detector, which is received by the communication module and forwarded to the MCU; An LED for real-time displaying the skin temperature value, the capillary arterial perfusion pressure and the skin water content of the patient and the corresponding skin acute failure detection result; A battery for power supply; The temperature sensor, the photoplethysmography sensor, and the capacitance sensor are respectively electrically connected to the DSP processor; The DSP processor, the communication module, the LED and the battery are respectively electrically connected to the MCU.

2. The non-invasive skin acute failure detector according to claim 1, wherein The capillary arterial perfusion pressure P perf is calculated as follows: Wherein: F is the amplitude of the pulse pressure signal detected by the sensor; E is the blood vessel elasticity coefficient (related to age and disease); R is the blood vessel radius (inverted through waveform analysis); η is the blood viscosity correction factor (default value is 0.15).

3. The non-invasive skin acute failure detector according to claim 1, characterized in that, The calculation method of the skin water content W is: ε s is the measured skin dielectric constant; ε d is the reference dielectric constant for dry skin (default 3.5); ε w is the dielectric constant of pure water (default 80); ρ is the tissue density correction coefficient (empirical value 0.93).

4. An in-vivo skin acute failure detection system according to claim 1, characterized in that, Including: The non-invasive skin acute failure detector according to any one of claims 1-3; A background server for recording and storing the skin temperature value, the capillary arterial perfusion pressure and the skin water content of the patient uploaded by the non-invasive skin acute failure detector; and, through a preset acute skin failure AI recognition model, performing data feature analysis on the skin temperature value, the capillary arterial perfusion pressure and the skin water content of the patient to identify whether the patient has acute skin failure: If so, output the corresponding acute skin failure symptom stage and the corresponding intervention strategy, and send them to the nursing terminal; at the same time, send them to the non-invasive skin acute failure detector; Otherwise, generate a skin normal notification and send it to the nursing terminal; A nursing terminal for receiving and displaying the acute skin failure symptom stage and the corresponding intervention strategy, or the skin normal notification; The non-invasive acute skin failure detector and the nursing terminal are respectively communicatively connected to the background server.

5. An on - non - invasive skin acute failure detection system according to claim 4, characterized in that, The method for generating the acute skin failure AI recognition model includes: Collecting the diagnostic data of a number of patients with acute skin failure; Performing data feature analysis, and statistically analyzing the skin temperature value, capillary arterial perfusion pressure, and skin water content in the diagnostic data; Performing feature annotation, and annotating the corresponding acute skin failure symptom stage and corresponding intervention strategy for the data features; Statistically analyzing the data features of each patient with acute skin failure, constructing a feature set, and dividing it into a training set and a validation set according to a preset ratio; Inputting the training set into the RF model, performing feature classification learning, and generating the acute skin failure AI recognition model; Using the validation set to verify the classification and recognition performance of the acute skin failure AI recognition model: If the verification is passed, the acute skin failure AI recognition model is deployed and applied to the background server; Otherwise, retrain.

6. A non-invasive method for detecting acute skin failure, characterized in that, It includes the following steps: Wearing and activating the non-invasive acute skin failure detector, and establishing data communication with the background server; The non-invasive acute skin failure detector continuously collects the skin temperature value, capillary arterial perfusion pressure, and skin water content of the patient, and reports them to the background server in real time; The background server records and stores the skin temperature value, the capillary arterial perfusion pressure, and the skin water content of the patient, and performs data feature analysis on the skin temperature value, the capillary arterial perfusion pressure, and the skin water content of the patient through a preset acute skin failure AI recognition model to identify whether the patient has acute skin failure: If so, output the corresponding acute skin failure symptom stage and corresponding intervention strategy, and send them to the nursing terminal; at the same time, send them to the non-invasive acute skin failure detector; Otherwise, generate a skin normal notification and send it to the nursing terminal; The nursing terminal is used to receive and display the acute skin failure symptom stage and the corresponding intervention strategy, or the skin normal notification.

7. An electronic device, characterized in that, It includes: A processor; A memory for storing processor-executable instructions; Wherein, when the processor is configured to execute the executable instructions, the method described in claim 7 is implemented.

Images