A patient physiological information monitoring system

Abstract

The present invention discloses a patient physiological information monitoring system, which specifically relates to the field of medical information monitoring, and is used to solve the problems of the accuracy of patient physiological monitoring data and the efficiency of nursing task allocation. The system obtains the patient's electrocardiogram data in real time, automatically identifies and corrects the electrode wearing status, and ensures the reliability of signal acquisition and the accuracy of data. Deep features are extracted through multi-level preprocessing and noise filtering technology to generate early warning signals with clinical indication significance, indicating potential health risks. The nurse's physiological indicators are monitored in real time through a smart watch, and a dynamic analysis is performed in combination with a multi-parameter integrated stress assessment model to generate an instant stress index to assess the nurse's work stress status. The time sensitivity of the early warning signal, the nurse's skill matrix and the current work stress status are collected, and the task fitness index is comprehensively calculated. High fitness tasks are allocated preferentially and the nurse status and skill matrix are updated in real time. Task allocation is dynamically optimized to improve overall nursing efficiency and quality.

Description

Technical Field

[0001] The present invention relates to the field of medical information monitoring, and more specifically, to a patient physiological information monitoring system. Background Art

[0002] In modern medical environments, monitoring instruments in wards are usually connected to central monitors via network cables. This configuration allows the central monitor to collect and process physiological data from monitoring instruments in each ward. Furthermore, the central monitor is connected to the medical staff's wristwatch via wireless communication to achieve seamless data transmission. This setting allows medical staff to view the patient's physiological characteristics, such as heart rate, blood pressure and other key vital signs, through the wristwatch anytime and anywhere, thereby improving the real-time monitoring ability of the patient's status and ensuring timely response to various medical needs. Although this monitoring system is designed to be highly convenient and effective, there are still some technical deficiencies in actual application, especially those related to the wearing of monitoring sensors and data accuracy.

[0003] During the initial setup, minor wearing errors of electrodes or other sensors, such as position deviation or poor contact, may not be detected immediately. This initial wearing error may gradually appear in the monitoring data over time, such as abnormal data fluctuations or deviations from the normal range, affecting the accuracy and reliability of the data. In addition, noise interference and signal instability are also challenges faced by current systems, affecting the clarity and accuracy of physiological data.

[0004] In terms of work stress management for nursing staff, the existing system also lacks real-time physiological data analysis and stress status assessment methods, and cannot dynamically adjust task allocation, which in turn affects nursing quality and work efficiency. The traditional task allocation method fails to fully consider the time sensitivity of the task, the skill matching of the nursing staff, and the immediate work stress status, resulting in unbalanced resource allocation and unreasonable workload.

[0005] In order to solve the above problems, a technical solution is now provided. Summary of the invention

[0006] In order to overcome the above-mentioned defects of the prior art, an embodiment of the present invention provides a patient physiological information monitoring system, which obtains the patient's ECG data in real time, and applies comprehensive abnormality detection technology to ensure the high quality and high integrity of the data, automatically identifies and corrects the electrode wearing status, and ensures the reliability of signal acquisition and the accuracy of the data. When the system does not obtain an early warning reminder during data acquisition, it extracts deep features through multi-level preprocessing and noise filtering technology, generates early warning signals with clinical indication significance, and prompts potential health risks. Through the smart watch, the nurse's physiological indicators are monitored in real time, and the multi-parameter integrated pressure assessment model is combined for dynamic analysis, and an instant pressure index is generated to assess the nurse's work pressure state, and stress problems are discovered and intervened in time. The system collects the time sensitivity of the early warning signal, the nurse's skill matrix and the current workload data, establishes a time sensitivity scoring model and a skill matching model, comprehensively calculates the task fitness index, prioritizes the high fitness tasks and updates the nurse's status and skill matrix in real time, dynamically optimizes the task allocation, and improves the overall nursing efficiency and quality to solve the problems raised in the above background technology.

[0007] To achieve the above object, the present invention provides the following technical solutions:

[0008] A patient physiological information monitoring system, comprising: a data aggregation module, a signal purification module, a feature analysis module, a pressure analysis module, a task adaptation module and a task allocation module;

[0009] Data aggregation module: obtains the patient's ECG data in real time through the ECG monitor, transmits the data to the central monitor, and then passes it to the signal purification module;

[0010] Signal purification module: Apply comprehensive anomaly detection technology in the central monitor to suppress noise and purify ECG signals, automatically identify and correct electrode wearing status, decide whether to issue a warning signal, and pass the purified ECG signals and decision results to the feature analysis module;

[0011] Feature analysis module: when no warning reminder is received during data collection, adaptive baseline correction technology and adaptive bandpass filter are used to ensure the clarity and stability of the signal, and deep features are extracted through heart rate variability analysis and ST segment dynamic change detection. Warning signals are issued for the extracted features, and the extracted ECG feature data are transmitted to the pressure analysis module;

[0012] Stress analysis module: monitors nurses' physiological parameters in real time through a wristwatch, analyzes dynamic physiological data based on a multi-parameter integrated stress assessment model, generates an instant stress index to assess the nurses' current work stress status, and transmits the nurses' instant stress index to the task adaptation module;

[0013] Task adaptation module: Using the rule engine and multi-factor scheduling algorithm, based on the time sensitivity of the warning signal, the nurse skill matrix and the immediate work pressure status, the task adaptation index model is constructed to analyze the adaptability of the task and the nurse, and the task adaptation index is transmitted to the task allocation module;

[0014] Task assignment module: By sorting all combinations of tasks and nurses, and assigning the most suitable tasks to the corresponding nurses based on the task fitness index, the nurses' work stress status and skill matrix are updated in real time, the task assignment process is dynamically optimized, and the updated nurses' work stress status and skill matrix data are fed back to the pressure analysis module and task adaptation module.

[0015] In a preferred embodiment, step S1 includes the following contents:

[0016] S1-1, obtaining the patient's ECG data in real time through an ECG monitor;

[0017] S1-2, the central monitor performs noise frequency domain characteristic analysis on the acquired ECG data;

[0018] S1-3, evaluate the amplitude fluctuation of the purified signal, calculate the RMS value and peak factor of the signal, and apply the dynamic segmentation method to identify abnormal amplitude changes;

[0019] S1-4, further performing wavelet transform on the signal to obtain the time-frequency distribution information of the signal, and identifying transient interference by analyzing the wavelet coefficients;

[0020] S1-5, apply consistency detection algorithm to evaluate the periodicity and stability of the signal, automatically identify and correct the electrode wearing status, and ensure data integrity. The specific steps include:

[0021] The consistency detection algorithm is used to evaluate the periodicity and stability of the signal and determine the wearing status of the electrode;

[0022] The signal consistency index is calculated using the following formula:

[0023] Among them, C is the consistency index; x(t i ) represents the signal value of the i-th sampling point; |x(t i )-μ| represents the absolute difference between the signal value and the mean; N is the total number of sampling points; σ is the standard deviation of the signal;

[0024] Convert the consistency index to a stability index ranging from 0 to 1:

[0025] Among them, S is the stability index; max(C) is the maximum value of the consistency index;

[0026] When the stability index is close to 1, it means that the signal is stable and the electrode is in good wearing condition; when the stability index is significantly lower than 1, it means that the signal fluctuates greatly and an early warning reminder is issued.

[0027] In a preferred embodiment, step S2 includes the following contents:

[0028] S2-1, when no warning reminder is obtained during data collection, the collected ECG data is used for multi-level preprocessing and noise filtering:

[0029] S2-1-1, using adaptive baseline correction technology, the original ECG signal is processed to remove the baseline drift by calculating the local average and subtracting the low-frequency components in the signal;

[0030] S2-1-2, then, applying an adaptive bandpass filter to the signal after removing the baseline drift to remove high-frequency and low-frequency noise;

[0031] S2-2, detect heart rate variability and ST segment dynamic changes on the preprocessed signal:

[0032] S2-2-1, in the heart rate variability analysis, by calculating the RR interval, and performing adaptive correction and feature extraction;

[0033] Next, the standard deviation and range of the RR interval were calculated;

[0034] S2-2-2, detecting the ST segment in the ECG signal by calculating the voltage elevation value of the ST segment relative to the PR segment;

[0035] S2-3, using classification algorithms, conduct in-depth analysis and multi-dimensional early warning evaluation of the extracted features:

[0036] S2-3-1, combining the comprehensive index of heart rate variability and the time-weighted average of ST segment characteristics into a vector;

[0037] S2-3-2, use the threshold function to transform and normalize the features, generate early warning signals through comprehensive evaluation, and obtain the predictive factors of dynamic abnormality of ECG signals:

[0038]

[0039] Among them, CAPF is the predictive factor of dynamic abnormality of ECG signal; θ is the threshold function, which is used to generate warning signal, indicating whether the warning is triggered; a represents the characteristic normalization coefficient, which represents the normalization of HRV comprehensive index; b represents the characteristic normalization coefficient, which represents the normalization of ST segment characteristics;

[0040] S2-4, the ECG signal dynamic abnormality predictive factor needs to be compared with the low-risk threshold and the high-risk threshold, and different levels of warning signals are generated according to the comparison results; the specific implementation process is as follows: if the ECG signal dynamic abnormality predictive factor value is less than the low-risk threshold, a green warning signal is generated, indicating a low risk; if the ECG signal dynamic abnormality predictive factor value is between the low-risk threshold and the high-risk threshold, a yellow warning signal is generated, indicating a medium risk; if the ECG signal dynamic abnormality predictive factor value is greater than the high-risk threshold, a red warning signal is generated, indicating a high risk.

[0041] In a preferred embodiment, S3 includes the following contents:

[0042] S3-1, use the wristwatch to monitor the nurse's physiological indicators in real time, including heart rate, blood pressure and skin galvanic response, and pre-process the collected data:

[0043] S3-1-1, performing sliding average on the heart rate data collected in real time;

[0044] S3-1-2, filtering the blood pressure data collected in real time;

[0045] S3-1-3, baseline correction of the real-time collected galvanic skin response data;

[0046] S3-2, extract relevant features from preprocessed physiological data:

[0047] S3-2-1, calculating the standard deviation of the heart rate data from the smoothed heart rate data as a heart rate variability indicator;

[0048] S3-2-2, extracting the mean and standard deviation from the filtered blood pressure data;

[0049] S3-2-3, extract the mean value and the rate of change between adjacent sampling points from the baseline-corrected galvanic skin response data;

[0050] S3-3, the characteristic values ​​of heart rate variability, blood pressure and skin electrical response are weighted and summed to generate an instant stress index, which is used to reflect the current work stress state of the nurse.

[0051] In a preferred embodiment, S4 includes the following contents:

[0052] S4-1, first, collect the time sensitivity of warning signals, nurses’ skill matrix, and current work stress data;

[0053] S4-1-1, for each warning signal, assign a time sensitivity score based on its urgency:

[0054]

[0055] Among them, S l represents the time sensitivity score corresponding to the warning signal l, that is, the time sensitivity score of task l, indicating the urgency of the warning signal; W p Indicates the weight of the warning signal, reflecting the importance of the warning signal; T e Indicates the expiration time of the warning signal, indicating the latest time when the task needs to be completed; T n Indicates the current time, indicating the current time node;

[0056] S4-1-2, construct a nurse skill matrix and calculate the skill matching between nurses and tasks:

[0057]

[0058] Among them, M ov represents the skill matching degree between nurse o and task l; K represents the total number of skills; SK ok″ represents the proficiency of nurse o in skill k″; TM lk″ represents the requirement of skill k″ required for task l;

[0059] S4-2, calculate the task suitability index assigned to each nurse for each task corresponding to each warning signal, taking into account time sensitivity, skill matching and immediate stress status. The calculation formula is as follows:

[0060]

[0061] Among them, CS ol It represents the task fitness index of task l assigned to nurse o;

[0062] S l represents the time sensitivity score of task l;

[0063] M ov represents the skill match between nurse o and task l;

[0064] YL o It represents the immediate stress index of nurse o;

[0065] τ represents the adjustment parameter used to control the effect of stress state on the score;

[0066] S4-1-2, construct a nurse skill matrix and calculate the skill matching between nurses and tasks:

[0067]

[0068] Among them, M ov represents the skill matching degree between nurse o and task l; K represents the total number of skills; SK ok″represents the proficiency of nurse o in skill k″; TM lk″ represents the requirement of skill k″ required for task l.

[0069] In a preferred embodiment, the weight of the warning signal is calculated based on the dynamic abnormality prediction factor of the ECG signal. The specific process is: first, the dynamic abnormality prediction factor value of the ECG signal of each warning signal is calculated, and then the dynamic abnormality prediction factor values ​​of the ECG signal of all warning signals are normalized so that their sum is 1. The weight of each warning signal is equal to its dynamic abnormality prediction factor value of the ECG signal divided by the sum of the dynamic abnormality prediction factor values ​​of the ECG signal of all warning signals.

[0070] In a preferred embodiment, S4-2, the task suitability index of each nurse assigned to each task corresponding to each warning signal is calculated, taking into account time sensitivity, skill matching and immediate stress status, and the calculation formula is as follows:

[0071]

[0072] Among them, CS ol It represents the task fitness index of task l assigned to nurse o;

[0073] S l represents the time sensitivity score of task l;

[0074] M ov represents the skill match between nurse o and task l;

[0075] YL o It represents the immediate stress index of nurse o;

[0076] τ represents the adjustment parameter used to control the impact of stress state on the score.

[0077] In a preferred embodiment, S5 includes the following contents:

[0078] S5-1, sort all combinations of tasks and nurses, and sort the tasks from high to low according to the calculated task fitness index;

[0079] S5-2, the task with the highest task suitability index is assigned to the corresponding nurse first, and a prompt processing signal is sent to the wristwatch of the corresponding nurse, and the assignment is performed according to the following steps:

[0080] S5-2-1, select the task and nurse combination with the highest fit index;

[0081] S5-2-2, assign the task to the corresponding nurse, and the central monitor sends a signal to the watch, prompting the nurse bound to the watch to handle it and record the task assignment;

[0082] S5-2-3, updating the immediate work pressure status of the assigned nurses, and re-execute step S3 for calculation;

[0083] Update the nurse's skill matrix and re-execute step S4-1-2 to calculate the skill matching between the nurse and the task;

[0084] S5-3, repeat step S5-2 to continue to assign the remaining tasks, each time selecting the task and nurse combination with the highest fitness index, and perform task assignment and status update until all tasks are assigned.

[0085] The technical effects and advantages of the patient physiological information monitoring system of the present invention are as follows:

[0086] 1. The present invention obtains the patient's ECG data in real time and applies comprehensive anomaly detection technology, including noise frequency domain feature analysis, signal amplitude fluctuation assessment, and complex spectrum characteristic analysis, to ensure high data quality and integrity. Automatic identification and correction of electrode wearing status further improves the reliability of signal acquisition and ensures data accuracy. This process can effectively filter out various types of noise and interference, extract meaningful physiological characteristics, and ensure the accuracy and reliability of subsequent data analysis and clinical evaluation. In addition, through the intelligent feedback mechanism, the system can monitor and adjust in real time to ensure the continued good state of electrode wearing, thereby significantly improving the overall effect and clinical application value of ECG monitoring.

[0087] 2. When the present invention does not obtain early warning reminders during data collection, it uses multi-level preprocessing and noise filtering, uses adaptive baseline correction technology to remove baseline drift, and applies adaptive bandpass filters to remove high-frequency and low-frequency noise to ensure the clarity and stability of the signal. It then performs heart rate variability analysis and ST segment dynamic change detection to extract deep features, and uses classification algorithms to perform multi-dimensional early warning evaluation on the extracted features to generate early warning signals with clinical indication significance. By comparing the dynamic abnormal predictive factors of the ECG signal with the risk threshold, early warning signals of different levels are generated to indicate potential health risks and help clarify the priority of task processing.

[0088] 3. The present invention uses a smart watch to monitor the nurses' heart rate, blood pressure, skin galvanic response and other physiological indicators in real time, and pre-processes the data to ensure its accuracy and stability, extracts relevant features from it, combines a multi-parameter integrated stress assessment model, performs dynamic physiological data analysis, and generates a real-time stress index to assess the nurses' current work stress status, thereby timely discovering and intervening in the nurses' work stress problems, facilitating dynamic decision-making for later task allocation, and improving work efficiency and nursing quality.

[0089] 4. The present invention establishes a time sensitivity scoring model and a nurse skill matching model by collecting data such as the time sensitivity of early warning signals, the nurse skill matrix and the current workload, and calculates the task fitness index based on these models by comprehensively considering the time sensitivity, skill matching and immediate stress status. Tasks with the highest fitness index are preferentially assigned to corresponding nurses, and the nurses' work stress status and skill matrix are updated in real time to dynamically optimize the task allocation process, thereby achieving efficient and reasonable task allocation and improving overall nursing efficiency and quality. BRIEF DESCRIPTION OF THE DRAWINGS

[0090] Figure 1 A schematic diagram of a flow chart of a patient physiological information monitoring system of the present invention;

[0091] Figure 2 The present invention is a schematic diagram of data transmission between an electrocardiogram monitor, a central monitor and a wristwatch in a patient physiological information monitoring system. DETAILED DESCRIPTION

[0092] 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.

[0093] Example 1

[0094] Figure 1 The present invention provides a patient physiological information monitoring system, comprising: a data aggregation module, a signal purification module, a feature analysis module, a pressure analysis module, a task adaptation module and a task allocation module;

[0095] Data aggregation module: obtains the patient's ECG data in real time through the ECG monitor, transmits the data to the central monitor, and passes the ECG data to the signal purification module;

[0096] Signal purification module: Apply comprehensive anomaly detection technology in the central monitor to suppress noise and purify ECG signals, automatically identify and correct electrode wearing status, decide whether to issue a warning signal, and pass the purified ECG signals and decision results to the feature analysis module;

[0097] Feature analysis module: when no warning reminder is received during data collection, adaptive baseline correction technology and adaptive bandpass filter are used to ensure the clarity and stability of the signal, and deep features are extracted through heart rate variability analysis and ST segment dynamic change detection. Warning signals are issued for the extracted features, and the extracted ECG feature data are transmitted to the pressure analysis module;

[0098] Stress analysis module: monitors nurses' physiological parameters in real time through a wristwatch, analyzes dynamic physiological data based on a multi-parameter integrated stress assessment model, generates an instant stress index to assess the nurses' current work stress status, and transmits the nurses' instant stress index to the task adaptation module;

[0099] Task adaptation module: Using the rule engine and multi-factor scheduling algorithm, based on the time sensitivity of the warning signal, the nurse skill matrix and the immediate work pressure status, the task adaptation index model is constructed to analyze the adaptability of the task and the nurse, and the task adaptation index is transmitted to the task allocation module;

[0100] Task assignment module: By sorting all combinations of tasks and nurses, and assigning the most suitable tasks to the corresponding nurses based on the task fitness index, the nurses' work stress status and skill matrix are updated in real time, the task assignment process is dynamically optimized, and the updated nurses' work stress status and skill matrix data are fed back to the pressure analysis module and task adaptation module to ensure dynamic optimization.

[0101] Step S1 includes the following contents:

[0102] S1-1, the patient's electrocardiogram (ECG) data is obtained in real time through an ECG monitor. The ECG data is recorded in the form of a time series, and each data point contains a voltage value and a timestamp; then the ECG data is sent to the central monitor.

[0103] S1-2, the central monitor performs noise frequency domain feature analysis on the acquired ECG data, obtains the main noise frequency components through spectrum analysis, and applies multi-level filtering technology to suppress noise based on the frequency domain analysis, and outputs the purified signal. The specific steps include:

[0104] The time domain signal is converted into a frequency domain signal, and the main noise frequency components are identified through spectrum analysis.

[0105] Use filtering technology to remove noise of specific frequencies, such as power frequency interference and myoelectric interference, etc. Combine multiple filtering technologies (such as bandpass filtering, low-pass filtering, etc.) to perform multi-level noise suppression.

[0106] S1-3, evaluate the amplitude fluctuation of the purified signal, calculate the root mean square value and peak factor of the signal, and apply the dynamic segmentation method to identify abnormal amplitude changes and evaluate the amplitude fluctuation of the signal. The specific steps include:

[0107] Calculate the root mean square (RMS) value of the signal to assess the overall energy level of the signal.

[0108] Calculate the crest factor (CrestFactor) and identify abnormal peaks in the signal by the crest factor.

[0109] The signal is segmented and the fluctuation range of each segment is calculated to identify areas with abnormal amplitude changes.

[0110] S1-4, further perform wavelet transform on the signal to obtain the time-frequency distribution information of the signal, and identify transient interference by analyzing the wavelet coefficients. The specific steps include:

[0111] Wavelet transform is used to decompose the signal at multiple scales to obtain the spectrum information of the signal at different time points.

[0112] Analyze the time-frequency distribution of wavelet coefficients and identify transient interference and non-stationary noise.

[0113] S1-5, apply consistency detection algorithm to evaluate the periodicity and stability of the signal, automatically identify and correct the electrode wearing status, and ensure data integrity. The specific steps include:

[0114] The periodicity and stability of the signal are evaluated through the consistency detection algorithm to determine the wearing status of the electrode.

[0115] The consistency index of the signal is calculated using the following formula to evaluate the stability of the signal in different time periods.

[0116]

[0117] Among them, C is the consistency index; x(t i ) represents the signal value of the i-th sampling point; |x(t i )-μ| represents the absolute difference between the signal value and the mean; N is the total number of sampling points; σ is the standard deviation of the signal.

[0118] The consistency index indicates the degree of deviation of the signal value from the mean. The smaller the value, the more stable the signal.

[0119] To facilitate judgment, the consistency index is converted into a stability index ranging from 0 to 1. The larger the value, the more stable the signal.

[0120]

[0121] Among them, S is the stability index; max(C) is the maximum value of the consistency index, which is the set threshold.

[0122] According to the calculated stability index, the wearing status of the electrode is judged:

[0123] When the stability index is close to 1, it means the signal is stable and the electrode is in good condition.

[0124] When the stability index is significantly lower than 1, it means that the signal fluctuates greatly and an early warning reminder is issued. The electrode may be loose or poorly worn.

[0125] Step S1 combines the consistency detection and stability evaluation of the signal, specifically by calculating the mean, standard deviation and consistency index of the signal to evaluate the fluctuation of the signal. If there is a fault in the ECG monitor, it will usually cause instability of all signals, not just fluctuations in a certain period of time. At the same time, combined with historical data, the possibility of equipment failure can be ruled out. Abnormalities in the patient's ECG itself usually have clinical characteristics and regularities, such as arrhythmias, rather than random fluctuations. Through the calculation of consistency index and stability index, the stability of the signal in different time periods can be effectively identified. If the signal consistency is high and the stability index is close to 1, it indicates that the signal fluctuation is small and the electrode wearing state is good; if the signal consistency is poor and the stability index is significantly lower than 1, it may be caused by poor electrode wearing. Signal fluctuations. This method combines dynamic threshold adjustment and multi-scale analysis to further eliminate interference from other factors, and monitors and corrects the electrode state in real time through an intelligent feedback mechanism to ensure the accuracy and uniqueness of the judgment.

[0126] The present invention obtains the patient's ECG data in real time and applies comprehensive anomaly detection technology, including noise frequency domain feature analysis, signal amplitude fluctuation assessment, and complex spectrum characteristic analysis, to ensure high quality and integrity of the data. Automatic identification and correction of electrode wearing status further improves the reliability of signal acquisition and ensures data accuracy. This process can effectively filter out various types of noise and interference, extract meaningful physiological features, and ensure the accuracy and reliability of subsequent data analysis and clinical evaluation. In addition, through the intelligent feedback mechanism, the system can monitor and adjust in real time to ensure the continued good state of electrode wearing, thereby significantly improving the overall effect and clinical application value of ECG monitoring.

[0127] Step S2 includes the following contents:

[0128] S2-1, when no warning reminder is received during data collection, the collected ECG data is used for multi-level preprocessing and noise filtering to ensure the clarity and accuracy of the signal:

[0129] S2-1-1, using adaptive baseline correction technology, the original ECG signal is processed to remove the baseline drift by calculating the local average and subtracting the low-frequency components in the signal. The specific formula is as follows:

[0130]

[0131] in, is the signal after removing the baseline drift, which represents the ECG signal value at time t after removing the baseline drift; x(t) is the original signal, which represents the ECG signal value at time t; L is the filter window length; w(k) is the adaptive weight function, which represents the weight, which is dynamically adjusted with k to adapt to different signal characteristics.

[0132] The local average is calculated by adaptive weighting and subtracted from the original signal to remove low-frequency baseline drift and ensure signal stability.

[0133] S2-1-2, then, apply an adaptive bandpass filter to the signal after removing the baseline drift to remove high-frequency and low-frequency noise. The characteristics of the filter are dynamically adjusted according to the local frequency components of the signal to ensure that the main frequency components of the ECG signal are retained while effectively suppressing noise interference. The specific formula is as follows:

[0134]

[0135] Among them, y(t) is the filtered signal, which represents the value of the filtered ECG signal at time t; h(k,t) is the time-varying impulse response of the bandpass filter, which represents the response function of the filter at time t and moment k, and dynamically adjusts the filter characteristics to adapt to different frequency components; N is the total length of the signal, which represents the response range of the filter.

[0136] Through the adaptive bandpass filter, the ECG signal is convolved to dynamically suppress high-frequency and low-frequency noise and retain the effective frequency components.

[0137] S2-2, detect heart rate variability and ST segment dynamic changes on the preprocessed signal and extract deep features:

[0138] S2-2-1, in the heart rate variability analysis, the RR interval (the time interval between R waves) is calculated, and adaptive correction and feature extraction are performed. The adaptively corrected RR interval and comprehensive indexes reflect the short-term variability and overall fluctuation of the heart rate.

[0139]

[0140] Among them, RR i represents the i-th RR interval, which is the time interval between the i-th R wave and the i+1-th R wave, including the adaptive correction term; The time of the i-th R wave indicates the moment when the i-th R wave occurs; The time of the i+1th R wave indicates the time when the i+1th R wave occurs; It represents the average value of RR intervals and the average time interval of all RR intervals; α is the adaptive coefficient, which is adjusted according to the fluctuation of heart rate and represents the size of the correction amount.

[0141] Next, the standard deviation and range of the RR interval are calculated to comprehensively reflect the short-term variability and overall fluctuation of heart rate. Adaptive correction coefficients are used to dynamically adjust according to the fluctuation of heart rate to improve the accuracy of HRV analysis.

[0142]

[0143] Among them, HRV index is a comprehensive index of HRV, which indicates a comprehensive measure of heart rate variability; N′ is the total number of RR intervals, which indicates the total number of RR intervals; β is a weight coefficient, which is used to balance the standard deviation and the range, indicating the adjustment of the influence of these two parts; It is the square of the difference between the i-th RR interval and the average RR interval, indicating the degree of fluctuation of the RR interval; max(RR) is the maximum value of the RR interval, indicating the maximum time interval among all RR intervals; min(RR) is the minimum value of the RR interval, indicating the minimum time interval among all RR intervals.

[0144] S2-2-2, detect the ST segment in the ECG signal, calculate the voltage elevation value of the ST segment relative to the PR segment, and use the periodic correction term to identify the possibility of myocardial ischemia or damage.

[0145]

[0146] Among them, ST elevation (t) represents the ST segment elevation value, which represents the voltage elevation value of the ST segment relative to the PR segment at time t; is the average voltage value of the PR segment, which is used as a benchmark and indicates the average voltage level of the PR segment; M is the number of sampling points of the ST segment, which indicates the number of sampling points when calculating the ST segment elevation value; x(t+k′) is the ECG signal value at time t+k′, which indicates the voltage value of the ST segment; γ is the correction coefficient, which adjusts the periodic fluctuation of the elevation value and indicates the magnitude of the correction; sin(2πft) is a periodic function, which indicates the periodic change of the signal; f is the correction frequency, which reflects the periodic change frequency of the signal.

[0147] The correction term is adjusted according to the periodic changes of the signal, making the detection of ST segment changes more sensitive and accurate.

[0148] S2-3, using classification algorithms, conduct in-depth analysis and multi-dimensional early warning evaluation of the extracted features to generate early warning signals with clinical significance:

[0149] S2-3-1, construct the HRV and ST segment features into a feature vector. Specifically, the time-weighted average of the heart rate variability index and the ST segment features is combined into a vector:

[0150]

[0151] Where F is the feature vector, which represents the combination of heart rate variability and ST segment features; δ is the time correction coefficient, which adjusts the decay of the feature over time and represents the magnitude of the time correction; e -λtis the time decay factor, which indicates the decay rate of the feature over time; λ is the time constant, which reflects the decay speed of the feature and indicates the decay rate.

[0152] The time-weighted average dynamically adjusts the time correction coefficient so that the eigenvector can reflect the changes of the signal in different time periods.

[0153] S2-3-2, using classification algorithms (such as support vector machines or random forests), conduct multi-dimensional early warning evaluation on the constructed feature vectors to generate early warning signals. Use threshold functions to transform and normalize the features, generate early warning signals through comprehensive evaluation, indicate potential health risks, and obtain dynamic abnormality predictive factors of ECG signals:

[0154] Among them, CAPF is the predictive factor of dynamic abnormality of ECG signal; θ is the threshold function, which is used to generate warning signal, indicating whether the warning is triggered; a represents the characteristic normalization coefficient, which represents the normalization of HRV comprehensive index; b represents the characteristic normalization coefficient, which represents the normalization of ST segment characteristics.

[0155] The threshold function can be adjusted according to specific circumstances to improve the accuracy and timeliness of early warning.

[0156] The dynamic abnormality predictive factor of ECG signals is used to indicate and reflect the comprehensive health status and potential risks of the patient's ECG signals. The larger the value of the dynamic abnormality predictive factor of ECG signals, the more abnormal features are detected in the ECG signals, indicating a higher risk of potential cardiovascular problems; the smaller the value of the dynamic abnormality predictive factor of ECG signals, the more normal the ECG signals are, and the better the patient's cardiovascular health status. The dynamic abnormality predictive factor of ECG signals integrates multi-dimensional features such as heart rate variability and dynamic changes in ST segments, and provides accurate early warning signals through complex algorithms and data processing to help clinicians detect and intervene in cardiovascular diseases early.

[0157] S2-4, the ECG signal dynamic abnormality predictive factor needs to be compared with the low risk threshold and the high risk threshold, and different levels of warning signals are generated according to the comparison results. The specific execution process is as follows: if the ECG signal dynamic abnormality predictive factor value is less than the low risk threshold, a green warning signal is generated, indicating low risk; if the ECG signal dynamic abnormality predictive factor value is between the low risk threshold and the high risk threshold, a yellow warning signal is generated, indicating medium risk; if the ECG signal dynamic abnormality predictive factor value is greater than the high risk threshold, a red warning signal is generated, indicating high risk.

[0158] When no early warning reminder is obtained during data collection, the present invention uses multi-level preprocessing and noise filtering, uses adaptive baseline correction technology to remove baseline drift, and applies adaptive bandpass filters to remove high-frequency and low-frequency noise to ensure the clarity and stability of the signal. Subsequently, heart rate variability analysis and ST segment dynamic change detection are performed to extract deep features, and a classification algorithm is used to perform multi-dimensional early warning evaluation on the extracted features to generate early warning signals with clinical indication significance. By comparing the dynamic abnormal predictive factors of the ECG signal with the risk threshold, early warning signals of different levels are generated to indicate potential health risks and help clarify the priority of task processing.

[0159] S3 includes the following:

[0160] S3-1, using the wristwatch to monitor nurses’ heart rate, blood pressure, skin galvanic response and other physiological indicators in real time, and pre-processing the collected data to ensure the accuracy and stability of the data:

[0161] S3-1-1, perform sliding average on the real-time collected heart rate data to reduce random fluctuations and noise interference and ensure the stability and reliability of the heart rate data.

[0162] S3-1-2, filter the real-time collected blood pressure data. Use bandpass filtering technology to remove high-frequency and low-frequency noise and retain the main frequency components of the blood pressure signal to improve the credibility of the data.

[0163] S3-1-3, perform baseline correction on the real-time collected galvanic skin response data. By calculating the local average and subtracting the low-frequency components in the original signal, the baseline drift is removed to ensure the accuracy of the galvanic skin response data.

[0164] S3-2, extract relevant features from the preprocessed physiological data for subsequent stress assessment model:

[0165] S3-2-1, calculate the heart rate variability index from the smoothed heart rate data. The specific method is to calculate the standard deviation of the heart rate data to reflect the fluctuation of the heart rate and evaluate the heart rate variability characteristics of the nurses.

[0166] S3-2-2, extract the mean value and variability characteristics from the filtered blood pressure data. The specific method is to calculate the mean value of the blood pressure data to reflect the overall level of blood pressure, and calculate the standard deviation of the blood pressure data to reflect the fluctuation of blood pressure.

[0167] S3-2-3, extract the average value and change rate features from the skin electrical response data after baseline correction. The specific method is to calculate the average value of the skin electrical response data to reflect the overall level of skin electrical response, and calculate the change rate between adjacent sampling points to reflect the fluctuation of skin electrical response.

[0168] S3-3, based on the multi-parameter integrated stress assessment model, combined with heart rate, blood pressure, skin galvanic response and other characteristics, conducts dynamic physiological data analysis to assess the nurses’ immediate work stress status:

[0169] S3-3-1, integrate the extracted heart rate variability, blood pressure and skin galvanic response features into a comprehensive stress assessment model. The specific method is to assign a certain weight to each feature to indicate its importance in stress assessment. These features are weighted and combined according to the weights to generate a comprehensive stress index.

[0170] S3-3-2, calculate the stress index according to the integrated model. The specific method is to sum the characteristic values ​​of heart rate variability, blood pressure and skin electrical response according to the preset weights to generate an instant stress index. The stress index reflects the current work stress state of the nurse.

[0171] Real-time monitoring of nurses' heart rate, blood pressure, galvanic skin response and other physiological indicators, data preprocessing and feature extraction, and dynamic physiological data analysis based on a multi-parameter integrated stress assessment model to assess nurses' immediate work stress status and help timely detect and intervene in nurses' work stress problems.

[0172] The present invention uses a smart watch to monitor the nurses' heart rate, blood pressure, skin electrical response and other physiological indicators in real time, and pre-processes the data to ensure its accuracy and stability, extracts relevant features from it, combines it with a multi-parameter integrated stress assessment model, performs dynamic physiological data analysis, and generates a real-time stress index to assess the nurses' current work stress status, thereby timely discovering and intervening in the nurses' work stress problems, facilitating dynamic decision-making for later task allocation, and improving work efficiency and nursing quality.

[0173] S4 includes the following:

[0174] S4-1, first, collect the time sensitivity of warning signals, nurses’ skill matrix, and current work stress data, and model these data.

[0175] S4-1-1, for each warning signal, assign a time sensitivity score based on its urgency:

[0176]

[0177] Among them, S l represents the time sensitivity score corresponding to the warning signal l, that is, the time sensitivity score of task l, indicating the urgency of the warning signal; W p Indicates the weight of the warning signal, reflecting the importance of the warning signal; T e Indicates the expiration time of the warning signal, indicating the latest time when the task needs to be completed; T nIndicates the current time and the current time node.

[0178] The weight of the warning signal W p It is calculated based on the ECG signal dynamic abnormality predictive factor (CAPF). The specific process is: first, calculate the ECG signal dynamic abnormality predictive factor value of each warning signal, and then normalize the ECG signal dynamic abnormality predictive factor values ​​of all warning signals so that their sum is 1. The weight of each warning signal is equal to its ECG signal dynamic abnormality predictive factor value divided by the sum of the ECG signal dynamic abnormality predictive factor values ​​of all warning signals, thereby reflecting the relative importance of the warning signal among all signals. In this way, the higher the ECG signal dynamic abnormality predictive factor value of the warning signal, the greater its weight, indicating that its importance is higher.

[0179] The time sensitivity score is determined by considering the importance of the warning signal and the latest time the task needs to be completed. The more important the warning signal is and the more urgent the time it needs to be completed, the higher the time sensitivity score.

[0180] S4-2, calculate the task suitability index assigned to each nurse for each task corresponding to each warning signal, taking into account time sensitivity, skill matching and immediate stress status. The calculation formula is as follows:

[0181]

[0182] Among them, CS ol It represents the task fitness index of task l assigned to nurse o;

[0183] S l represents the time sensitivity score of task l;

[0184] M ov represents the skill match between nurse o and task l;

[0185] YL o It represents the immediate stress index of nurse o;

[0186] τ represents the adjustment parameter used to control the impact of stress state on the score.

[0187] The calculation of the task fitness index is to evaluate the fitness of the task assigned to a specific nurse under the current conditions by comprehensively considering the time sensitivity of the task, the skill matching between the nurse and the task, and the nurse's immediate stress state. Logarithmic transformation and nonlinear adjustment are used in the calculation process to make time sensitivity and skill matching have a more significant impact on fitness. At the same time, the introduction of exponential function and sigmoid function refines the adjustment of stress state to ensure the balance and rationality of the comprehensive score, so as to give priority to assigning tasks with higher fitness index to corresponding nurses.

[0188] The task suitability index is used to indicate the comprehensive suitability of the task corresponding to each warning signal when assigned to each nurse, taking into account the time sensitivity of the task, the skill matching of the nurse, and the immediate work pressure status. The larger the suitability index, the more suitable the task is for the nurse under the current conditions, the more urgent the task is, the nurse's skills are highly matched with the task requirements, and the work pressure is low; the smaller the suitability index, the less suitable the task is for the nurse, which may be due to the fact that the task is not urgent, the skills are not matched, or the nurse's work pressure is high.

[0189] S5 includes the following:

[0190] S5-1, sort all combinations of tasks and nurses, and sort the tasks from high to low according to the calculated task fitness index.

[0191] S5-2, the task with the highest task suitability index is assigned to the corresponding nurse first, and a prompt processing signal is sent to the wristwatch of the corresponding nurse, and the assignment is performed according to the following steps:

[0192] S5-2-1, select the task and nurse combination with the highest fitness index.

[0193] S5-2-2, assign the task to the corresponding nurse, and the central monitor sends a signal to the watch, prompting the nurse bound to the watch to handle it and record the task assignment.

[0194] S5-2-3, update the immediate work pressure status of the assigned nurse, and re-execute step S3 for calculation.

[0195] Assume that executing the task can improve the proficiency of certain skills of the nurse, update the nurse's skill matrix, and re-execute step S4-1-2 to calculate the skill matching between the nurse and the task.

[0196] S5-3, repeat step S5-2 to continue to assign the remaining tasks. Each time, the task and nurse combination with the highest fitness index is selected to perform task assignment and status update until all tasks are assigned.

[0197] The present invention establishes a time sensitivity scoring model and a nurse skill matching model by collecting data such as the time sensitivity of early warning signals, the nurse skill matrix and the current workload, and calculates the task fitness index based on these models by comprehensively considering the time sensitivity, skill matching and immediate stress status. Tasks with the highest fitness index are preferentially assigned to corresponding nurses, and the nurses' work stress status and skill matrix are updated in real time to dynamically optimize the task allocation process, thereby achieving efficient and reasonable task allocation and improving overall nursing efficiency and quality.

[0198] Figure 2 The present invention is a schematic diagram of data transmission between an electrocardiogram monitor, a central monitor and a wristwatch in a patient physiological information monitoring system.

[0199] The above formulas are all dimensionless and numerical calculations. The formula is a formula for the most recent real situation obtained by collecting a large amount of data and performing software simulation. The preset parameters and thresholds in the formula are set by technicians in this field according to actual conditions.

[0200] The above embodiments can be implemented in whole or in part by software, hardware, firmware or any other combination. When implemented using software, the above embodiments can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, the process or function described in the embodiment of the present application is generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable devices. The computer instructions can be stored in a computer-readable storage medium, or transmitted from one computer-readable storage medium to another computer-readable storage medium. For example, the computer instructions can be transmitted from one website site, computer, server or data center to another website site, computer, server or data center by wired (e.g., infrared, wireless, microwave, etc.). The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that contains one or more available media sets. The available medium can be a magnetic medium (e.g., a floppy disk, a hard disk, a tape), an optical medium (e.g., a DVD), or a semiconductor medium. The semiconductor medium can be a solid-state hard disk.

[0201] Those of ordinary skill in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.

[0202] In the several embodiments provided in the present application, it should be understood that the disclosed systems and devices can be implemented in other ways. For example, the device embodiments described above are only schematic. For example, the division of the units is only a logical function division. There may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be an indirect coupling or communication connection through some interfaces, devices or units, which can be electrical, mechanical or other forms.

[0203] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0204] In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit.

[0205] If the functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application can be essentially or partly embodied in the form of a software product that contributes to the prior art. The computer software product is stored in a storage medium and includes several instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps described in each embodiment of the present application. The aforementioned storage medium includes: various media that can store program codes, such as a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk.

[0206] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art who is familiar with the present technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

[0207] Finally: The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. 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 patient physiological information monitoring system, characterized in that: include: Data aggregation module, signal purification module, feature analysis module, pressure analysis module, task adaptation module and task allocation module; Data aggregation module: obtains the patient's ECG data in real time through the ECG monitor, transmits the data to the central monitor, and then passes it to the signal purification module; Signal purification module: Apply comprehensive anomaly detection technology in the central monitor to suppress noise and purify ECG signals, automatically identify and correct electrode wearing status, decide whether to issue a warning signal, and pass the purified ECG signals and decision results to the feature analysis module; Feature analysis module: when no warning reminder is received during data collection, adaptive baseline correction technology and adaptive bandpass filter are used to ensure the clarity and stability of the signal, and deep features are extracted through heart rate variability analysis and ST segment dynamic change detection. Warning signals are issued for the extracted features, and the extracted ECG feature data are transmitted to the pressure analysis module; Stress analysis module: monitors the physiological parameters of the caregiver in real time through the watch, and performs dynamic physiological data analysis based on a multi-parameter integrated stress assessment model, generates an instant stress index to assess the current work stress status of the caregiver, and transmits the instant stress index of the caregiver to the task adaptation module; Task adaptation module: Using the rule engine and multi-factor scheduling algorithm, based on the time sensitivity of the warning signal, the nursing staff skill matrix and the immediate work pressure status, the task adaptation index model is constructed to analyze the adaptability of the task and the nursing staff, and the task adaptation index is transmitted to the task allocation module; S4 includes the following: S4-1, first, collect the time sensitivity of warning signals, nursing skills matrix and current work pressure data; S4-1-1, for each warning signal, assign a time sensitivity score based on its urgency: Among them, S l represents the time sensitivity score corresponding to the warning signal l, that is, the time sensitivity score of task l, indicating the urgency of the warning signal; W p Indicates the weight of the warning signal, reflecting the importance of the warning signal; T e Indicates the expiration time of the warning signal, indicating the latest time when the task needs to be completed; T n Indicates the current time, indicating the current time node; S4-1-2, build a nursing skill matrix and calculate the skill matching between nursing staff and tasks: Among them, M ov represents the skill matching degree between care worker o and task l; K represents the total number of skills; SK ok″ represents the proficiency of caregiver o in skill k″; TM lk″ represents the requirement of skill k″ required for task l; S4-2, calculate the task suitability index assigned to each caregiver for each task corresponding to each warning signal, taking into account time sensitivity, skill matching and immediate stress status. The calculation formula is as follows: Among them, CS ol represents the task suitability index of task l assigned to caregiver o; S l represents the time sensitivity score of task l; M ov represents the skill matching degree between care worker o and task l; YL o represents the immediate stress index of caregiver o; τ represents the adjustment parameter used to control the impact of stress status on the score; Task allocation module: sorts all combinations of tasks and caregivers, and prioritizes tasks with the highest suitability to corresponding caregivers based on the task suitability index, updates the work stress status and skill matrix of caregivers in real time, dynamically optimizes the task allocation process, and feeds back the updated work stress status and skill matrix data of caregivers to the pressure analysis module and task adaptation module; S5 includes the following: S5-1, sort all combinations of tasks and caregivers, and sort the tasks from high to low according to the calculated task fitness index; S5-2, the task with the highest task suitability index is assigned to the corresponding caregiver first, and a prompt processing signal is sent to the watch of the corresponding caregiver, and the assignment is performed according to the following steps: S5-2-1, select the task and caregiver combination with the highest fitness index; S5-2-2, assign the task to the corresponding caregiver, and the central monitor sends a signal to the watch, prompting the caregiver bound to the watch to handle it and record the task assignment; S5-2-3, updating the immediate work pressure status of the assigned caregiver, and re-execute step S3 for calculation; Update the caregiver's skill matrix and re-execute step S4-1-2 to calculate the skill matching between the caregiver and the task; S5-3, repeat step S5-2 to continue to assign the remaining tasks, each time selecting the task and caregiver combination with the highest fitness index, and perform task assignment and status update until all tasks are assigned.

2. A patient physiological information monitoring system according to claim 1, characterized in that: Step S1 includes the following contents: S1-1, obtaining the patient's ECG data in real time through an ECG monitor; S1-2, the central monitor performs noise frequency domain characteristic analysis on the acquired ECG data; S1-3, evaluate the amplitude fluctuation of the purified signal, calculate the RMS value and peak factor of the signal, and apply the dynamic segmentation method to identify abnormal amplitude changes; S1-4, further performing wavelet transform on the signal to obtain the time-frequency distribution information of the signal, and identifying transient interference by analyzing the wavelet coefficients; S1-5, apply consistency detection algorithm to evaluate the periodicity and stability of the signal, automatically identify and correct the electrode wearing status, and ensure data integrity. The specific steps include: The consistency detection algorithm is used to evaluate the periodicity and stability of the signal and determine the wearing status of the electrode; The signal consistency index is calculated using the following formula: Where C is the consistency index; x(ti) represents the signal value of the i-th sampling point; |x(ti)-μ represents the absolute difference between the signal value and the mean; N is the total number of sampling points; σ is the standard deviation of the signal; Convert the consistency index to a stability index ranging from 0 to 1: Among them, S is the stability index; max(C) is the maximum value of the consistency index; When the stability index is close to 1, it means that the signal is stable and the electrode is in good wearing condition; when the stability index is significantly lower than 1, it means that the signal fluctuates greatly and an early warning reminder is issued.

3. A patient physiological information monitoring system according to claim 2, characterized in that: Step S2 includes the following contents: S2-1, when no warning reminder is obtained during data collection, the collected ECG data is used for multi-level preprocessing and noise filtering: S2-1-1, using adaptive baseline correction technology, the original ECG signal is processed to remove the baseline drift by calculating the local average and subtracting the low-frequency components in the signal; S2-1-2, then, applying an adaptive bandpass filter to the signal after removing the baseline drift to remove high-frequency and low-frequency noise; S2-2, detect heart rate variability and ST segment dynamic changes on the preprocessed signal: S2-2-1, in the heart rate variability analysis, by calculating the RR interval, and performing adaptive correction and feature extraction; Next, the standard deviation and range of the RR interval were calculated; S2-2-2, detecting the ST segment in the ECG signal by calculating the voltage elevation value of the ST segment relative to the PR segment; S2-3, using classification algorithms, conduct in-depth analysis and multi-dimensional early warning evaluation of the extracted features: S2-3-1, combining the comprehensive index of heart rate variability and the time-weighted average of ST segment characteristics into a vector; S2-3-2, use the threshold function to transform and normalize the features, generate early warning signals through comprehensive evaluation, and obtain the predictive factors of dynamic abnormality of ECG signals: Among them, CAPF is the predictive factor of dynamic abnormality of ECG signal; θ is the threshold function, which is used to generate warning signal, indicating whether the warning is triggered; a represents the characteristic normalization coefficient, which represents the normalization of HRV comprehensive index; b represents the characteristic normalization coefficient, which represents the normalization of ST segment characteristics; S2-4, the ECG signal dynamic abnormality predictive factor needs to be compared with the low-risk threshold and the high-risk threshold, and different levels of warning signals are generated according to the comparison results; the specific implementation process is as follows: if the ECG signal dynamic abnormality predictive factor value is less than the low-risk threshold, a green warning signal is generated, indicating a low risk; if the ECG signal dynamic abnormality predictive factor value is between the low-risk threshold and the high-risk threshold, a yellow warning signal is generated, indicating a medium risk; if the ECG signal dynamic abnormality predictive factor value is greater than the high-risk threshold, a red warning signal is generated, indicating a high risk.

4. A patient physiological information monitoring system according to claim 3, characterized in that: S3 includes the following: S3-1, use the wristwatch to monitor the caregiver's physiological indicators in real time, including heart rate, blood pressure and skin galvanic response, and pre-process the collected data: S3-1-1, performing sliding average on the heart rate data collected in real time; S3-1-2, filtering the blood pressure data collected in real time; S3-1-3, baseline correction of the real-time collected galvanic skin response data; S3-2, extract relevant features from preprocessed physiological data: S3-2-1, calculating the standard deviation of the heart rate data from the smoothed heart rate data as a heart rate variability indicator; S3-2-2, extracting the mean and standard deviation from the filtered blood pressure data; S3-2-3, extract the mean value and the rate of change between adjacent sampling points from the baseline-corrected galvanic skin response data; S3-3, weighted summation of the characteristic values ​​of heart rate variability, blood pressure and skin electrical response is performed to generate an instant stress index, which is used to reflect the current work stress state of the caregiver.

5. A patient physiological information monitoring system according to claim 1, characterized in that: The weight of the warning signal is calculated based on the dynamic abnormality predictive factor of the ECG signal. The specific process is: first, calculate the dynamic abnormality predictive factor value of the ECG signal of each warning signal, and then normalize the dynamic abnormality predictive factor values ​​of the ECG signal of all warning signals so that their sum is 1. The weight of each warning signal is equal to its dynamic abnormality predictive factor value of the ECG signal divided by the sum of the dynamic abnormality predictive factor values ​​of the ECG signal of all warning signals.

Images