A safe diagnostic method for medical monitoring

Abstract

The application discloses a safety diagnosis method for medical monitoring, and relates to the technical field of intelligent medical monitoring, and comprises the following steps: collecting physiological data of a patient through a sensor, and pre-processing the physiological data; constructing a multi-modal data analysis model; inputting the pre-processed physiological data into the multi-modal data analysis model for feature extraction and fusion, and outputting physiological state features; constructing a safety diagnosis model by using a deep neural network; inputting the physiological state features into the machine learning model, and outputting a health risk value; obtaining an early warning threshold value through a support vector machine (SVM) according to historical physiological data of the patient; and sending an alarm to medical staff when the health risk value exceeds the early warning threshold value.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of intelligent medical monitoring, and in particular to a safety diagnosis method for medical monitoring. BACKGROUND

[0002] In recent years, with the rapid development of sensor technology and computer science, the field of medical monitoring has made significant progress. Traditional medical monitoring usually relies on manual observation and evaluation by doctors and nurses, which is not only time-consuming and labor-intensive, but also susceptible to human factors, leading to increased subjectivity and uncertainty in monitoring results. With the rise of the Internet of Things technology, various types of sensors have been widely used in medical monitoring, such as heart rate monitors, blood pressure monitors, blood glucose meters, and thermometers. These sensors can collect patients' physiological data in real time, providing doctors with more objective and continuous monitoring means. In terms of data processing, traditional data processing methods often focus on a single type of physiological signal, such as electrocardiogram (ECG) or blood oxygen saturation (SpO2), which limits doctors' understanding of patients' overall health status. In recent years, multi-modal data analysis technology has gradually become a research hotspot. This method can simultaneously process multiple types of data, providing more comprehensive health assessment. In addition, with the development of machine learning and deep learning technologies, it has become possible to use these technologies for automatic feature extraction and model training, greatly improving the accuracy and efficiency of disease diagnosis. SUMMARY

[0003] In view of the above-mentioned existing problems, the present application is proposed.

[0004] Therefore, the present application provides a safety diagnosis method for medical monitoring to solve the problem that traditional preprocessing methods often fail to effectively remove noise and outliers, especially in the context of multi-modal data fusion, how to ensure data quality is a problem that needs to be solved. Secondly, existing multi-modal data analysis methods mostly use simple linear combination methods, which may ignore the internal relationship and interaction between different physiological data, thereby affecting the accuracy of the final diagnosis result. In addition, existing diagnosis models often focus on the analysis of a single indicator, ignoring the complex interaction between multiple physiological parameters, which may result in an incomplete and inaccurate diagnosis result. Finally, regarding the setting of the warning threshold, existing technologies often rely on fixed thresholds or simple statistical methods, lacking consideration of individual differences, which may result in insufficient sensitivity and specificity of the warning system.

[0005] To solve the above technical problems, the present application provides the following technical solutions:

[0006] In a first aspect, the present application provides a safety diagnosis method for medical monitoring, which comprises:

[0007] Collect physiological data of the patient through a sensor, and pre-process the physiological data;

[0008] Construct a multi-modal data analysis model, input the pre-processed physiological data into the multi-modal data analysis model for feature extraction and fusion, and output physiological state features;

[0009] Construct a safety diagnosis model using a deep neural network, input the physiological state features into the machine learning model, and output a health risk value;

[0010] According to the historical physiological data of the patient, obtain a warning threshold value through a support vector machine (SVM), and send an alarm to medical staff when the health risk value exceeds the warning threshold value.

[0011] As a preferred scheme of the safety diagnosis method for medical monitoring, the physiological data of the patient is collected through a sensor, and the physiological data is pre-processed, and the specific steps are as follows:

[0012] The physiological data of the patient is collected using a heart rate monitor, a sphygmomanometer, a blood glucose meter, and a thermometer, the physiological data is de-noised and abnormal values are removed, normalized, and converted into a zero-mean unit variance form;

[0013] The collected physiological data set is denoted as x, which can be expressed as x=(x1, x2,..., x4), x1 represents heart rate, x2 represents blood pressure, x3 represents blood glucose, and x4 represents body temperature.

[0014] As a preferred scheme of the safety diagnosis method for medical monitoring, the physiological data of the patient is collected through a sensor, and the physiological data is pre-processed, and the specific steps are as follows:

[0015] The collected physiological data x is extracted using a nonlinear feature mapping, and the expression is as follows:

[0016]

[0017] where f k represents the feature value of the kth physiological data, x k,i represents the value of the ith sampling point in the kth physiological data, n represents the number of sampling points, and alpha, beta, and gamma represent positive real number parameters, log represents a logarithmic function, and exp represents an exponential function.

[0018] As a preferred scheme of the safety diagnosis method for medical monitoring, the physiological data of the patient is collected through a sensor, and the physiological data is pre-processed, and the specific steps are as follows:

[0019] The features of different physiological data are fused into a comprehensive feature vector, and the expression is:

[0020]

[0021] wherein F represents the fused comprehensive feature vector, w k represents the weight of the kth physiological data, m represents the number of physiological data types, and w m f m represents the product of the feature value of the mth physiological data and its corresponding weight.

[0022] As a preferred scheme of the safety diagnosis method for medical monitoring, the output physiological state feature has the following specific steps:

[0023] Based on the fused feature vector, the physiological state feature is output, and the expression is:

[0024]

[0025] wherein S represents the physiological state feature, η represents the nonlinear transformation degree parameter of the physiological state feature S, and θ represents the scaling degree parameter of the comprehensive feature vector F in the exponential function.

[0026] As a preferred scheme of the safety diagnosis method for medical monitoring, the safety diagnosis model is constructed using a deep neural network, the physiological state feature is input into the machine learning model, and the health risk value is output, and the specific steps are:

[0027] Based on the physiological state feature S, a safety diagnosis model is constructed using a deep neural network, the physiological state feature is input into the safety diagnosis model, and the health risk value R is output, and the expression is:

[0028]

[0029] wherein sin represents a sine function, π represents a circular constant, t represents an integral variable, cos represents a cosine function, and dt represents a small change of t.

[0030] As a preferred scheme of the safety diagnosis method for medical monitoring, the safety diagnosis model is trained according to the historical physiological data of the patient, and the specific steps are:

[0031] The historical physiological data of the patient is collected, input into the safety diagnosis model, and the historical physiological state feature is output, and the best hyperplane is found by solving an optimization problem, and the expression is

[0032] w·S+b=0;

[0033] Wherein, w is the normal vector of the safety diagnosis model hyperplane, and b is the intercept of the multi-modal data analysis model hyperplane.

[0034] As a preferred scheme of the safety diagnosis method for medical care, the pre-warning threshold is obtained by the support vector machine, and when the health risk value exceeds the pre-warning threshold, an alarm is sent to the medical staff, and the specific steps are as follows:

[0035] The score is predicted by the safety diagnosis model, and the pre-warning threshold T is determined based on the score, and the expression is as follows:

[0036]

[0037] Wherein, alpha i represents the Lagrange multiplier, C represents the penalty parameter, N represents the number of samples, y i represents the label of the i-th sample, S i represents the physiological state feature vector of the i-th sample, and w and b represent key parameters.

[0038] When R>T, there is a health risk, and the system immediately sends an alarm to the medical staff;

[0039] When r≤T, there is no health risk, and the system remains in a monitoring state.

[0040] In a second aspect, the embodiments of the present application provide a computer device, comprising a memory and a processor, and the memory stores a computer program, wherein when the computer program is executed by the processor, any step of the safety diagnosis method for medical care according to the first aspect of the present application is realized.

[0041] In a third aspect, the embodiments of the present application provide a computer readable storage medium, which stores a computer program, wherein when the computer program is executed by the processor, any step of the safety diagnosis method for medical care according to the first aspect of the present application is realized.

[0042] The present application has the following beneficial effects: the present application realizes efficient collection, preprocessing, feature extraction, fusion and health risk assessment of physiological data of patients, finally informs the medical staff in time through a dynamic pre-warning mechanism, improves the accuracy and timeliness of diagnosis, and also significantly improves the quality and efficiency of medical services through intelligent means. BRIEF DESCRIPTION OF DRAWINGS

[0043] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings needed in the embodiment description will be briefly introduced as follows. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative labor.

[0044] Figure 1 Flow chart for the safety diagnosis method for medical monitoring in Example 1.

[0045] Figure 2 Flow chart for obtaining physiological state features in Example 1. DETAILED DESCRIPTION

[0046] In order to make the above objectives, characteristics and advantages of the present application more apparent, specific embodiments of the present application will be described in detail below with reference to the accompanying drawings.

[0047] In the following description, numerous specific details are set forth in order to provide a thorough understanding of the present application. However, it will be apparent to one skilled in the art that the present application can be practiced without the specific details given herein, that the present application can be practiced with other than the described implementations, and that variations of the described implementations can be made. Thus, the present application is not intended to be limited to the particular implementations described.

[0048] Secondly, the "one embodiment" or "embodiment" referred to herein means that the specific features, structures or characteristics can be included in at least one implementation of the present application. "In one embodiment" appearing in different places in the specification does not mean the same embodiment, nor is it an embodiment that is independent or alternative to other embodiments.

[0049] Example 1, with reference to Figure 1 and Figure 2 , is the first embodiment of the present application, which provides a safety diagnosis method for medical monitoring, comprising the following steps:

[0050] S1 collects physiological data of the patient through a sensor and pre-processes the physiological data, the specific steps being:

[0051] The physiological data of the patient is collected using a heart rate monitor, a sphygmomanometer, a blood glucose meter and a thermometer, the physiological data is de-noised and de-abnormalized, normalized and converted into a form of zero mean unit variance;

[0052] The collected physiological data set is denoted as x, which can be expressed as x = (x1, x2,..., x4), x1 represents heart rate, x2 represents blood pressure, x3 represents blood glucose and x4 represents body temperature;

[0053] By de-noising and de-abnormalizing, the accuracy of subsequent feature extraction is ensured, misdiagnosis caused by data quality problems is avoided, the physiological data is converted into a form of zero mean unit variance, different types of physiological data can be compared on the same scale, the generalization ability of the model is improved, the quality of the data is ensured, and reliable basic data is provided for subsequent feature extraction and fusion.

[0054] S2 constructs a multi-modal data analysis model, and inputs the preprocessed physiological data into the multi-modal data analysis model for feature extraction, and the specific steps are as follows:

[0055] The collected physiological data x is subjected to feature extraction by using nonlinear feature mapping, and the expression is as follows:

[0056]

[0057] wherein f k represents the feature value of the kth physiological data, x k,i represents the value of the ith sampling point in the kth physiological data, n represents the number of sampling points, a, b, g represent positive real number parameters, log represents a logarithmic function, and exp represents an exponential function;

[0058] By extracting features through nonlinear mapping, the complex relationship between physiological data can be captured, and the recognition ability of the model for physiological state is enhanced.

[0059] S3 fuses the feature vectors of different physiological data, and the specific steps are as follows:

[0060] The features of different physiological data are fused into a comprehensive feature vector, and the expression is as follows:

[0061]

[0062] wherein F represents the fused comprehensive feature vector, w k represents the weight of the kth physiological data, m represents the number of physiological data types, w m f m represents the product of the feature value of the mth physiological data and its corresponding weight;

[0063] The features of different physiological data are fused into a comprehensive feature vector, and the information of multi-modal data is fully utilized, the comprehensive judgment ability of the model is improved, and the physiological state of the patient can be more accurately represented, providing a rich information foundation for subsequent risk assessment.

[0064] S4 outputs the physiological state feature, and the specific steps are as follows:

[0065] Based on the fused feature vector, the physiological state feature is output, and the expression is as follows:

[0066]

[0067] wherein S represents the physiological state feature, h represents the nonlinear transformation degree parameter of the physiological state feature S, and q represents the scaling degree parameter of the comprehensive feature vector F in the exponential function.

[0068] S5 uses a deep neural network to build a safety diagnosis model, inputs the physiological state features into the machine learning model, and outputs a health risk value. The specific steps are as follows:

[0069] Based on the physiological state features S, a safety diagnosis model is built using a deep neural network, the physiological state features are input into the safety diagnosis model, and a health risk value R is output. The expression is:

[0070]

[0071] Where sin represents the sine function, π represents the circular constant, t represents the integral variable, cos represents the cosine function, and dt represents a small change in t.

[0072] Using a deep neural network to build a safety diagnosis model can automatically learn and extract complex features, improving the prediction accuracy of the model. By outputting a health risk value, it can provide quantitative evaluation for medical staff, which helps to early detect potential health problems.

[0073] S6 trains the safety diagnosis model according to the patient's historical physiological data. The specific steps are as follows:

[0074] Collect the patient's historical physiological data, input it into the safety diagnosis model, and output the historical physiological state features. By solving the optimization problem, the best hyperplane is found, and the expression is

[0075] w·S+b=0;

[0076] Where w is the normal vector of the hyperplane of the multi-modal data analysis model, and b is the intercept of the hyperplane of the multi-modal data analysis model.

[0077] The support vector machine determines the early warning threshold, solving the problem that the fixed threshold cannot adapt to individual differences.

[0078] S7 gets the early warning threshold through the support vector machine, and sends an alarm to the medical staff when the health risk value exceeds the early warning threshold. The specific steps are as follows:

[0079] Predict the score through the safety diagnosis model, and determine the early warning threshold T based on the score. The expression is:

[0080]

[0081] Where α i represents the Lagrange multiplier, C represents the penalty parameter, N represents the number of samples, y i represents the label of the i-th sample, S i represents the physiological state feature vector of the i-th sample, and w and b represent the key parameters.

[0082] When R > T, there is a health risk, and the system immediately sends an alarm to medical staff;

[0083] When R≤T, there is no health risk, and the system remains in a monitoring state;

[0084] When the health risk value exceeds the early warning threshold, an alarm can be immediately sent to medical staff, ensuring that measures can be taken in a timely manner when a health risk occurs.

[0085] The embodiment also provides a computer device suitable for the case of the safety diagnosis method for medical care, including a memory and a processor; the memory is used to store computer executable instructions, and the processor is used to execute the computer executable instructions to realize the safety diagnosis method for medical care proposed in the above embodiment.

[0086] The computer device can be a terminal, and the computer device includes a processor, a memory, a communication interface, a display screen and an input device connected through a system bus. The processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operating system and the computer program in the non-volatile storage medium to run. The communication interface of the computer device is used to communicate with external terminals in a wired or wireless manner. The wireless manner can be achieved through WIFI, an operator network, NFC (near field communication) or other technologies. The display screen of the computer device can be a liquid crystal display screen or an electronic ink display screen. The input device of the computer device can be a touch layer overlaid on the display screen, or a key, trackball or touchpad arranged on the shell of the computer device, or an external keyboard, touchpad or mouse, etc.

[0087] The embodiment also provides a storage medium on which a computer program is stored, the program being executed by a processor to implement the safety diagnosis method for medical care proposed in the above embodiment; the storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as a static random access memory (SRAM), an electrically erasable programmable read-only memory (EEPROM), an erasable programmable read-only memory (EPROM), a programmable read-only memory (PROM), a read-only memory (ROM), a magnetic memory, a flash memory, a magnetic disk or an optical disk.

[0088] To sum up, the present application realizes efficient collection, preprocessing, feature extraction, fusion and health risk assessment of physiological data of a patient, finally timely informs medical staff through a dynamic early warning mechanism, improves the accuracy and timeliness of diagnosis, and significantly improves the quality and efficiency of medical services through intelligent means.

[0089] In the embodiment 2, referring to Table 1, the experimental simulation data of the safety diagnosis method for medical care are given to further verify the technical solution of the present application.

[0090] The embodiment aims to verify the effectiveness and superiority of the proposed safety diagnosis method for medical care, especially in improving the accuracy of health risk assessment. The embodiment tests on a group of volunteers by comparing the traditional method with the method proposed in the present application.

[0091] Ten healthy adults aged 25 to 55 years old are selected as test objects, each participant wears a smart bracelet equipped with a heart rate monitor, a sphygmomanometer, a blood glucose meter and a thermometer, which can continuously collect physiological data of heart rate, blood pressure, blood glucose and temperature, the data is collected every minute, and the continuous collection lasts for 24 hours.

[0092] The original data is denoised, outliers are removed, and standardized to zero mean unit variance using pandas and scikit-learn libraries in Python programming language, nonlinear feature mapping is used to extract features, and different physiological data features are fused into a comprehensive feature vector by weighted average, a deep neural network model is built using TensorFlow library, and the above fused feature vector is used for training, outputting health risk value, collecting the physiological data of each participant in the past week, inputting the multi-modal data analysis model, outputting the historical physiological state characteristics, and finding the best hyperplane to determine the warning threshold, when the health risk value exceeds the warning threshold, the system will immediately send an alarm to the medical staff.

[0093] For comparison, a traditional health risk assessment method based on a single physiological indicator is also used, in addition, a shallow neural network-based model is also used as another comparison benchmark.

[0094] Specifically as shown in Table 1:

[0095] Table 1 Experimental Record Table

[0096]

[0097]

[0098] The embodiment compares three different health risk assessment methods to verify the advantages of the safety diagnosis method of the present application in improving the assessment accuracy.

[0099] The method of the present application: the safety diagnosis model constructed by the multi-modal data analysis model and the deep neural network outputs a lower health risk value, indicating that this method can more accurately assess health risks.

[0100] Comparison method 1: the traditional method using a single physiological indicator has a generally higher health risk value, indicating that this method may cause false positives and increase unnecessary medical resource consumption.

[0101] Comparison method 2: the method based on shallow neural network has improved compared to the traditional method, but is still less accurate in health risk assessment than the method of the present application.

[0102] From the table data, it can be seen that the health risk value of the method of the present application is generally lower than that of the other two methods, which indicates that the present application has the following significant advantages:

[0103] More accurate health risk assessment: due to the use of multi-modal data analysis and deep neural networks, the method of the present application can more comprehensively consider the interaction between multiple physiological data, thereby improving the accuracy of the assessment.

[0104] Lower false positive rate: Compared with the other two methods, the health risk value of the present method is lower, which means it can reduce unnecessary alarms and reduce the waste of medical resources.

[0105] More timely warning: The warning threshold obtained by the support vector machine can more accurately identify situations that need attention, thereby achieving more timely warning.

[0106] In summary, the present method has shown significant advantages in improving the accuracy and timeliness of health risk assessment. Compared with existing technologies, the present method can significantly improve the quality and efficiency of medical services in practical applications.

[0107] It should be noted that the above examples are only used to illustrate the technical solutions of the present application and are not limiting. Although the present application has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present application can be modified or replaced equivalently without departing from the spirit and scope of the present application, and they should be covered in the scope of the claims of the present application.

Claims

1. A safe diagnostic method for medical monitoring, characterized in that: The application relates to a safety diagnosis method for medical monitoring. physiological data of a patient is collected through a sensor, and the physiological data is preprocessed; a multi-modal data analysis model is constructed, the preprocessed physiological data is input into the multi-modal data analysis model for feature extraction and fusion, and physiological state features are output; a safety diagnosis model is constructed using a deep neural network, the physiological state features are input into the machine learning model, and a health risk value is output; the safety diagnosis model is constructed using the deep neural network, the physiological state features are input into the machine learning model, and the health risk value is output, and the specific steps are as follows: Using a deep neural network model as a basis for a safety diagnostic model, physiological state features are input into the safety diagnostic model, and a health risk value is output , expressed as: ； wherein, denotes a sine function, denotes the ratio of the circumference of a circle to its diameter, denotes the integral variable, denotes a cosine function, denotes a small change in an early warning threshold is obtained through a support vector machine SVM according to historical physiological data of the patient, and an alarm is sent to medical staff when the health risk value exceeds the early warning threshold; the feature extraction is performed, and the specific steps are as follows: Using nonlinear feature mapping on collected physiological data performing feature extraction, expressed as: ； in Indicates the The characteristic value of physiological data, Indicates the The first type of physiological data The value of the sampling point, represents the number of sampling points, represents a positive real parameter, represents the logarithmic function, represents the exponential function; features of different physiological data are fused into a comprehensive feature vector, and the expression is as follows: ； in represents the integrated feature vector after fusion, Indicates the The weight of physiological data, Indicates the number of types of physiological data, Indicates the The product of the characteristic value of the physiological data and its corresponding weight; the physiological state features are output based on the fused feature vector, and the expression is as follows: ； wherein, denotes a physiological status feature, denotes a physiological status feature a non-linear transformation degree parameter, denotes a combined feature vector a scaling degree parameter within an exponential function.

2. The secure diagnostic method for medical monitoring of claim 1, wherein: the physiological data of the patient is collected through the sensor, and the physiological data is preprocessed, and the specific steps are as follows: a heart rate monitor, a sphygmomanometer, a blood glucose meter and a thermometer are used to collect the physiological data of the patient, the physiological data is subjected to noise removal and abnormal value removal, is subjected to normalization treatment, and is converted into a zero-mean unit variance form; The set of collected physiological data is denoted as , may be denoted as , denotes heart rate, denotes blood pressure, denotes blood glucose, denotes body temperature.

3. The secure diagnostic method for medical monitoring of claim 2, wherein: the safety diagnosis model is trained according to the historical physiological data of the patient, and the specific steps are as follows: the historical physiological data of the patient is collected, is input into the safety diagnosis model, and historical physiological state features are output, an optimal hyperplane is found by solving an optimization problem, and the expression is as follows ； wherein, represents a normal vector of the safety diagnostic model hyperplane, represents an intercept of the multi-modal data analysis model hyperplane.

4. The secure diagnostic method for medical monitoring of claim 3, wherein: the early warning threshold is obtained through the support vector machine, and the alarm is sent to the medical staff when the health risk value exceeds the early warning threshold, and the specific steps are as follows: Predicting scores by a safety diagnostic model and determining a warning threshold based on the scores , the expression being: ； wherein, denotes a Lagrange multiplier, denotes a penalty parameter, denotes a number of samples, denotes a label of a -th sample, denotes a physiological state feature vector of a -th sample; When there is a health risk, the system immediately sends an alert to healthcare personnel; When there is no health risk, the system remains in a monitoring state.

5. A computer device comprising a memory and a processor, the memory storing a computer program, characterized in that: the processor executes the computer program to realize the steps of the safety diagnosis method for medical monitoring according to any one of claims 1-4.

6. A computer readable storage medium having stored thereon a computer program, characterized in that: the computer program is executed by the processor to realize the steps of the safety diagnosis method for medical monitoring according to any one of claims 1-4.

Images