Method for monitoring cerebral apoplexy, device for monitoring cerebral apoplexy and storage medium

Abstract

The invention relates to a method for monitoring cerebral apoplexy, a device for monitoring cerebral apoplexy and a storage medium, and the method comprises the steps: collecting physiological parameters such as electroencephalogram (EEG), oxyhemoglobin saturation (SpO2), electrocardiogram (ECG) and blood pressure through wearable equipment or a family health monitoring system, and carrying out data preprocessing, feature extraction and anomaly detection; the EEG signals are processed by utilizing frequency domain characteristics and short-time energy analysis, and the cerebral apoplexy risk is identified by adopting a support vector machine (SVM) classification model. The system can analyze and detect abnormal conditions in real time, and sends out early warning information through a mobile phone App, a short message or a telephone to remind a user to see a doctor in time. The system comprises wearable equipment, a family health system, a hospital platform and a mobile health application, and is suitable for personal health monitoring and centralized monitoring of medical institutions. The real-time performance and accuracy of cerebral apoplexy monitoring are remarkably improved, the user experience is enhanced, the medical response speed is increased, and the method has wide application prospects and practical value.

Description

Technical Field

[0001] The present invention relates to the technical field of rectal cancer treatment auxiliary systems, and in particular to a method for monitoring cerebral stroke, a device for monitoring cerebral stroke and a storage medium. Background Art

[0002] Stroke is a serious neurological disease, usually caused by sudden rupture or blockage of brain blood vessels, resulting in ischemia and hypoxia of brain tissue, thus causing neurological damage. Stroke often occurs suddenly and develops rapidly, greatly affecting the patient's quality of life and may even be fatal. Therefore, early identification and timely treatment of stroke are crucial to reducing mortality and improving the patient's quality of life. However, existing stroke monitoring technologies have the following significant problems:

[0003] 1. Limitations of traditional detection methods

[0004] Currently, the diagnosis of stroke mainly relies on imaging examinations in hospitals, such as computed tomography (CT) and magnetic resonance imaging (MRI). These methods can clearly show the structure and lesions of the brain, but their main problems are:

[0005] Delayed diagnosis: CT and MRI are usually performed only after the patient develops obvious clinical symptoms, and the early symptoms of stroke may not be obvious. This delayed diagnosis will result in the best treatment time being missed when the clinical manifestations of stroke are obvious.

[0006] The equipment is expensive and inconvenient for daily monitoring: CT and MRI equipment are expensive, complicated to operate, and need to be used in a hospital environment. This limits their application in daily life, especially for high-risk groups (such as the elderly or individuals with a family history of stroke), where continuous real-time monitoring is difficult to achieve.

[0007] 2. Shortcomings of real-time monitoring technology

[0008] Although modern medical technology has seen the emergence of wearable devices such as smart watches and health monitoring bands that can monitor electrocardiogram (ECG) and blood oxygen saturation (SpO2), they still fall short in stroke monitoring:

[0009] Lack of electroencephalogram (EEG) monitoring: EEG can provide detailed information about the brain's electrical activity, which is crucial for early identification of stroke. However, existing wearable devices usually do not integrate EEG monitoring functions, which limits the detection and analysis of stroke-related brain activities.

[0010] Limitations of a single monitoring indicator: Most existing devices focus on monitoring a single physiological parameter and fail to comprehensively consider multiple physiological indicators. This limitation makes it difficult for the device to comprehensively assess the risk of stroke, thus affecting the accuracy and effectiveness of monitoring.

[0011] 3. Lack of data analysis and early warning mechanisms

[0012] In the existing technology, the data analysis and early warning system of stroke usually has the following problems:

[0013] Single indicator analysis: Traditional systems often rely on a single physiological indicator (such as electrocardiogram or blood oxygen saturation) and fail to comprehensively analyze multiple indicators. For example, the frequency domain characteristics and short-term energy analysis in the electroencephalogram signal are particularly important for early stroke identification, but many existing systems cannot effectively integrate this information.

[0014] Lack of comprehensive analysis capabilities: Existing systems usually lack effective algorithms for comprehensive analysis when processing multiple physiological data. Many systems can only provide basic abnormality detection and cannot fully assess the risk of stroke. This situation may result in inaccurate monitoring results and cannot meet the needs of real-time warning.

[0015] Imperfect early warning mechanism: Although some systems are able to issue alarms when abnormalities are detected, they lack timely and effective response mechanisms. For example, when the monitoring system detects the risk of stroke, there are often insufficient notification means (such as real-time push notifications, text messages, phone calls, etc.) to alert users and their families. This deficiency may lead to missing the best time to seek medical treatment.

[0016] 4. Low level of personalization and intelligence

[0017] Most existing stroke monitoring systems lack personalized and intelligent analysis capabilities:

[0018] Lack of personalized settings: Many systems cannot dynamically adjust based on the user's health status and historical data. Personalized settings can optimize for specific risk factors of different users, thereby improving the accuracy of monitoring.

[0019] Insufficient intelligent analysis: At present, intelligent data analysis technology has not been widely used in stroke monitoring. Intelligent analysis can combine machine learning and data mining technology to conduct in-depth analysis of a large amount of physiological data, thereby improving the ability to predict stroke risk. However, most existing systems are still at the traditional rule-based analysis stage, with a low level of intelligence.

[0020] In summary, existing stroke monitoring technologies have significant deficiencies in early detection, real-time monitoring, data analysis, early warning mechanisms, and personalized intelligence levels. These problems limit the ability to identify stroke early and effectively intervene. Therefore, developing a stroke monitoring system that integrates multiple physiological parameters and has real-time monitoring and intelligent analysis functions will significantly enhance the ability to detect stroke early and improve the health management level of patients. Summary of the invention

[0021] The purpose of the present invention is to address the deficiencies of the prior art and provide a method for monitoring cerebral stroke, a device for monitoring cerebral stroke and a storage medium to solve the problems raised in the background technology.

[0022] To achieve the above object, the present invention provides the following technical solution: a method for monitoring stroke, the method comprising the following steps:

[0023] Collecting physiological parameter data of the user through sensors, wherein the physiological parameter data includes electroencephalogram (EEG), blood oxygen saturation (SpO2), electrocardiogram (ECG) and blood pressure;

[0024] Preprocess the collected physiological parameter data, including filtering, denoising and normalization;

[0025] Extracting stroke-related features from the preprocessed physiological parameter data, wherein the features include frequency domain features, time domain features, and hemodynamic changes of EEG signals;

[0026] The frequency domain features of EEG are calculated by the following formula:

[0027]

[0028] Where P(f) represents the spectrum of the EEG signal, x(n) represents the preprocessed EEG signal, N is the number of sampling points, and f is the frequency;

[0029] The extracted features are analyzed using a pre-trained machine learning model to detect whether there are any abnormalities; when an abnormality is detected, a stroke warning signal is issued.

[0030] As a preferred technical solution of the present invention, the time domain feature extraction step of the EEG signal includes calculating the short-time energy of the EEG signal, and the energy calculation formula is:

[0031]

[0032] Among them, E(t) is the EEG energy in a certain period of time, x(n) is the EEG signal in this period of time, and N is the number of sampling points.

[0033] As a preferred technical solution of the present invention, the abnormal detection of blood oxygen saturation is performed by comparing the current blood oxygen saturation value with a set threshold value. When the blood oxygen saturation is lower than the set threshold value S threshold When the stroke risk warning is triggered, the details are as follows:

[0034] IfSpO2 threshold then trigger alarm

[0035] As a preferred technical solution of the present invention, the machine learning model is a classification model based on a support vector machine (SVM), and the SVM model classifies the data into a normal state and a stroke risk state based on physiological data features. The objective function of the SVM classification model is:

[0036]

[0037] Among them, w is the weight vector, x i is the extracted feature, y i is the classification label, C is the penalty parameter, and b is the bias term.

[0038] As a preferred technical solution of the present invention, the SVM classification model is trained by using historical stroke data and physiological data of healthy individuals, and the model parameters are optimized by using a k-fold cross-validation method. The formula of the k-fold cross-validation is:

[0039]

[0040] Among them, Error i is the classification error in the i-th fold, and k is the number of folds.

[0041] The present invention also provides a device for monitoring stroke, comprising:

[0042] Sensor module: used to collect the user's physiological parameter data, including EEG, SpO2, ECG and blood pressure;

[0043] Data processing module: used for preprocessing the physiological parameter data, including filtering, denoising and normalization;

[0044] Feature extraction module: used to extract features related to stroke;

[0045] Analysis module: using a machine learning model to analyze the features and detect whether there is a risk of stroke;

[0046] Early warning module: issues an early warning signal when stroke risk is detected.

[0047] ​As a preferred technical solution of the present invention, the sensor module is a wearable device that can collect electroencephalogram (EEG), electrocardiogram (ECG), blood oxygen saturation (SpO2) and blood pressure data in real time.

[0048] As a preferred technical solution of the present invention, the data processing module includes a filter and a normalization algorithm for removing noise from physiological signals and converting the signals into a standardized format to improve the accuracy of analysis.

[0049] The present invention also provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a computer, a method for monitoring stroke is implemented.

[0050] As a preferred technical solution of the present invention, the storage medium is a cloud server, a hard disk, a solid state drive, a USB flash drive or a CD, which is used to store and transmit the user's physiological data and analysis results.

[0051] Compared with the prior art, the present invention has the following beneficial effects:

[0052] The present invention significantly improves the real-time and convenience of stroke monitoring by adopting comprehensive monitoring of multiple physiological parameters, including electroencephalogram (EEG), blood oxygen saturation (SpO2), electrocardiogram (ECG) and blood pressure. Users can obtain relevant physiological data in real time in daily life through wearable devices or home health monitoring systems, and promptly detect possible signs of stroke. Especially in high-risk groups (such as the elderly or individuals with a family history of stroke), the system can provide early warning and reduce the risk of sudden events and potential serious consequences caused by stroke.

[0053] By extracting frequency domain features and analyzing short-time energy of EEG signals, the present invention can identify early signs of stroke with high accuracy. In addition, the support vector machine (SVM) classification model is used for abnormal detection, and the k-fold cross-validation is combined to optimize the model parameters, which significantly improves the accuracy and reliability of stroke detection. The system can not only effectively distinguish between normal and risk states, but also reduce false positives and false negatives, thereby improving the ability to predict stroke.

[0054] The early warning module in the present invention sends real-time alerts to users and their families through various means (such as mobile phone apps, text messages, and phone calls), enhancing user experience and timely response capabilities. The hospital application platform ensures that medical staff can quickly obtain patients' health data and respond through centralized monitoring and real-time early warning systems. These functions enable users and medical institutions to take countermeasures more quickly, thereby improving the efficiency of stroke prevention and intervention, and ultimately promoting the health and safety of patients. BRIEF DESCRIPTION OF THE DRAWINGS

[0055] Figure 1 A method flow chart of a method for monitoring stroke, a device for monitoring stroke, and a storage medium;

[0056] Figure 2 The present invention is a diagram showing the structure of a method for monitoring stroke, a device for monitoring stroke and a storage medium. DETAILED DESCRIPTION

[0057] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. 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 creative work are within the scope of protection of the present invention.

[0058] The specific implementation of the present invention is described in detail below in conjunction with the accompanying drawings and multiple embodiments.

[0059] In order to better explain the embodiments of the present invention, the following describes in detail the method and system for predicting the efficacy of neoadjuvant chemoradiotherapy for rectal cancer based on magnetic resonance image analysis through several specific examples. Each example covers the complete process from image acquisition and processing to efficacy prediction.

[0060] Example 1: Wearable stroke monitoring device

[0061] Equipment composition:

[0062] Sensor modules: These include electroencephalogram (EEG) sensors, blood oxygen sensors, electrocardiogram (ECG) sensors, and blood pressure sensors. These sensors are integrated into a portable wearable device such as a head-mounted device, watch, or patch.

[0063] Data processing module: Built into the wearable device, it is responsible for filtering and denoising the collected raw data. The EEG signal is processed using a digital filter (such as a low-pass filter) to remove high-frequency noise.

[0064] Feature extraction module: Perform frequency domain analysis on EEG signals and extract EEG spectrum features P(f):

[0065]

[0066] And calculate the short-term energy E(t) of the EEG signal:

[0067]

[0068] Analysis module: Use the SVM classification model to analyze the extracted features. The SVM model uses the following objective function for classification:

[0069]

[0070] Among them, w is the weight vector, x i Characterized by y i is the classification label, C is the penalty parameter, and b is the bias term.

[0071] Early warning module: When the analysis module detects an abnormal signal, it sends real-time early warning information through the mobile phone app, suggesting that users seek medical treatment in time.

[0072] Usage scenario: Users wear the device in their daily lives, and the device monitors their physiological state in real time and provides early warnings when signs of stroke are detected. This device is particularly suitable for high-risk groups such as the elderly or individuals with a family history of stroke.

[0073] Example 2: Home Health Monitoring System

[0074] System composition:

[0075] Sensor system: Consists of an integrated health monitoring device that includes an EEG sensor, a blood oxygen sensor, an ECG sensor, and a blood pressure sensor, installed in the patient's home environment, such as a bedside or living room.

[0076] Data processing unit: connected to the sensor system, responsible for receiving and processing physiological data. The data processing unit has built-in filters and denoising algorithms to ensure data accuracy.

[0077] Feature extraction and analysis system: extracts the frequency domain features and short-term energy of the EEG, and analyzes abnormalities through the SVM model. The model uses the k-fold cross-validation method to optimize parameters:

[0078]

[0079] Among them, Error i is the classification error in the i-th fold, and k is the number of folds.

[0080] Notification module: When the system detects the risk of stroke, it notifies the patient and his / her family members via SMS, phone call or App push, provides emergency advice and recommends medical treatment.

[0081] Use scenarios:

[0082] Suitable for elderly people or patients with chronic diseases in the family, through the home health monitoring system for 24 hours real-time monitoring. The system can process data in real time and immediately notify family members or medical institutions when abnormalities are detected.

[0083] Example 3: Hospital Stroke Monitoring Platform

[0084] Platform composition:

[0085] Multi-point sensor network: The hospital sets up multiple monitoring points, including wards, emergency rooms, etc., equipped with EEG sensors, blood oxygen sensors, ECG sensors and blood pressure sensors.

[0086] Central data processing system: responsible for collecting data from each monitoring point, filtering, denoising and normalizing the data. The frequency domain characteristics of the EEG signal are calculated using the following formula:

[0087]

[0088] Where P(f) represents the spectrum of the EEG signal, x(n) represents the preprocessed EEG signal, N is the number of sampling points, and f is the frequency;

[0089] Advanced feature analysis system: Uses SVM models and other machine learning algorithms to comprehensively analyze the extracted features and assess the patient's stroke risk. The system records and analyzes a large amount of patient data and continuously optimizes the model.

[0090] Real-time early warning and reporting system: The system automatically generates risk reports and sends them to doctors and nurses through the hospital's internal network. When a high-risk patient is detected, an alarm is immediately issued and relevant data and suggestions are provided.

[0091] Usage scenario: Applicable to centralized monitoring of high-risk patients in hospitals. Doctors can obtain patients' health data in real time and respond quickly when potential stroke risks are discovered.

[0092] Example 4: Mobile health application

[0093] Application composition:

[0094] Smart bracelet: integrated with EEG sensor, blood oxygen sensor, electrocardiogram sensor and blood pressure sensor for daily monitoring.

[0095] Mobile application: Connect with smart bracelet, transmit data via Bluetooth, and perform data processing, feature extraction and anomaly detection on the mobile phone. The data analysis module in the application uses the SVM classification model for real-time analysis.

[0096] Data storage and analysis: Store users' health data and perform historical data analysis to detect long-term trends and abnormal changes.

[0097] Early warning system: When abnormal data is detected, the app will notify users via push notifications and provide health advice and medical advice.

[0098] Usage scenario: Suitable for personal health management. Users can monitor their health status in real time through mobile applications and receive timely warnings and guidance when abnormalities are found.

[0099] The contents not described in detail in this specification belong to the prior art known to professional and technical personnel in the field. Although the present invention has been described in detail with reference to the aforementioned embodiments, it is still possible for those skilled in the art to modify the technical solutions described in the aforementioned embodiments, or to make equivalent substitutions for some of the technical features therein. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A method for monitoring stroke, characterized in that: The method comprises the following steps: The user's physiological parameter data is collected through sensors, wherein the physiological parameter data includes electroencephalogram (EEG), blood oxygen saturation (SpO2), electrocardiogram (ECG) and blood pressure; Preprocess the collected physiological parameter data, including filtering, denoising and normalization; Extracting stroke-related features from the preprocessed physiological parameter data, wherein the features include frequency domain features, time domain features, and hemodynamic changes of EEG signals; The frequency domain features of EEG are calculated by the following formula: Where P(f) represents the spectrum of the EEG signal, x(n) represents the preprocessed EEG signal, N is the number of sampling points, and f is the frequency; The extracted features are analyzed using a pre-trained machine learning model to detect whether there are any abnormalities; when an abnormality is detected, a stroke warning signal is issued.

2. The method for monitoring stroke according to claim 1, characterized in that: The time domain feature extraction step of the EEG signal includes calculating the short-time energy of the EEG signal, and the energy calculation formula is: Among them, E(t) is the EEG energy in a certain period of time, x(n) is the EEG signal in this period of time, and N is the number of sampling points.

3. The method for monitoring stroke according to claim 1 or 2, characterized in that: The abnormal detection of blood oxygen saturation is performed by comparing the current blood oxygen saturation value with the set threshold value. When the blood oxygen saturation is lower than the set threshold value S threshold When the stroke risk warning is triggered, the details are as follows: ifSpO2<S threshold then trigger alarm。 4. The method for monitoring stroke according to claim 1, characterized in that: The machine learning model is a classification model based on a support vector machine (SVM). The SVM model classifies data into a normal state and a stroke risk state based on physiological data features. The objective function of the SVM classification model is: Among them, w is the weight vector, x i is the extracted feature, y i is the classification label, C is the penalty parameter, and b is the bias term.

5. The method for monitoring stroke according to claim 4, characterized in that: The SVM classification model is trained by using historical stroke data and physiological data of healthy individuals, and the model parameters are optimized by using the k-fold cross-validation method. The formula of the k-fold cross-validation is: Among them, Error i is the classification error in the i-th fold, and k is the number of folds.

6. A device for monitoring stroke, comprising: Sensor module: used to collect the user's physiological parameter data, including EEG, SpO2, ECG and blood pressure sensors; Data processing module: used for preprocessing the physiological parameter data, including filtering, denoising and normalization; Feature extraction module: used to extract features related to stroke; Analysis module: using a machine learning model to analyze the features and detect whether there is a risk of stroke; Early warning module: issues an early warning signal when stroke risk is detected.

7. The device for monitoring stroke according to claim 6, characterized in that: The sensor module is a wearable device that can collect electroencephalogram (EEG), electrocardiogram (ECG), blood oxygen saturation (SpO2) and blood pressure data in real time.

8. The device for monitoring stroke according to claim 6 or 7, characterized in that: The data processing module includes filters and normalization algorithms for removing noise from physiological signals and converting the signals into a standardized format.

9. A computer-readable storage medium for executing any one of the methods of claims 1 to 5, wherein the computer-readable storage medium stores a computer program, and the computer program implements any one of the methods of claims 1 to 5 when executed by a computer.

10. The computer-readable storage medium according to claim 9, wherein: The storage medium is a cloud server, a hard disk, a solid state drive, a USB flash drive or a CD, which is used to store and transmit the user's physiological data and analysis results.

Images