Nursing process monitoring system

Abstract

The invention relates to the technical field of nursing monitoring, in particular to a nursing process monitoring system which comprises a state classification unit, an intelligent display unit, a state monitoring unit and a state matching unit and further comprises an electrocardio monitoring unit used for conducting electrocardio monitoring on a target patient through an electrocardio monitoring instrument, the electrocardiosignal acquisition unit is used for acquiring an electrocardiosignal of the target patient and transmitting the electrocardiosignal of the target patient to the electrocardiosignal preprocessing unit. According to the system, multiple physiological parameters such as electrocardiogram, blood pressure, oxyhemoglobin saturation and body temperature can be monitored at the same time, comprehensive health condition evaluation on a target patient is provided, diagnosis deviation possibly caused by a single data source is avoided, the electrocardiogram signals are expanded, filtered and normalized through the electrocardiogram preprocessing unit, and the diagnosis accuracy is improved. And the characteristics of the standard electrocardiosignal are extracted, so that the accuracy and availability of the electrocardiogram data can be improved.

Description

Technical Field

[0001] The present invention relates to the technical field of nursing monitoring, and in particular to a nursing process monitoring system. Background Art

[0002] Nursing is comprehensive health care for individuals, families and communities, including prevention, diagnosis, treatment and rehabilitation. It aims to maintain or improve health, alleviate the effects of illness or injury, and support the patient's overall well-being. Nursing includes not only direct physical care, but also psychological, emotional and educational support to promote the patient's overall health and quality of life.

[0003] Traditional systems may only monitor a single physiological parameter, which may lead to an incomplete assessment of the patient's health status and increase the risk of diagnostic bias; and many traditional systems may lack advanced data processing and analysis capabilities, such as signal filtering, normalization processing, and feature extraction, which may lead to low data accuracy and availability, thereby affecting the accuracy of diagnosis; and traditional systems may not be able to provide real-time health data updates and feedback, resulting in caregivers being unable to quickly understand changes in the patient's health status and making it difficult to take timely measures to deal with emergencies; and traditional systems may lack clear graph structures or models to represent the relationships between different physiological indicators, which makes it difficult to understand the relationship between various physiological parameters and their impact on the overall health status. Summary of the invention

[0004] The purpose of the present invention is to propose a nursing process monitoring system in view of the problems existing in the background technology.

[0005] The technical solution of the present invention is a nursing process monitoring system, comprising a state classification unit, an intelligent display unit, a state monitoring unit and a state matching unit, and also comprising:

[0006] An ECG monitoring unit, the ECG monitoring unit is used to perform ECG monitoring on a target patient through an ECG monitoring instrument to obtain an ECG signal of the target patient, and transmit the ECG signal of the target patient to an ECG preprocessing unit;

[0007] An ECG preprocessing unit, wherein the ECG preprocessing unit receives the ECG signal of the target patient transmitted by the ECG monitoring unit, and sets a filtering window, performs expansion processing on the ECG signal of the target patient based on the filtering window to obtain an expanded ECG signal, performs filtering processing on the expanded ECG signal to obtain a first ECG signal, performs signal normalization processing on the first ECG signal to obtain a second ECG signal, takes the R wave of the second ECG signal as the center, performs interception processing on the second ECG signal through a preset equal-length threshold to obtain a standard ECG signal, and transmits the standard ECG signal of the target patient to a feature extraction unit;

[0008] A feature extraction unit receives the standard ECG signal of the target patient transmitted by the ECG preprocessing unit, and performs feature extraction on the standard ECG signal of the target patient to obtain a feature set corresponding to the standard ECG signal, and transmits the feature set corresponding to the standard ECG signal to the state classification unit.

[0009] Preferably, the state classification unit receives the feature set corresponding to the standard ECG signal transmitted by the feature extraction unit, and inputs the feature set corresponding to the standard ECG signal into a trained state classification model, outputs the heart rhythm state corresponding to the target patient through the trained state classification model, and transmits the heart rhythm state corresponding to the target patient to the intelligent display unit and the state matching unit.

[0010] Preferably, the state monitoring unit is used to perform state monitoring on the target patient through a blood pressure sensor, a blood oxygen saturation sensor and a body temperature sensor to obtain state data of the target patient, wherein the state data includes blood pressure, blood oxygen saturation and body temperature, and a health real-time monitoring directed graph is constructed based on a preset defined relationship between heart rate, blood pressure, blood oxygen saturation and body temperature, wherein the health real-time monitoring directed graph includes multiple nodes and multiple undirected edges, wherein the nodes correspond to electrocardiogram monitoring instruments, blood pressure sensors, blood oxygen saturation sensors and body temperature sensors, and the undirected edges correspond to a preset defined relationship between heart rate, blood pressure, blood oxygen saturation and body temperature, and weights of the nodes are assigned based on the state data of the target patient, and weights of the undirected edges are assigned based on a preset defined relationship between heart rate, blood pressure, blood oxygen saturation and body temperature to obtain a weighted health real-time monitoring directed graph, and the weighted health real-time monitoring directed graph is transmitted to the intelligent display unit and the state matching unit.

[0011] Preferably, the state matching unit receives the heart rhythm state corresponding to the target patient transmitted by the state classification unit and the weighted health real-time monitoring directed graph transmitted by the state monitoring unit, and matches the heart rhythm state corresponding to the target patient with a standard weighted health monitoring directed graph sequence pre-stored in a database to obtain a matching standard weighted health monitoring directed graph, compares the matching standard weighted health monitoring directed graph with the weighted health real-time monitoring directed graph to obtain an error between the two, matches the error between the two with a preset error threshold table to obtain the health status level of the target patient, and transmits the health status level of the target patient to the intelligent display unit.

[0012] Preferably, the intelligent display unit receives the heart rhythm state corresponding to the target patient transmitted by the state classification unit, the weighted health real-time monitoring directed graph transmitted by the state monitoring unit, and the health status level of the target patient transmitted by the state matching unit, and displays the heart rhythm state corresponding to the target patient, the weighted health real-time monitoring directed graph, and the health status level of the target patient on the intelligent display terminal, and uploads the heart rhythm state corresponding to the target patient, the weighted health real-time monitoring directed graph, and the health status level of the target patient to the caregiver's intelligent terminal in real time.

[0013] Preferably, the formula for the expansion process is as follows:

[0014]

[0015] Where, d(n) represents the nth sample point after expansion, L represents the record length of each ECG signal, and k represents the filter window;

[0016] The formula for the filtering process is as follows:

[0017] s(o)=f med {f sort [d(0):d(o+k-1)]};

[0018] Among them, s(o) represents the filtered data of the 0th point in the expanded ECG signal, and f sort represents the sorting function, f med Represents the median function;

[0019] The formula for the normalization process is as follows:

[0020]

[0021] Among them, s * (o) represents the filtered data of the oth point in the second ECG signal, s min Indicates the minimum sample amplitude value in the current first ECG signal, s max Indicates the maximum value of the sample amplitude in the current first ECG signal;

[0022] The formula for the interception process is as follows:

[0023] U u (δ) = s * (δ)*(R u -90+δ);

[0024] Among them, U u (δ) represents the δth sampling point data in the uth heartbeat, and u=1,2,...,N beat , N beatIndicates the number of R peak marks of the current ECG signal, R u represents the R wave in the u-th heart beat.

[0025] Preferably, feature extraction is performed on the standard ECG signal of the target patient to obtain a feature set corresponding to the standard ECG signal, including:

[0026] The RR interval of the standard electrocardiogram signal is feature extracted by a first formula to obtain a first feature corresponding to the standard electrocardiogram signal. The first formula is as follows:

[0027] RR(n)=R(n+1)-R(n);

[0028] Where RR(n) represents the nth RR interval, and R(n) represents the nth R wave;

[0029] Calculating the average value, standard deviation, standard deviation of the difference value and root mean square of the difference value of the first feature corresponding to the standard electrocardiogram signal to obtain the second feature corresponding to the standard electrocardiogram signal;

[0030] The DRR interval of the standard ECG signal is feature extracted based on the first feature corresponding to the standard ECG signal by a second formula to obtain a third feature corresponding to the standard ECG signal. The second formula is as follows:

[0031] DRR(n)=RR(n+1)-RR(n);

[0032] Where DRR(n) represents the nth DRR interval;

[0033] The skewness and kurtosis of the standard electrocardiogram signal are extracted by a third formula to obtain a fourth feature corresponding to the standard electrocardiogram signal. The third formula is as follows:

[0034]

[0035] Among them, K represents the skewness of the standard ECG signal, S represents the kurtosis of the standard ECG signal, and N hos Indicates the number of samples in this segment, U u_mean Represents the average value of the sampling point data in the u-th heart beat.

[0036] Preferably, performing feature extraction on the standard ECG signal of the target patient to obtain a feature set corresponding to the standard ECG signal also includes:

[0037] Performing an inverse cosine transform on the heart rate value corresponding to the standard electrocardiogram signal to convert the heart rate value into an angle value in a polar coordinate system, and converting the timestamp corresponding to the standard electrocardiogram signal into a radius in a polar coordinate system;

[0038] constructing a polar coordinate matrix based on the angle value and radius in the polar coordinate system, wherein the polar coordinates of the elements in the polar coordinate matrix are represented by the angle between the timestamp and the corresponding heart rate value, and converting the polar coordinate data into Cartesian coordinates based on a conversion formula from polar coordinates to Cartesian coordinates;

[0039] Performing an inner product on the data in the Cartesian coordinates to obtain a Gramian matrix, and mapping corresponding values ​​of the Gramian matrix onto an image to obtain a GAF image;

[0040] Perform feature extraction on the GAF image to obtain image features corresponding to the GAF image, and combine the first feature, the second feature, the third feature, the fourth feature and the image features corresponding to the GAF image to obtain a feature set corresponding to the standard electrocardiogram signal.

[0041] Preferably, the state classification model adopts an improved support vector machine model, and the state classification model improves the support vector machine model through an optimization algorithm.

[0042] Preferably, the support vector machine model is improved by an optimization algorithm, including:

[0043] Set the initialization parameters of the optimization algorithm to initialize the population position. The population position initialization formula is as follows:

[0044]

[0045] Among them, x t and x t+1 represents the position of the individual at the t-th iteration and the t+1-th iteration, and p represents the preset parameter for controlling the nonlinear mapping;

[0046] Calculate the fitness value of the individual and calculate the elite individual position. The elite individual position calculation formula is as follows:

[0047]

[0048] in, represents the reverse solution of the j-th dimension individual at the t-th iteration, x i,，j (t) represents the elite solution of the j-th dimension individual at the t-th iteration, k represents a random number in the interval [0,1], u j (t) represents the maximum position of the individual at the tth iteration, l j (t) represents the minimum position of the individual at the tth iteration.

[0049] Preferably, the support vector machine model is improved by an optimization algorithm, further comprising:

[0050] Marking a preset number of the individuals with the best fitness values ​​as leader individuals, and updating the positions of the remaining individuals in the population;

[0051] Calculate and sort the updated individual fitness values, obtain elite individuals, and update the elite individual positions;

[0052] Repeat the above steps until the optimal solution is found or the maximum number of iterations of the individual optimization algorithm is reached, thereby obtaining the optimal penalty factor and kernel function parameters of the support vector machine model, and improving the support vector machine model based on the optimal penalty factor and kernel function parameters.

[0053] Compared with the prior art, the above technical solution of the present invention has the following beneficial technical effects:

[0054] 1. The present invention can simultaneously monitor multiple physiological parameters such as electrocardiogram, blood pressure, blood oxygen saturation and body temperature, provide a comprehensive health status assessment of the target patient, avoid diagnostic bias that may be caused by a single data source, and expand, filter and normalize the electrocardiogram signal through the electrocardiogram preprocessing unit, and extract the features of the standard electrocardiogram signal, thereby improving the accuracy and availability of the electrocardiogram data, which is helpful for more accurate diagnosis of the heart rhythm state, and classifying the extracted features using a trained model through the state classification unit, providing automated heart rhythm state recognition, and defining a weighted health real-time monitoring directed graph through the state monitoring unit to clarify the preset relationship between heart rate, blood pressure, blood oxygen saturation and body temperature, so that the system can clearly represent the relationship between various physiological indicators through a graph structure, thereby helping to better understand the relationship between different physiological parameters and their impact on the overall health status.

[0055] 2. The present invention can provide each patient with a personalized health assessment by comparing the real-time data of each patient with the standard chart through the state matching unit, so that a treatment plan that is more in line with the patient's actual situation can be formulated based on the specific health status level, and the health data of the target patient, including heart rhythm status, health monitoring chart and health status level, can be updated and displayed in real time through the intelligent display unit. This real-time feedback function helps nursing staff to quickly understand the patient's health changes and take corresponding measures in time. BRIEF DESCRIPTION OF THE DRAWINGS

[0056] Figure 1 A schematic diagram of the system architecture of an overall system in an embodiment of the present invention.

[0057] Figure numerals: 1. ECG monitoring unit; 2. ECG preprocessing unit; 3. Feature extraction unit; 4. State classification unit; 5. Intelligent display unit; 6. State monitoring unit; 7. State matching unit. DETAILED DESCRIPTION

[0058] Embodiment 1, as Figure 1 As shown, a nursing process monitoring system proposed by the present invention includes a state classification unit 4, an intelligent display unit 5, a state monitoring unit 6 and a state matching unit 7, and also includes:

[0059] The ECG monitoring unit 1 is used to perform ECG monitoring on a target patient through an ECG monitoring instrument to obtain an ECG signal of the target patient, and transmit the ECG signal of the target patient to the ECG preprocessing unit 2;

[0060] ECG preprocessing unit 2, ECG preprocessing unit 2 receives the ECG signal of the target patient transmitted by the ECG monitoring unit 1, and sets a filtering window, performs expansion processing on the ECG signal of the target patient based on the filtering window to obtain an expanded ECG signal, performs filtering processing on the expanded ECG signal to obtain a first ECG signal, performs signal normalization processing on the first ECG signal to obtain a second ECG signal, takes the R wave of the second ECG signal as the center, intercepts the second ECG signal through a preset equal-length threshold to obtain a standard ECG signal, and transmits the standard ECG signal of the target patient to the feature extraction unit 3;

[0061] The feature extraction unit 3 receives the standard ECG signal of the target patient transmitted by the ECG preprocessing unit 2, and performs feature extraction on the standard ECG signal of the target patient to obtain a feature set corresponding to the standard ECG signal, and transmits the feature set corresponding to the standard ECG signal to the state classification unit 4.

[0062] In the present invention, the ECG monitoring instrument is a medical device used to record and analyze the electrical activity of the heart. It is placed on the patient's skin through electrode patches to detect the electrical signals generated by the heart at each heartbeat; the filter window refers to a device used to remove noise or unnecessary frequency components in the signal; and the R-wave center interception is based on the position of the R wave in the ECG signal to intercept a section of the ECG signal.

[0063] In an optional embodiment, the state classification unit 4 receives the feature set corresponding to the standard ECG signal transmitted by the feature extraction unit 3, and inputs the feature set corresponding to the standard ECG signal into a trained state classification model, outputs the heart rhythm state corresponding to the target patient through the trained state classification model, and transmits the heart rhythm state corresponding to the target patient to the intelligent display unit 5 and the state matching unit 7.

[0064] It should be noted that the heart rhythm state includes but is not limited to normal sinus rhythm, tachycardia, bradycardia, atrial fibrillation, atrial flutter, ventricular premature beats, ventricular fibrillation and atrioventricular conduction block.

[0065] In an optional embodiment, the state monitoring unit 6 is used to perform state monitoring on the target patient through a blood pressure sensor, a blood oxygen saturation sensor and a body temperature sensor to obtain the state data of the target patient, the state data including blood pressure, blood oxygen saturation and body temperature, and construct a health real-time monitoring directed graph based on a preset defined relationship between heart rate, blood pressure, blood oxygen saturation and body temperature. The health real-time monitoring directed graph includes multiple nodes and multiple undirected edges, the nodes correspond to the electrocardiogram monitoring instrument, the blood pressure sensor, the blood oxygen saturation sensor and the body temperature sensor, and the undirected edges correspond to the preset defined relationship between heart rate, blood pressure, blood oxygen saturation and body temperature. The weights of the nodes are assigned based on the state data of the target patient, and the weights of the undirected edges are assigned based on the preset defined relationship between heart rate, blood pressure, blood oxygen saturation and body temperature to obtain a weighted health real-time monitoring directed graph, and the weighted health real-time monitoring directed graph is transmitted to the intelligent display unit 5 and the state matching unit 7.

[0066] It should be noted that a blood pressure sensor refers to a device that measures the blood pressure value (such as systolic and diastolic pressure) of a target patient; a blood oxygen saturation sensor refers to an instrument that monitors the oxygen saturation in the blood; a body temperature sensor refers to a device that measures the body temperature of a target patient; the preset defined relationship between heart rate, blood pressure, blood oxygen saturation and body temperature usually refers to the normal range of these physiological parameters in a healthy state and their mutual influence. For example, under normal circumstances, blood pressure and heart rate have a certain correlation. An increase in heart rate may lead to an increase in blood pressure. A faster heart rate may lead to insufficient blood oxygen supply, affecting blood oxygen saturation. An increase in body temperature may cause a slight increase in blood pressure. Excessive body temperature may affect the delivery and utilization of oxygen.

[0067] In an optional embodiment, the state matching unit 7 receives the heart rhythm state corresponding to the target patient transmitted by the state classification unit 4 and the weighted health real-time monitoring directed graph transmitted by the state monitoring unit 6, and matches the heart rhythm state corresponding to the target patient with a standard weighted health monitoring directed graph sequence pre-stored in the database to obtain a matching standard weighted health monitoring directed graph, compares the matching standard weighted health monitoring directed graph with the weighted health real-time monitoring directed graph to obtain an error between the two, matches the error between the two with a preset error threshold table to obtain the health status level of the target patient, and transmits the health status level of the target patient to the intelligent display unit 5.

[0068] It should be noted that the standard weighted health monitoring directed graph sequence is pre-defined and represents the relationship between physiological parameters of different heart rate states; the heart rate state corresponding to the target patient is matched with the standard weighted health monitoring directed graph sequence pre-stored in the database, and the similarity of the two graphs is determined by graph matching and similarity calculation; according to the matching results of the error and threshold table, the health status of the target patient is classified into different levels (such as normal, mild abnormality, severe abnormality, etc.), and this level indicates the degree of deviation of the patient's health status from the standard health graph.

[0069] In an optional embodiment, the intelligent display unit 5 receives the heart rhythm state corresponding to the target patient transmitted by the state classification unit 4, the weighted health real-time monitoring directed graph transmitted by the state monitoring unit 6, and the health status level of the target patient transmitted by the state matching unit 7, and displays the heart rhythm state corresponding to the target patient, the weighted health real-time monitoring directed graph, and the health status level of the target patient on the intelligent display terminal, and uploads the heart rhythm state corresponding to the target patient, the weighted health real-time monitoring directed graph, and the health status level of the target patient to the caregiver's intelligent terminal in real time.

[0070] It should be noted that a smart terminal generally refers to a device that integrates computing power, communication functions and a user interaction interface. It can be used to display information, process data and communicate with other devices or systems.

[0071] Embodiment 2, a nursing process monitoring system proposed by the present invention, compared with embodiment 1, this embodiment further includes: the formula for extended processing is as follows:

[0072]

[0073] Where, d(n) represents the nth sample point after expansion, L represents the record length of each ECG signal, and k represents the filter window;

[0074] The filtering formula is as follows:

[0075] s(o)=f med {f sort [d(o):d(o+k-1)]};

[0076] Where s(o) represents the filtered data of the oth point in the expanded ECG signal, and f sort represents the sorting function, f med Represents the median function;

[0077] The normalization formula is as follows:

[0078]

[0079] Among them, s* (o) represents the filtered data of the oth point in the second ECG signal, s min Indicates the minimum sample amplitude value in the current first ECG signal, s max Indicates the maximum value of the sample amplitude in the current first ECG signal;

[0080] The formula for interception processing is as follows:

[0081] U u (δ) = s * (δ)*(R u -90+δ);

[0082] Among them, U u (δ) represents the δth sampling point data in the uth heartbeat, and u=1, 2, ..., N beat , N beat Indicates the number of R peak marks of the current ECG signal, R u represents the R wave in the u-th heart beat.

[0083] In an optional embodiment, feature extraction is performed on the standard ECG signal of the target patient to obtain a feature set corresponding to the standard ECG signal, including:

[0084] The RR interval of the standard ECG signal is feature extracted by the first formula to obtain the first feature corresponding to the standard ECG signal. The first formula is as follows:

[0085] RR(n)=R(n+1)-R(n);

[0086] Where RR(n) represents the nth RR interval, and R(n) represents the nth R wave;

[0087] Calculating the average value, standard deviation, standard deviation of the difference value and root mean square of the difference value of the first feature corresponding to the standard electrocardiogram signal to obtain the second feature corresponding to the standard electrocardiogram signal;

[0088] The DRR interval of the standard ECG signal is feature extracted based on the first feature corresponding to the standard ECG signal by the second formula to obtain the third feature corresponding to the standard ECG signal. The second formula is as follows:

[0089] DRR(n)=RR(n+1)-RR(n);

[0090] Where DRR(n) represents the nth DRR interval;

[0091] The skewness and kurtosis of the standard ECG signal are extracted by the third formula to obtain the fourth feature corresponding to the standard ECG signal. The third formula is as follows:

[0092]

[0093] Among them, K represents the skewness of the standard ECG signal, S represents the kurtosis of the standard ECG signal, and N hos Indicates the number of samples in this segment, U u_mean Represents the average value of the sampling point data in the u-th heart beat.

[0094] In an optional embodiment, feature extraction is performed on the standard ECG signal of the target patient to obtain a feature set corresponding to the standard ECG signal, further comprising:

[0095] Performing an inverse cosine transform on the heart rate value corresponding to the standard ECG signal to convert the heart rate value into an angle value in a polar coordinate system, and converting the timestamp corresponding to the standard ECG signal into a radius in a polar coordinate system;

[0096] A polar coordinate matrix is ​​constructed based on the angle value and radius in the polar coordinate system, wherein the polar coordinates of the elements in the polar coordinate matrix are represented by the angle between the timestamp and the corresponding heart rate value, and the polar coordinate data is converted into Cartesian coordinates based on a conversion formula from polar coordinates to Cartesian coordinates;

[0097] Perform inner product on the data in Cartesian coordinates to obtain a Gramian matrix, and map the corresponding values ​​of the Gramian matrix onto the image to obtain a GAF image;

[0098] Feature extraction is performed on the GAF image to obtain image features corresponding to the GAF image, and the first feature, the second feature, the third feature, the fourth feature and the image features corresponding to the GAF image are combined to obtain a feature set corresponding to the standard electrocardiogram signal.

[0099] It should be noted that the Gramian image is an image representation method that converts time series data from polar coordinates to Cartesian coordinates and then calculates it through inner product calculation. The GAF image can be used to represent the characteristics of time series data and is often used in pattern recognition and data analysis; the image features corresponding to the GAF image include but are not limited to texture features, shape features, spatial features and local features.

[0100] In an optional embodiment, the state classification model adopts an improved support vector machine model, and the state classification model improves the support vector machine model through an optimization algorithm.

[0101] It should be noted that support vector machine (SVM) is a supervised learning algorithm used for classification and regression tasks. SVM separates data points of different categories by finding the best separating hyperplane (i.e., decision boundary); optimization algorithm is an algorithm used to adjust model parameters to improve model performance. In SVM, the optimization algorithm is used to find the best separating hyperplane to maximize the interval between categories (i.e., support vector spacing).

[0102] In an optional embodiment, the support vector machine model is improved by an optimization algorithm, including:

[0103] Set the initialization parameters of the optimization algorithm and initialize the population position. The population position initialization formula is as follows:

[0104]

[0105] Among them, x t and x t+1 represents the position of the individual at the t-th iteration and the t+1-th iteration, and p represents the preset parameter for controlling the nonlinear mapping;

[0106] Calculate the fitness value of the individual and the position of the elite individual. The calculation formula for the position of the elite individual is as follows:

[0107]

[0108] in, represents the reverse solution of the j-th dimension individual at the t-th iteration, x i,j (t) represents the elite solution of the j-th dimension individual at the t-th iteration, k represents a random number in the interval [0,1], u j (t) represents the maximum position of the individual at the tth iteration, l j (t) represents the minimum position of the individual at the tth iteration.

[0109] In an optional embodiment, the support vector machine model is improved by an optimization algorithm, further comprising:

[0110] Mark a preset number of individuals with the best fitness values ​​as leader individuals, and update the positions of the remaining individuals in the population;

[0111] Calculate and sort the updated individual fitness values, obtain elite individuals, and update the elite individual positions;

[0112] Repeat the above steps until the optimal solution is found or the maximum number of iterations of the individual optimization algorithm is reached, so as to obtain the optimal penalty factor and kernel function parameters of the support vector machine model, and improve the support vector machine model based on the optimal penalty factor and kernel function parameters.

[0113] The embodiments of the present invention are described in detail above with reference to the accompanying drawings, but the present invention is not limited thereto, and various changes can be made within the knowledge scope of technicians in the relevant technical field without departing from the purpose of the present invention.

Claims

1. A nursing process monitoring system, comprising a state classification unit (4), an intelligent display unit (5), a state monitoring unit (6) and a state matching unit (7), characterized in that: An electrocardiogram monitoring unit (1), the electrocardiogram monitoring unit (1) being used to perform electrocardiogram monitoring on a target patient through an electrocardiogram monitoring instrument to obtain an electrocardiogram signal of the target patient, and to transmit the electrocardiogram signal of the target patient to an electrocardiogram preprocessing unit (2); An ECG preprocessing unit (2), wherein the ECG preprocessing unit (2) receives the ECG signal of the target patient transmitted by the ECG monitoring unit (1), sets a filtering window, performs expansion processing on the ECG signal of the target patient based on the filtering window to obtain an expanded ECG signal, performs filtering processing on the expanded ECG signal to obtain a first ECG signal, performs signal normalization processing on the first ECG signal to obtain a second ECG signal, takes the R wave of the second ECG signal as the center, performs interception processing on the second ECG signal through a preset equal length threshold to obtain a standard ECG signal, and transmits the standard ECG signal of the target patient to a feature extraction unit (3); A feature extraction unit (3), wherein the feature extraction unit (3) receives the standard ECG signal of the target patient transmitted by the ECG preprocessing unit (2), and extracts features from the standard ECG signal of the target patient to obtain a feature set corresponding to the standard ECG signal, and transmits the feature set corresponding to the standard ECG signal to the state classification unit (4).

2. A nursing process monitoring system according to claim 1, characterized in that: The state classification unit (4) receives the feature set corresponding to the standard electrocardiogram signal transmitted by the feature extraction unit (3), inputs the feature set corresponding to the standard electrocardiogram signal into a trained state classification model, outputs the heart rhythm state corresponding to the target patient through the trained state classification model, and transmits the heart rhythm state corresponding to the target patient to the intelligent display unit (5) and the state matching unit (7); The state monitoring unit (6) is used to monitor the state of the target patient through a blood pressure sensor, a blood oxygen saturation sensor and a body temperature sensor to obtain state data of the target patient, wherein the state data includes blood pressure, blood oxygen saturation and body temperature, and a health real-time monitoring directed graph is constructed based on the preset relationship between heart rate, blood pressure, blood oxygen saturation and body temperature. The health real-time monitoring directed graph includes multiple nodes and multiple undirected edges, wherein the nodes correspond to the electrocardiogram monitoring instrument, the blood pressure sensor, the blood oxygen saturation sensor and the body temperature sensor, and the undirected edges correspond to the preset relationship between heart rate, blood pressure, blood oxygen saturation and body temperature. The weights of the nodes are assigned based on the state data of the target patient, and the weights of the undirected edges are assigned based on the preset relationship between heart rate, blood pressure, blood oxygen saturation and body temperature to obtain a weighted health real-time monitoring directed graph, and the weighted health real-time monitoring directed graph is transmitted to the intelligent display unit (5) and the state matching unit (7).

3. A nursing process monitoring system according to claim 2, characterized in that: The state matching unit (7) receives the heart rhythm state corresponding to the target patient transmitted by the state classification unit (4) and the weighted health real-time monitoring directed graph transmitted by the state monitoring unit (6), and matches the heart rhythm state corresponding to the target patient with a standard weighted health monitoring directed graph sequence pre-stored in a database to obtain a matching standard weighted health monitoring directed graph, compares the matching standard weighted health monitoring directed graph with the weighted health real-time monitoring directed graph to obtain an error between the two, matches the error between the two with a preset error threshold table to obtain the health status level of the target patient, and transmits the health status level of the target patient to the intelligent display unit (5).

4. A nursing process monitoring system according to claim 3, characterized in that: The intelligent display unit (5) receives the heart rhythm state corresponding to the target patient transmitted by the state classification unit (4), the weighted health real-time monitoring directed graph transmitted by the state monitoring unit (6), and the health status level of the target patient transmitted by the state matching unit (7), and displays the heart rhythm state corresponding to the target patient, the weighted health real-time monitoring directed graph, and the health status level of the target patient on the intelligent display terminal, and uploads the heart rhythm state corresponding to the target patient, the weighted health real-time monitoring directed graph, and the health status level of the target patient to the intelligent terminal of the nursing staff in real time.

5. A nursing process monitoring system according to claim 1, characterized in that: The formula for the expansion process is as follows: Where, d(n) represents the nth sample point after expansion, L represents the record length of each ECG signal, and k represents the filter window; The formula for the filtering process is as follows: s(o)=f med {f sort [d(o):d(o+k-1)]}; Where s(o) represents the filtered data of the oth point in the expanded ECG signal, and f sort represents the sorting function, f med Represents the median function; The formula for the normalization process is as follows: Among them, s * (o) represents the filtered data of the oth point in the second ECG signal, s min Indicates the minimum sample amplitude value in the current first ECG signal, s max Indicates the maximum value of the sample amplitude in the current first ECG signal; The formula for the interception process is as follows: U u (δ)=s * (d)*(R u -90+d); Among them, U u (δ) represents the δth sampling point data in the uth heartbeat, and u=1,2,...,N beat , N beat Indicates the number of R peak marks of the current ECG signal, R u represents the R wave in the u-th heart beat.

6. A nursing process monitoring system according to claim 1, characterized in that: Extracting features from the target patient's standard ECG signal to obtain a feature set corresponding to the standard ECG signal includes: The RR interval of the standard electrocardiogram signal is feature extracted by a first formula to obtain a first feature corresponding to the standard electrocardiogram signal. The first formula is as follows: RR(n)=R(n+1)-R(n); Where RR(n) represents the nth RR interval, and R(n) represents the nth R wave; Calculating the average value, standard deviation, standard deviation of the difference value and root mean square of the difference value of the first feature corresponding to the standard electrocardiogram signal to obtain the second feature corresponding to the standard electrocardiogram signal; The DRR interval of the standard ECG signal is feature extracted based on the first feature corresponding to the standard ECG signal by a second formula to obtain a third feature corresponding to the standard ECG signal. The second formula is as follows: DRR(n)=RR(n+1)-RR(n); Where DRR(n) represents the nth DRR interval; The skewness and kurtosis of the standard electrocardiogram signal are extracted by a third formula to obtain a fourth feature corresponding to the standard electrocardiogram signal. The third formula is as follows: Among them, K represents the skewness of the standard ECG signal, S represents the kurtosis of the standard ECG signal, and N hos Indicates the number of samples in this segment, U u_mean Represents the average value of the sampling point data in the u-th heart beat.

7. A nursing process monitoring system according to claim 6, characterized in that: Extracting features from the standard ECG signal of the target patient to obtain a feature set corresponding to the standard ECG signal also includes: Performing an arc cosine transform on the heart rate value corresponding to the standard ECG signal to convert the heart rate value into an angle value in a polar coordinate system, and converting the timestamp corresponding to the standard ECG signal into a radius in a polar coordinate system; constructing a polar coordinate matrix based on the angle value and radius in the polar coordinate system, wherein the polar coordinates of the elements in the polar coordinate matrix are represented by the angle between the timestamp and the corresponding heart rate value, and converting the polar coordinate data into Cartesian coordinates based on a conversion formula from polar coordinates to Cartesian coordinates; Performing an inner product on the data in the Cartesian coordinates to obtain a Gramian matrix, and mapping corresponding values ​​of the Gramian matrix onto an image to obtain a GAF image; Perform feature extraction on the GAF image to obtain image features corresponding to the GAF image, and combine the first feature, the second feature, the third feature, the fourth feature and the image features corresponding to the GAF image to obtain a feature set corresponding to the standard electrocardiogram signal.

8. A nursing process monitoring system according to claim 2, characterized in that: The state classification model adopts an improved support vector machine model, and the state classification model improves the support vector machine model through an optimization algorithm.

9. A nursing process monitoring system according to claim 8, characterized in that: The support vector machine model is improved by an optimization algorithm, including: Set the initialization parameters of the optimization algorithm to initialize the population position. The population position initialization formula is as follows: Among them, x t and x t+1 represents the position of the individual at the t-th iteration and the t+1-th iteration, and p represents the preset parameter for controlling the nonlinear mapping; Calculate the fitness value of the individual and calculate the elite individual position. The elite individual position calculation formula is as follows: in, represents the reverse solution of the j-th dimension individual at the t-th iteration, x i,j (t) represents the elite solution of the j-th dimension individual at the t-th iteration, k represents a random number in the interval [0,1], u j (t) represents the maximum position of the individual at the tth iteration, l j (t) represents the minimum position of the individual at the tth iteration.

10. A nursing process monitoring system according to claim 9, characterized in that: Improving the support vector machine model by an optimization algorithm also includes: Marking a preset number of the individuals with the best fitness values ​​as leader individuals, and updating the positions of the remaining individuals in the population; Calculate and sort the updated individual fitness values, obtain elite individuals, and update the elite individual positions; Repeat the above steps until the optimal solution is found or the maximum number of iterations of the individual optimization algorithm is reached, thereby obtaining the optimal penalty factor and kernel function parameters of the support vector machine model, and improving the support vector machine model based on the optimal penalty factor and kernel function parameters.

Images