A dynamic risk early warning system based on intelligent situation awareness

Abstract

The present invention provides a dynamic risk early warning system based on intelligent situation awareness, which relates to the technical field of situation awareness data risk early warning. The dynamic risk early warning system based on intelligent situation awareness includes: a data acquisition module; a real-time monitoring module; a situation awareness module; and a risk early warning module. By collecting the perception data of the patient to be visited in the preset collection area in real time, and simultaneously monitoring the changes of the perception data of the patient to be visited within the preset time period to obtain the correlation index, then monitoring the changes of the situation awareness of the patient to be visited within the preset time period according to the obtained correlation index to obtain the situation awareness data, and finally inputting the obtained correlation index and situation awareness data into the risk early warning model to obtain the risk early warning index, the real-time performance of the situation awareness risk early warning is improved, and the problem of low real-time performance of the situation awareness abnormal data risk early warning in the prior art is solved.

Description

Technical Field

[0001] The present invention relates to the technical field of situation awareness data risk early warning, and particularly to a dynamic risk early warning system based on intelligent situation awareness. Background Art

[0002] With the continuous progress of medical technology and the in-depth promotion of medical informatization construction, the amount of medical data shows an explosive growth trend. These data include multi-dimensional information such as patients' clinical information, test results, medical imaging data, genetic information, and lifestyle. The accumulation of these data provides a rich data source for intelligent situation awareness technology, making it possible to predict and warn diseases based on big data analysis. Secondly, with the improvement of medical level and the enhancement of patients' health awareness, precision medicine and personalized treatment have become important development directions in the medical field. The dynamic risk early warning system based on intelligent situation awareness can, through comprehensive analysis and real-time monitoring of patients' individual data, timely discover potential disease risks, provide personalized health management and medical treatment suggestions for patients, and thus achieve early prevention and treatment of diseases.

[0003] The existing dynamic risk early warning system forms a security situation view by integrating data from preset data sources, and at the same time uses big data analysis technology to process the acquired data in real time to identify abnormal patterns, and automatically triggers corresponding security protection measures according to the early warning information to achieve the extraction and early warning of risks and effective response. However, in the prior art, the collection of medical data often depends on various medical devices and sensors, and the real-time performance of these devices may be affected by network latency and device performance, resulting in a time lag for medical data to reach the early warning system. In addition, due to the real-time update of medical data, the real-time analysis efficiency of medical data is low, and there is a problem of low real-time performance in the risk early warning of situation awareness abnormal data. Summary of the Invention

[0004] The embodiments of the present invention solve the problem of low real-time performance in the risk early warning of situation awareness abnormal data in the prior art by providing a dynamic risk early warning system based on intelligent situation awareness, and achieve an improvement in the real-time performance of situation awareness risk early warning.

[0005] The embodiments of the present invention provide a dynamic risk early warning system based on intelligent situation awareness, including:

[0006] A data acquisition module, a real-time monitoring module, a situation awareness module, and a risk warning module; wherein, the data acquisition module is used to collect in real time the perception data of the patient to be visited within a preset acquisition area, the perception data includes face image data and physiological parameter data, the face image data is used to reflect in real time the facial features of the patient to be visited, the facial features are used to reflect the influence degree of the facial situation awareness of the patient to be visited, the physiological parameter data is used to reflect in real time the changes in the action behaviors of the patient to be visited, and the influence degree of the facial situation awareness is used to measure the influence degree of the facial features on the facial situation awareness of the patient to be visited within a preset time period; the real-time monitoring module is used to monitor in real time the changes in the perception data of the patient to be visited within a preset time period to obtain a correlation index, and the correlation index is used to measure the degree of association between the perception data of the patient to be visited and the action behaviors of the patient to be visited within a preset time period; the situation awareness module is used to monitor in real time the changes in the situation awareness of the patient to be visited within a preset time period according to the obtained correlation index to obtain situation awareness data, and the situation awareness data is used to reflect in real time the situation awareness state of the patient to be visited at the current time point; the risk warning module is used to input the obtained correlation index and situation awareness data into a risk warning model to obtain a risk warning index, and at the same time generate a corresponding risk level according to the risk warning index, and the risk warning index is used to measure the degree of risk of the situation awareness of the patient to be visited at the current time point.

[0007] Optionally, the specific steps for obtaining the influence degree of the facial situation awareness are as follows: detecting the skin color area in the face image of the patient to be visited within the preset acquisition area, and comparing the detected skin color area with the standard skin color sample in the preset database to obtain a skin color difference value; the influence degree of the facial situation awareness is calculated by the following formula:

[0008]

[0009] In the formula, h is the number of the patient to be visited, h = 1, 2,..., H, H is the total number of patients to be visited, q is the number of the preset acquisition area, q = 1, 2,..., Q, Q is the total number of preset acquisition areas, e is the natural constant, BAI h represents the influence degree of the facial situation awareness of the h-th patient to be visited within the preset acquisition area, F h.q represents the skin color difference value of the h-th patient to be visited in the q-th preset acquisition area, ΔF 0 represents the reference skin color difference value, β 1 represents the skin color area contrast correction factor, L h.q represents the skin color area contrast of the skin color area in the face image of the h-th patient to be visited in the q-th preset acquisition area, β 2 represents the skin color area saturation correction factor, Bh.q Indicates the skin color region saturation of the skin color region in the h-th patient face image to be visited within the q-th preset acquisition region.

[0010] Optionally, the relevance index is obtained by the following method: aligning the perception data within a preset time period at the same time stamp to obtain an influence rate, where the influence rate includes a first influence rate and a second influence rate; combining the obtained facial situation perception influence degree to obtain a relevance index, and the relevance index is calculated by the following formula:

[0011]

[0012] In the formula, t is the number of the same time stamp, t = 1, 2,..., T, and T is the total number of the same time stamps, GUAN h Indicates the relevance index of the h-th patient to be visited within the same time stamp. Indicates the first influence rate of the face image data of the h-th patient to be visited within the t-th same time stamp. Indicates the reference first influence rate. Indicates the second influence rate of the physiological parameter data of the h-th patient to be visited within the t-th same time stamp. Indicates the reference second influence rate, δ represents the situation perception risk influence factor, BAI 0 Indicates the reference influence degree of facial situation perception.

[0013] Optionally, the situation perception module includes a situation perception data acquisition unit, a data preprocessing unit, and a situation perception anomaly recognition unit; the situation perception data acquisition unit: used to monitor the situation perception change of the patient to be visited in real time within a preset time period and record the situation perception data of the patient to be visited; the data preprocessing unit: used to convert the format of the situation perception data; the situation perception anomaly recognition unit: used to judge whether the situation perception of the patient to be visited is equal to a preset situation perception value by monitoring the fluctuation of the situation perception data in real time within a preset time period. If not, it is recorded as situation perception anomaly data and fed back to a preset person, otherwise it is stored in a preset database.

[0014] Optionally, after recording the situation perception anomaly data and feeding it back to a preset person, it further includes classifying the situation perception anomaly data according to the obtained situation perception anomaly degree. The specific acquisition steps of the situation perception anomaly degree are as follows: obtaining an anomaly deviation degree according to the deviation degree between the situation perception anomaly data and the situation perception data; obtaining an anomaly duration according to the time difference between the anomaly start time and the anomaly end time of the situation perception anomaly data; combining the obtained anomaly deviation degree, anomaly duration, and the action behavior deviation of the patient to be visited within a preset time period to obtain the situation perception anomaly degree.

[0015] Optionally, after inputting the obtained correlation index and situation awareness data into the risk early warning model, the method further includes obtaining a risk early warning index through the risk early warning model. The specific obtaining steps of the risk early warning index include: obtaining a risk probability value according to the situation awareness abnormal data and the situation awareness data, and simultaneously combining the obtained correlation index and the situation awareness abnormal degree to obtain the risk early warning index.

[0016] Optionally, the expression for obtaining the risk early warning index through the risk early warning model is:

[0017]

[0018] In the formula, q is the number of the preset acquisition area, q = 1, 2,..., Q, Q is the total number of the preset acquisition areas, e is the natural constant, GUAN h represents the correlation index of the h-th patient to be visited within the same time stamp, GUAN 0 represents the reference correlation index, YI h represents the situation awareness abnormal degree of the situation awareness abnormal data of the h-th patient to be visited within the preset time period, YI 0 represents the reference situation awareness abnormal degree, G h.q represents the risk probability value of the h-th patient to be visited within the q-th preset acquisition area, G 0 represents the reference risk probability value.

[0019] The above technical solution has at least the following beneficial effects compared with the prior art:

[0020] In the above solution, a dynamic risk early warning system based on intelligent situation awareness collects the perception data of the patient to be visited in the preset acquisition area in real time, and simultaneously monitors the change of the perception data of the patient to be visited within the preset time period to obtain the correlation index. Then, according to the obtained correlation index, it monitors the change of the situation awareness of the patient to be visited within the preset time period to obtain the situation awareness data. Finally, it inputs the obtained correlation index and the situation awareness data into the risk early warning model to obtain the risk early warning index, thereby realizing more accurate acquisition of the risk early warning index, further realizing the improvement of the real-time performance of the situation awareness risk early warning, and effectively solving the problem of low real-time performance of the situation awareness abnormal data risk early warning in the prior art. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following will briefly introduce the drawings required for the description of the embodiments. Obviously, the following drawings are only some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0022] Figure 1 It is a schematic structural diagram of a dynamic risk early warning system based on intelligent situation awareness provided by an embodiment of the present invention;

[0023] Figure 2 It is a three-dimensional coordinate analysis diagram of the influence degree of facial situation awareness provided by an embodiment of the present invention;

[0024] Figure 3 It is a schematic structural diagram of a situation awareness module in a dynamic risk early warning system based on intelligent situation awareness provided by an embodiment of the present invention. Detailed implementation manners

[0025] To make the objectives, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are some, but not all, of the embodiments of the present invention. All other embodiments obtained by those of ordinary skill in the art based on the described embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0026] Unless otherwise defined, the technical terms or scientific terms used in the present invention shall have the ordinary meanings understood by those of ordinary skill in the art to which the present invention pertains. The "first", "second" and similar terms used in the present invention do not denote any order, quantity or importance, but are only used to distinguish different components. Similarly, the terms such as "a", "an" or "the" do not denote a quantity limitation, but mean that there is at least one. The terms such as "including" or "comprising" mean that the elements or objects appearing before this term cover the elements or objects listed after this term and their equivalents, without excluding other elements or objects. The terms such as "connected" or "coupled" are not limited to physical or mechanical connections, but may include electrical connections, whether direct or indirect.

[0027] It should be noted that the "upper", "lower", "left", "right", "front", "rear", etc. used in the present invention are only used to represent relative positional relationships. When the absolute position of the object being described changes, the relative positional relationship may also change accordingly.

[0028] In view of the fact that the real-time acquisition of medical data in the present invention may be affected by network latency and device performance, resulting in a time lag for medical data to reach the early warning system. In addition, due to the real-time update of medical data, the real-time analysis efficiency of medical data is low, thereby reducing the real-time performance of the situation awareness abnormal data risk early warning. The present invention provides a dynamic risk early warning system based on intelligent situation awareness and effectively improves the real-time performance of the situation awareness abnormal data risk early warning.

[0029] As Figure 1 shown, it is a schematic structural diagram of a dynamic risk early warning system based on intelligent situation awareness provided by an embodiment of the present invention. A dynamic risk early warning system based on intelligent situation awareness provided by an embodiment of the present invention includes: a data acquisition module, a real-time monitoring module, a situation awareness module, and a risk early warning module; wherein, the data acquisition module is used to collect in real time the perception data of the patient to be visited within a preset acquisition area, and the perception data includes face image data and physiological parameter data. The face image data is used to reflect in real time the facial features of the patient to be visited, and the facial features are used to reflect the influence degree of the facial situation awareness of the patient to be visited. The physiological parameter data is used to reflect in real time the change of the action behavior of the patient to be visited, and the influence degree of the facial situation awareness is used to measure the influence degree of the facial features on the facial situation awareness of the patient to be visited within a preset time period; the real-time monitoring module is used to monitor in real time the change of the perception data of the patient to be visited within a preset time period to obtain a correlation index, and the correlation index is used to measure the correlation degree between the perception data of the patient to be visited and the action behavior of the patient to be visited within a preset time period; the situation awareness module is used to monitor in real time the change of the situation awareness of the patient to be visited within a preset time period according to the obtained correlation index to obtain situation awareness data, and the situation awareness data is used to reflect in real time the situation awareness state of the patient to be visited at the current time point; the risk early warning module is used to input the obtained correlation index and situation awareness data into a risk early warning model to obtain a risk early warning index, and at the same time generate a corresponding risk level according to the risk early warning index. The risk early warning model is used to obtain the risk early warning index, and the risk early warning index is used to measure the risk degree of the situation awareness of the patient to be visited at the current time point.

[0030] In this embodiment, in practical applications, although the degree of facial pallor is usually directly related to health problems such as anemia and low blood pressure and can be preliminarily judged by visual observation, in many cases, this judgment may be affected by various factors such as light, skin type, and makeup, thus being inaccurate or incomplete. Therefore, the degree of impact on facial situation awareness of the patient is usually used to replace the degree of facial pallor. Among them, the degree of impact on facial situation awareness is not limited to the single dimension of skin color, but comprehensively considers the overall changes of the face and the impact of these changes on the patient's health status. This indicator can be obtained through advanced image processing technology and biosensors, which can more accurately capture the subtle changes in the face and infer the patient's health status based on this; physiological parameter data usually includes blood circulation state data and blood loss degree data. Among them, the blood circulation state data is used to reflect the patient's blood circulation state, such as blood flow velocity and vascular resistance, to evaluate whether the patient's circulatory system function is normal; in the application scenario of hemorrhagic shock, the blood loss degree data is used to evaluate the patient's blood loss degree; when the patient is in the early stage of hemorrhagic shock, abnormal expressions usually appear due to emotions such as pain, tension, and fear. These abnormal expressions may be manifested as frowning, closing eyes, clenching teeth, etc., reflecting the patient's discomfort and pain. As hemorrhagic shock develops, the patient usually shows confusion, apathy, or even coma. Therefore, the degree of abnormal expression can be used as a preliminary indicator for evaluating the risk of hemorrhagic shock in the patient, especially in the early stage of shock; during hemorrhagic shock, due to the sharp decrease in blood circulation volume, the patient's face often shows obvious pallor, and the degree of facial pallor is positively correlated with the risk of hemorrhagic shock, that is, the more severe the pallor, the higher the risk of hemorrhagic shock.

[0031] It should be noted that the situation awareness module plays a core role in the dynamic risk warning system. It monitors the changes in the situation awareness of the patient to be visited in real time according to the obtained correlation indicators (including face image data and physiological parameter data) within a preset time period. Usually, the situation awareness data is collected in real time through medical sensor devices. Among them, these situation awareness data usually include heart rate, blood pressure, and blood oxygen saturation, which can reflect the situation awareness state of the patient at the current time point in real time. For example, during hemorrhagic shock, due to the reduction of blood volume and compensatory sympathetic nerve excitement, the patient's heart rate usually increases (the increase in heart rate is an early manifestation of hemorrhagic shock), which improves the accuracy of real-time monitoring and data processing of medical sensors; for patients with chronic diseases such as diabetes and hypertension, the situation awareness module can continuously monitor the patient's physiological indicators, timely detect abnormal situations, and prevent the occurrence of complications; for the elderly living alone, the situation awareness module can monitor their daily activities, sleep quality, etc., evaluate their independent living ability, and timely detect potential health risks; in summary, the situation awareness data can not only reflect the patient's current health status in real time, but also predict health risks, achieving an improvement in the real-time nature of situation awareness risk warning.

[0032] Optionally, the specific steps for obtaining the degree of influence of facial situation perception are as follows: Detect the skin color area in the face image of the patient to be visited within the preset acquisition area, compare the detected skin color area with the standard skin color samples in the preset database to obtain the skin color difference value. The face image is collected by an image sensor within the preset acquisition area. The skin color area represents the facial area of the patient to be visited during the face detection process. The skin color difference value is used to measure the deviation degree between the skin color in the skin color area of the patient to be visited and the standard skin color samples in the preset database. The degree of influence of facial situation perception is calculated by the following formula:

[0033]

[0034] In the formula, h is the number of the patient to be visited, h = 1, 2,..., H, where H is the total number of patients to be visited; q is the number of the preset acquisition area, q = 1, 2,..., Q, where Q is the total number of preset acquisition areas; e is the natural constant, BAI h represents the degree of influence of facial situation perception of the h-th patient to be visited within the preset acquisition area, F h.q represents the skin color difference value of the h-th patient to be visited in the q-th preset acquisition area, ΔF 0 represents the reference skin color difference value, β 1 represents the skin color area contrast correction factor, L h.q represents the skin color area contrast of the skin color area in the face image of the h-th patient to be visited in the q-th preset acquisition area, β 2 represents the skin color area saturation correction factor, B h.q represents the skin color area saturation of the skin color area in the face image of the h-th patient to be visited in the q-th preset acquisition area.

[0035] Among them, the skin color area contrast correction factor and the skin color area saturation correction factor are obtained from the preset database. The skin color area contrast correction factor is used to correct the deviation between the skin color area contrast in the face image of the patient to be visited and the degree of influence of facial situation perception. The skin color area saturation correction factor is used to correct the deviation between the skin color area saturation in the face image of the patient to be visited and the degree of influence of facial situation perception.

[0036] In this embodiment, a face detection algorithm is usually used to locate the face region in the image, crop the image according to the detected face region, and scale it to a suitable size for subsequent processing. The image is converted from the RGB color space to a color space more suitable for skin color detection (such as YCrCb). YCrCb is commonly used in skin color detection because the Cr and Cb components are relatively insensitive to illumination changes. A machine learning-based method (such as SVM, decision tree) is used to segment the skin color region in the converted color space, and color histograms, average colors, and color moment features are extracted from the detected skin color region. In the preset database, there should be multiple standard skin color samples, which may come from different ethnic groups, genders, and age groups. For the skin color characteristics of the patient to be visited, the difference value is calculated with each standard skin color sample in the database, usually obtained by calculating the Euclidean distance between the features.

[0037] Specifically, based on the standard skin color samples in the preset database, a series of thresholds for evaluating the impact degree of facial situation awareness are set. These thresholds can be numerical ranges based on skin color characteristics and are used to distinguish different degrees of facial situation awareness. For example, the following thresholds can be set: normal skin color range: representing the skin color characteristic range in a healthy state; mild pallor threshold: slightly deviating from the normal skin color range but still within an acceptable lower value; moderate pallor threshold: significantly lower than the normal skin color range but not reaching the boundary of severe pallor; severe pallor threshold: significantly lower than the normal skin color range, usually indicating a serious problem with the health condition. If the skin color difference value is within the normal skin color range, it is considered that there is no significant pallor on the patient's current face; if the skin color difference value exceeds the mild pallor threshold but does not reach the moderate pallor threshold, it is considered that the patient's face shows mild pallor; if the skin color difference value exceeds the moderate pallor threshold but does not reach the severe pallor threshold, it is considered that the patient's face shows moderate pallor; if the skin color difference value exceeds the severe pallor threshold, it is considered that the patient's face shows severe pallor, and immediate medical treatment should be sought.

[0038] Specifically, for the sake of simplifying the analysis, define B1 = β 1 *L h.q +β 2 *B h.q , where F1 represents the skin color difference value coefficient of the hth patient to be visited in the qth preset acquisition area, and B1 represents the skin color area impact coefficient of the hth patient to be visited in the qth preset acquisition area. The simplified calculation formula for the impact degree of facial situation awareness is: Among them, the reference skin color difference value is usually represented by the result of summing and averaging the skin color data in the preset database, such as Figure 2As shown in the figure, it is a three-dimensional coordinate analysis diagram of the influence degree of facial gesture perception provided by the embodiment of the present invention. It can be seen from the figure that the influence degree of facial gesture perception increases with the increase of the skin color difference value coefficient and the skin color area influence coefficient, and the influence of the skin color difference coefficient on the influence degree of facial gesture perception is more obvious than that of the skin color area influence coefficient.

[0039] It should be understood that the skin color area contrast correction factor is obtained from a preset database. In a specific embodiment, the relationship between the skin color area contrast and the influence degree of facial gesture perception is fitted through the historical skin color image contrast data and historical gesture perception data in the preset database to obtain a fitting curve. According to the fitting curve, the relationship between the skin color area contrast and the influence degree of facial gesture perception is determined, and the real-time skin color area contrast is substituted into the fitting curve to obtain the corresponding skin color area contrast correction factor.

[0040] The skin color area saturation correction factor is obtained from a preset database. In a specific embodiment, the relationship between the skin color area saturation and the influence degree of facial gesture perception is fitted through the historical skin color image saturation data and historical gesture perception data in the preset database to obtain a fitting curve. According to the fitting curve, the relationship between the skin color area saturation and the influence degree of facial gesture perception is determined, and the real-time skin color area saturation is substituted into the fitting curve to obtain the corresponding skin color area saturation correction factor, which improves the accuracy and reliability of obtaining the influence degree of facial gesture perception, and further improves the real-time performance of gesture perception risk warning, effectively solving the problem of low real-time performance of gesture perception abnormal data risk warning in the prior art.

[0041] Optionally, the relevance index is obtained by the following method: Align the perception data within a preset time period at the same time stamp to obtain the influence rate, which includes the first influence rate and the second influence rate. The first influence rate is used to measure the influence degree of the face image data on the gesture perception risk within the same time stamp, and the second influence rate is used to measure the influence degree of the physiological parameter data on the gesture perception risk within the same time stamp; Combine the obtained influence degree of facial gesture perception to obtain the relevance index, and the relevance index is calculated by the following formula:

[0042]

[0043] In the formula, t is the number of the same time stamp, t = 1, 2,..., T, T is the total number of the same time stamps, GUAN h represents the relevance index of the hth patient to be visited within the same time stamp, represents the first influence rate of the face image data of the hth patient to be visited at the tth same time stamp, represents the reference first influence rate, It represents the second impact rate of the physiological parameter data of the h-th patient to be visited within the t-th same time stamp. It represents the reference second impact rate, δ represents the situation awareness risk impact factor, and BAI 0 It represents the reference impact degree of facial situation awareness.

[0044] Among them, the situation awareness risk impact factor is obtained from a preset database, and the situation awareness risk impact factor is used to measure the impact degree of the change of the situation awareness of the patient to be visited within a preset time period on the situation awareness impact degree.

[0045] In this embodiment, the same time stamp is the time stamp shared by the face image data and the physiological parameter data within a preset time period. Features related to situation awareness risk are extracted from the preprocessed face image, such as the degree of paleness of the skin color and the change of facial blood vessel distribution. These features can be monitored in real time by medical sensors, while the physiological parameter features can directly extract features related to situation awareness risk from the physiological parameter data, such as heart rate variability and blood pressure drop rate. These features can usually be directly used for analysis; in the field of medical applications, the first impact rate is usually obtained by the ratio of the total number of abnormal face situation awareness data of the patient to the total number of face situation awareness data, and the second impact rate is obtained by monitoring the fluctuation frequency of the patient's heart rate within a preset time period and the fluctuation frequency at each time stamp within a preset time period by a medical sensor device. Among them, the second impact rate represents the ratio of the fluctuation frequency of the patient at each time stamp to the fluctuation frequency of the patient within a preset time period.

[0046] Specifically, the reference first impact rate is usually represented by the result of summing and averaging the face image data in the preset database, the reference second impact rate is usually represented by the result of summing and averaging the physiological parameter data in the preset database, and the reference impact degree of facial situation awareness is usually represented by the result of summing and averaging the facial situation awareness data in the preset database.

[0047] Specifically, the situation awareness risk impact factor is obtained from a preset database. In a specific embodiment, the relationship between the facial situation awareness impact degree and the correlation index is fitted by the historical facial situation awareness data and the historical situation awareness abnormal data in the preset database to obtain a fitting curve. According to the fitting curve, the relationship between the facial situation awareness impact degree and the correlation index is determined, and the real-time facial situation awareness data is substituted into the fitting curve to obtain the corresponding situation awareness risk impact factor.

[0048] It should be understood that the algorithm in this embodiment combines the first influence rate, the second influence rate, and the influencing factors of facial situation perception degree, and comprehensively analyzes to obtain the correlation index. The first influence rate and the influencing factors of facial situation perception degree in this formula do not only unidirectionally affect the value of the correlation index. The first influence rate also indirectly affects the value of the influencing factors of facial situation perception degree. When medical interventions such as drug treatment or surgery effectively improve the patient's physiological condition (i.e., the first influence rate increases), the patient's emotional state often also improves accordingly, thus showing more positive postures (such as relaxation, smiling) in facial expressions, that is, the amplitude of the patient's action behavior increases, which is the improvement of the influencing factors of facial situation perception degree. On the contrary, if the patient expresses a positive attitude towards treatment (such as actively cooperating with treatment, remaining optimistic) through non-verbal means such as facial expressions, this psychological state may also promote physiological recovery in turn, thereby further increasing the first influence rate, achieving the improvement of the accuracy and efficiency of obtaining the influencing factors of facial situation perception degree, and then achieving the improvement of the real-time performance of situation perception risk warning, effectively solving the problem of low real-time performance of situation perception abnormal data risk warning in the prior art.

[0049] Optionally, as Figure 3 shown, it is a schematic structural diagram of a situation perception module in a dynamic risk warning system based on intelligent situation perception provided by an embodiment of the present invention. The situation perception module includes a situation perception data acquisition unit, a data preprocessing unit, and a situation perception anomaly recognition unit; Situation perception data acquisition unit: It is used to monitor the situation perception change of the patient to be visited in real time within a preset time period and record the situation perception data of the patient to be visited; Data preprocessing unit: It is used to perform format conversion on the situation perception data. The format conversion is used to convert non-numerical data in the situation perception data into numerical data; Situation perception anomaly recognition unit: It is used to judge whether the situation perception of the patient to be visited is equal to a preset situation perception value by monitoring the fluctuation of the situation perception data in real time within a preset time period. If it is not equal, it is recorded as situation perception abnormal data and fed back to a preset person, otherwise it is stored in a preset database. The situation perception abnormal data is situation perception data that exceeds the preset situation perception threshold.

[0050] In this embodiment, the situation awareness data acquisition unit is responsible for real-time monitoring of the situation awareness of the user to be accessed within a preset time period, such as heart rate, blood pressure, respiratory rate, and body temperature. This is usually achieved by connecting to medical monitoring devices of the patient (such as electrocardiogram monitors, sphygmomanometers, and blood oxygen saturation monitors). These devices can continuously and automatically collect the situation awareness data of the patient, improving the accuracy of the acquisition of situation awareness data. In practical applications, a sphygmomanometer is usually used to measure the blood pressure of the patient within a preset time period, that is, ambulatory blood pressure monitoring. Ambulatory blood pressure monitoring can continuously record the blood pressure changes within a preset time period (such as 24 hours), avoiding the contingency and instability of single measurement. Secondly, ambulatory blood pressure monitoring helps to detect masked hypertension, that is, the phenomenon that the patient's blood pressure is normal when measured in the hospital but increases in daily life. It should be noted that two types of hypertension, postural hypertension and postprandial hypotension, are easily overlooked in routine blood pressure monitoring, but ambulatory blood pressure monitoring can accurately identify them.

[0051] Therefore, in practical applications, the blood pressure of the patient within a preset time period is usually used as the situation awareness data, and it can also be selected according to the specific application scenario. The situation awareness data in the following embodiments are all the blood pressure of the patient within a preset time period, and the corresponding situation awareness abnormal data are all the abnormal blood pressure of the patient within a preset time period. Among them, the blood pressure of the patient within a preset time period is usually obtained by real-time monitoring with a sphygmomanometer. At this time, the preset situation awareness value is the preset blood pressure value, which can accurately reflect the change of the patient's health status. It is usually represented by the result of summing and averaging the historical blood pressure data in the preset database. If the patient's blood pressure is greater than the preset blood pressure value, it is marked as abnormal blood pressure. These blood pressure abnormal data may indicate that the patient has some health problems or risks (such as cerebral infarction caused by hypertension), which requires medical staff to pay timely attention and treatment, improving the real-time and accuracy of the patient's situation awareness monitoring.

[0052] Optionally, it is recorded as situation awareness abnormal data and fed back to the preset personnel. After that, it further includes classifying the situation awareness abnormal data according to the obtained degree of situation awareness abnormality. The degree of situation awareness abnormality is used to measure the abnormality degree of situation awareness within a preset time period. The classification includes mild abnormality, moderate abnormality, and severe abnormality. The specific steps for obtaining the degree of situation awareness abnormality are as follows: obtaining the abnormal deviation degree according to the deviation degree between the situation awareness abnormal data and the situation awareness data. The abnormal deviation degree represents the absolute value of the difference between the situation awareness data and the situation awareness abnormal data; obtaining the abnormal duration according to the time difference between the abnormal start time and the abnormal end time of the situation awareness abnormal data. The abnormal duration is the duration of the abnormal state within the preset time period; combining the obtained abnormal deviation degree, abnormal duration, and the action behavior deviation of the patient to be visited within the preset time period to obtain the degree of situation awareness abnormality. The action behavior deviation is used to measure the deviation degree between the action behavior of the patient to be visited within the preset time period and the preset behavior threshold.

[0053] In this embodiment, the content is the same as that of the previous embodiment. The situation awareness abnormal data is usually the abnormal blood pressure of the patient within a preset time period. Usually, the normal range of blood pressure (i.e., the normal range of situation awareness) is obtained by summing and averaging the historical blood pressure data in the preset database. When the patient's blood pressure deviates from the normal range of blood pressure, the patient needs to be treated, such as performing cardiopulmonary resuscitation and using first aid drugs; for each situation awareness abnormal data point (such as blood pressure), calculate its deviation from the normal range of blood pressure. The deviation can be the absolute difference (i.e., the abnormal value minus the central value or the lower / upper limit of the normal range), or the deviation percentage (i.e., the deviation value divided by half of the normal range or a certain fixed value, and then multiplied by 100%). According to the calculated deviation, determine the abnormal deviation degree. This usually requires mapping the deviation value to the corresponding classification, such as "mild deviation", "moderate deviation", and "severe deviation". In practical applications, assume that the normal heart rate range of a certain patient is 60 - 100 beats per minute, and the real-time monitored heart rate is 120 beats per minute. Then the absolute deviation of the heart rate is 120 - 100 = 20 beats per minute, and the deviation percentage is (20 / 50) * 100% = 40%. According to the preset deviation degree classification standard, this heart rate value may be regarded as "moderate deviation"; usually, by wearing a speed sensor on the patient, the patient's movement speed (such as the speed of lying down, getting up, and walking) can be recorded in real time, which is used to reflect the patient's exercise intensity, and then compared with the preset movement speed to obtain the patient's action behavior deviation. In practical applications, the action behavior of the patient within the preset time period is usually the walking speed. At this time, the preset movement speed is the preset walking speed. Among them, the preset walking speed is usually represented by the result of summing and averaging the historical walking speed data in the preset database. Therefore, the action behavior deviation in this embodiment represents the difference between the walking speed of the patient within the preset time period and the preset walking speed.

[0054] Specifically, the degree of situation awareness anomaly is calculated by the following formula:

[0055]

[0056] In the formula, h is the number of the patient to be visited, h = 1, 2,..., H, where H is the total number of patients to be visited; t is the number of the same time stamps in the preset time period, t = 1, 2,..., T, where T is the total number of the same time stamps in the preset time period; m is the number of the situation awareness anomaly data, m = 1, 2,..., M, where M is the total number of the situation awareness anomaly data, and YI h represents the degree of situation awareness anomaly of the h-th patient to be visited, γ 1 represents the weight factor of anomaly deviation degree, D h.m represents the anomaly deviation degree of the m-th situation awareness anomaly data of the h-th patient to be visited within the preset time period, D 0 represents the reference anomaly deviation degree, γ 2 represents the weight factor of anomaly duration, J h.m represents the anomaly duration of the m-th situation awareness anomaly data of the h-th patient to be visited within the preset time period, J 0 represents the reference anomaly duration, γ 3 represents the weight factor of action behavior deviation, S h.t represents the action behavior deviation of the h-th patient to be visited at the t-th same time stamp, S 0 represents the reference action behavior deviation.

[0057] Among them, the weight factor of anomaly deviation degree, the weight factor of anomaly duration, and the weight factor of action behavior deviation are obtained through a preset database. The weight factor of anomaly deviation degree is used to measure the proportion of the anomaly deviation degree relative to the degree of situation awareness anomaly. The weight factor of anomaly duration is used to measure the proportion of the anomaly duration relative to the degree of situation awareness anomaly. The weight factor of action behavior deviation is used to measure the proportion of the action behavior deviation relative to the degree of situation awareness anomaly.

[0058] It should be noted that during the calculation of the degree of situation awareness anomaly, when D h.m ≥D 0 , J h.m ≥J 0 and S h.t ≥S 0 , where D 0 is the reference anomaly deviation degree (i.e., the minimum value of the anomaly deviation degree), J 0 is the reference anomaly duration (i.e., the minimum value of the anomaly duration), S 0 is the reference action behavior deviation (i.e., the minimum value of the action behavior deviation), when Dh.m <D 0 or J h.m <J 0 or S h.t <S 0 When it is the case of <D> or <J> or <S>, there is no need to calculate the degree of anomaly in situation awareness. At this time, it indicates that the situation awareness data is normal.

[0059] Specifically, the action behavior deviation is used to measure the deviation degree between the action behavior of the patient to be visited within a preset time period and the preset behavior threshold. This includes various behaviors of the patient such as daily activity patterns, sleep quality, eating habits, etc., and can evaluate whether the patient's lifestyle is healthy and whether there are behavior deviations that may lead to health problems.

[0060] Among them, the preset behavior threshold is usually represented by the result of summing and averaging the historical patient behavior data in the preset medical database; the reference anomaly deviation degree is usually represented by the result of summing and averaging the historical situation awareness data in the preset medical database, the reference duration is usually represented by the result of summing and averaging the historical duration data in the preset database, and the action behavior reference deviation is usually represented by the result of summing and averaging the historical patient behavior data in the preset medical database.

[0061] It should be understood that the algorithm in this embodiment combines factors such as anomaly deviation degree, anomaly duration, and action behavior deviation, and comprehensively analyzes to obtain the degree of anomaly in situation awareness. The anomaly deviation degree and action behavior deviation in this formula not only unilaterally affect the value of the degree of anomaly in situation awareness, but also the action behavior deviation indirectly affects the value of the degree of anomaly in situation awareness. When it is found that the patient's action behavior deviation is large (such as sitting still for a long time, frequently consuming high-sugar foods), and is accompanied by abnormal deviation of physiological indicators (such as continuous increase in blood pressure, large blood sugar fluctuations), the algorithm will comprehensively evaluate and obtain a higher degree of anomaly in situation awareness. At this time, the dynamic risk warning system will automatically send a health reminder to the patient, such as suggesting increasing physical activity and adjusting the diet structure, and at the same time send the evaluation result to the doctor so that the doctor can timely adjust the treatment plan or perform optional medical interventions.

[0062] The anomaly deviation degree weight factor is obtained from the preset database. In a specific embodiment, the relationship between the anomaly deviation degree and the degree of anomaly in situation awareness is obtained by fitting the historical situation awareness data and historical situation awareness anomaly data in the preset database to obtain a fitting curve. According to the fitting curve, the relationship between the anomaly deviation degree and the degree of influence on situation awareness is determined, and the real-time anomaly deviation degree is substituted into the fitting curve to obtain the corresponding anomaly deviation degree weight factor, where the anomaly deviation degree represents the ratio of the total number of situation awareness anomaly data to the total number of situation awareness data.

[0063] The abnormal duration weight factor is obtained from a preset database. In a specific embodiment, the relationship between the abnormal duration and the situation awareness abnormal degree is obtained by fitting the historical situation awareness data duration and the historical situation awareness abnormal data duration in the preset database to obtain a fitting curve. According to the fitting curve, the relationship between the abnormal duration and the situation awareness influence degree is determined, and the abnormal duration of the real-time situation awareness abnormal data is substituted into the fitting curve to obtain the corresponding abnormal duration weight factor.

[0064] The action behavior deviation weight factor is obtained from a preset database. In a specific embodiment, the relationship between the action behavior deviation and the situation awareness abnormal degree is obtained by fitting the historical facial situation awareness data and the historical action behavior deviation data in the preset database to obtain a fitting curve. According to the fitting curve, the relationship between the action behavior deviation and the situation awareness influence degree is determined, and the action behavior deviation of the real-time patient is substituted into the fitting curve to obtain the corresponding action behavior deviation weight factor, realizing a more accurate acquisition of the situation awareness abnormal degree, and further realizing an improvement in the real-time performance of the situation awareness risk warning, effectively solving the problem of low real-time performance of the situation awareness abnormal data risk warning in the prior art.

[0065] Optionally, before inputting the obtained correlation index and situation awareness data into the risk warning model, it further includes preprocessing the obtained situation awareness data. The preprocessing includes data deduplication, data normalization, and data smoothing. Data deduplication is used to remove duplicate situation awareness data records. Data normalization is used to perform timestamp partitioning on the time series data in the situation awareness data. Timestamp partitioning is used to ensure that the situation awareness data is within the same timestamp. Data smoothing is used to eliminate the noise data and short-term fluctuations in the situation awareness data.

[0066] In this embodiment, a database query statement (such as the DISTINCT keyword in SQL) is used to identify and delete duplicate records. For time series data, it may be necessary to consider the combination of timestamp and specific values to determine whether it is a duplicate record; the respective feature values in the situation awareness data are scaled to the same dimension to eliminate the influence of the dimension difference between different features on model training. For example, the data is converted into a distribution with a mean of 0 and a standard deviation of 1. For data collected by different devices or at different time frequencies, interpolation or resampling operations are performed so that all data points are on the same timestamp; the low-pass filter is a commonly used technique in signal processing. It can remove high-frequency components (i.e., short-term fluctuations) while retaining low-frequency components (i.e., long-term trends). In the processing of situation awareness data, the time series data can be regarded as a signal, and a low-pass filter is applied to smooth the data, realizing an improvement in the accuracy of low-pass filter data processing, and further realizing an improvement in the accuracy and reliability of situation awareness data acquisition.

[0067] Optionally, the obtained correlation index and situation awareness data are input into a risk early warning model, and then obtaining a risk early warning index through the risk early warning model. The specific steps for obtaining the risk early warning index include: obtaining a risk probability value based on the situation awareness abnormal data and the situation awareness data, and at the same time combining the obtained correlation index and the degree of situation awareness abnormality to obtain the risk early warning index. The risk probability value is used to measure the proportion of the situation awareness abnormal data in the situation awareness data.

[0068] In this embodiment, in the patient visit application scenario, the risk probability value is usually obtained by the ratio of the abnormal blood pressure of the patient to the blood pressure within a preset time period; in addition to the risk probability value, it is also necessary to evaluate the severity of the situation awareness abnormality, which can be achieved by comparing the deviation degree of the abnormal data from the normal range, the duration of the abnormal data, and the frequency of the abnormal data. The risk probability value, the correlation index, and the degree of situation awareness abnormality are subjected to fuzzy comprehensive evaluation to generate the risk early warning index, and at the same time trigger the early warning mechanism to remind medical staff to pay attention to the health status of the patient, improving the accuracy of obtaining the risk early warning index.

[0069] Optionally, the expression for obtaining the risk early warning index through the risk early warning model is:

[0070]

[0071] In the formula, q is the number of the preset collection area, q = 1, 2,..., Q, Q is the total number of the preset collection areas, e is the natural constant, GUAN h represents the correlation index of the hth patient to be visited at the same time stamp, GUAN 0 represents the reference correlation index, YI h represents the degree of situation awareness abnormality of the situation awareness abnormal data of the hth patient to be visited within the preset time period, YI 0 represents the reference situation awareness abnormal degree, G h.q represents the risk probability value of the hth patient to be visited in the qth preset collection area, G 0 represents the reference risk probability value.

[0072] In this embodiment, it should be noted that in the calculation process of the risk early warning index, when GUAN h ≤GUAN 0 , G h.q ≥G 0 and YI h ≥YI 0 where GUAN 0 is the reference correlation index (i.e., the maximum value of the correlation index), G 0 is the reference risk probability value (i.e., the minimum value of the risk probability), YI 0is the abnormal degree of situation awareness (i.e., the minimum value of the abnormal degree of situation awareness). When GUAN h >GUAN 0 or G h.q <G 0 or YI h <YI 0 , then there is no need to calculate the risk warning index, and there is no need to conduct risk warning at this time.

[0073] To simplify the analysis, define In the formula, GU1 represents the correlation coefficient of the situation awareness abnormal data of the h-th patient to be visited within the preset time period, G1 represents the risk probability coefficient of the situation awareness abnormal data of the h-th patient to be visited within the preset time period, and Y1 represents the situation awareness abnormal coefficient of the situation awareness abnormal data of the h-th patient to be visited within the preset time period. The expression of the simplified risk warning index is: The change statistical table of the risk warning index is shown in Table 1:

[0074] Table 1 Change Statistical Table of Risk Warning Index

[0075]

[0076] It should be understood that the reference correlation index is usually represented by the result of summing and averaging the situation awareness correlation data in the preset database, the reference risk probability value is usually represented by the result of summing and averaging the risk probability data in the preset database, and the reference abnormal degree of situation awareness is usually represented by the result of summing and averaging the abnormal data in the preset situation awareness database. The risk warning index decreases as the correlation index increases, and increases as the abnormal degree of situation awareness and the risk probability value increase. When GUAN h =GUAN 0 , G h.q =G 0 and YI h =YI 0 , the situation awareness risk degree of the patient to be visited at the current time point is the smallest at this time, improving the accuracy and real-time performance of obtaining the risk warning index, and then improving the real-time performance of situation awareness risk warning, effectively solving the problem of low real-time performance of situation awareness abnormal data risk warning in the prior art.

[0077] Optionally, the risk level is determined by mapping the obtained risk warning index to a preset threshold interval; the threshold interval includes a low value interval, a medium value interval, and a high value interval, and the preset threshold interval is set by the risk level division standard; the risk level division standard is represented by the result of summing and averaging the risk level data in the preset database.

[0078] In this embodiment, historical risk level data is collected from a preset database. These data usually come from multiple cases, patients, or time periods to ensure their representativeness and comprehensiveness. The collected data is cleaned to remove outliers and missing values to ensure the accuracy and integrity of the data. Then, the cleaned data is summed and the average value is calculated. This average value can be used as a reference point or reference line for risk level division (i.e., the threshold range corresponding to the risk level). Among them, the low-value range indicates a lower risk. The upper limit of this range should be lower than the average value of the risk level division standard, and the specific value can be determined according to the actual situation and requirements; the median range is between the low-value range and the high-value range, indicating a medium risk. The upper and lower limits of this range can be set according to the average value of the risk level division standard and the data distribution; the high-value range indicates a higher risk. The lower limit of this range should be higher than the average value of the risk level division standard, and the specific value also needs to be determined according to the actual situation and requirements; the calculated risk warning indicator is compared with the set threshold range to determine which risk level it belongs to. If the risk warning indicator falls within the low-value range, it is determined as a low risk level; if the risk warning indicator falls within the median range, it is determined as a medium risk level; if the risk warning indicator falls within the high-value range or exceeds the upper limit of the high-value range, it is determined as a high risk level; it helps to provide timely and accurate risk assessment and warning information for medical staff, and realizes a more accurate setting of the risk level corresponding to the risk warning indicator.

[0079] Optionally, after generating the corresponding risk level according to the risk warning indicator, it further includes generating a monitoring and warning rainbow chart. The specific generation steps include: mapping the real-time situation awareness data and the corresponding risk level to the basic framework of the monitoring and warning rainbow chart, and at the same time using pre-set colors and color scales for rendering to generate the monitoring and warning rainbow chart. The basic framework includes coordinate axes, legends, and titles. The monitoring and warning rainbow chart shows the real-time changes of the risk level corresponding to the situation awareness data through a graph containing pre-set colors and color scales.

[0080] In this embodiment, the basic framework of the monitoring and warning rainbow chart usually includes coordinate axes, legends, and titles. Among them, the coordinate axes include a horizontal coordinate axis and a vertical coordinate axis. The horizontal coordinate axis can represent time (such as hours, minutes) to show the real-time changes of the situation awareness data; the vertical coordinate axis represents the specific value of the situation awareness or the processed risk warning indicator value; the legend usually clearly marks the risk levels corresponding to different colors or color scales (such as green representing low risk, yellow representing medium risk, and red representing high risk); the theme of the chart is marked, such as "Situation Awareness Risk Monitoring and Warning Rainbow Chart"; according to the calculated risk level, pre-set colors and color scales are selected to render the situation awareness data.

[0081] Plot the rendered data points (or data lines if it is time-series data) at the corresponding positions on the monitoring and warning rainbow chart, ensuring that the chart can be updated in real time to reflect the latest changes in the situation awareness data and risk levels; transmit the generated monitoring and warning rainbow chart to the interactive interfaces of medical staff (such as computer screens, tablets or smartphones) through the network or other communication methods, which helps medical staff quickly understand the situation awareness risk level of patients, realizes a more in-depth and comprehensive analysis of the situation awareness risk level of patients by medical staff, and further improves the accuracy and real-time performance of the generation of the monitoring and warning rainbow chart.

[0082] The following points need to be explained:

[0083] (1) The attached drawings of the embodiments of the present invention only relate to the structures involved in the embodiments of the present invention, and other structures can refer to the general design.

[0084] (2) For clarity, in the attached drawings used to describe the embodiments of the present invention, the thickness of layers or regions is enlarged or reduced, that is, these drawings are not drawn according to the actual scale. It can be understood that when an element such as a layer, film, region or substrate is referred to as being "on" or "under" another element, the element can be "directly" on or under the other element or there can be intermediate elements.

[0085] (3) Without conflict, the embodiments of the present invention and the features in the embodiments can be combined with each other to obtain new embodiments.

[0086] The above is only the specific implementation manner of the present invention, but the protection scope of the present invention is not limited thereto. The protection scope of the present invention shall be subject to the protection scope of the claims.

Claims

1. A dynamic risk warning system based on intelligent situational awareness, characterized in that: include: Data acquisition module, real-time monitoring module, situation awareness module and risk warning module; Wherein, the data acquisition module is used to collect the perception data of the patient to be visited in a preset collection area in real time, and the perception data includes facial image data and physiological parameter data, the facial image data is used to reflect the facial features of the patient to be visited in real time, the facial features are used to reflect the degree of influence of facial situation perception of the patient to be visited, the physiological parameter data is used to reflect the changes in the movement and behavior of the patient to be visited in real time, and the degree of influence of facial situation perception is used to measure the degree of influence of facial features on facial situation perception of the patient to be visited within a preset time period; The real-time monitoring module is used to monitor in real time the changes in the perception data of the patient to be visited within a preset time period to obtain a correlation index, and the correlation index is used to measure the correlation between the perception data of the patient to be visited within the preset time period and the action behavior of the patient to be visited; The situation awareness module is used to monitor the changes in the situation awareness of the patient to be visited within a preset time period in real time according to the acquired correlation indicators to obtain situation awareness data, and the situation awareness data is used to reflect the situation awareness state of the patient to be visited at the current time point in real time, and the situation awareness data includes heart rate, blood pressure and blood oxygen saturation; The risk warning module is used to input the acquired correlation index and situation awareness data into the risk warning model to obtain the risk warning index, and generate the corresponding risk level according to the risk warning index. The risk warning index is used to measure the situation awareness risk level of the patient to be visited at the current time point; The specific steps for obtaining the degree of influence of facial situation awareness are as follows: Detecting the skin color area in the face image of the patient to be visited within the preset acquisition area, and comparing the detected skin color area with the standard skin color sample in the preset database to obtain a skin color difference value; The facial situation awareness impact degree is calculated by the following formula: ； In the formula, h is the number of the patient to be visited, , H is the total number of patients to be visited, q is the number of the preset collection area, , Q is the total number of preset collection areas, e is a natural constant, It represents the impact degree of facial situation awareness of the hth patient to be visited in the preset acquisition area, represents the skin color difference value of the hth patient to be visited in the qth preset acquisition area, Indicates the reference skin color difference value, represents the contrast correction factor of the skin color area, represents the skin color area contrast of the hth patient's face image to be visited within the qth preset acquisition area, Indicates the skin color area saturation correction factor, represents the saturation of the skin color area in the hth patient's face image to be visited within the qth preset acquisition area; The correlation index is obtained by the following method: aligning the sensing data within a preset time period at the same timestamp to obtain an influence rate, wherein the influence rate includes a first influence rate and a second influence rate; The obtained facial situation awareness influence degree is combined to obtain a correlation index, which is calculated using the following formula: ； Where t is the number of the same timestamp, , T is the total number of the same timestamp, represents the correlation index of the hth patient to be visited within the same timestamp, represents the first impact rate of the face image data of the hth patient to be visited at the same time stamp of tth time, represents the reference first impact rate, represents the second impact rate of the physiological parameter data of the hth patient to be accessed at the same time stamp of tth time, represents the reference second impact rate, represents the situation awareness risk influencing factor, Indicates the reference influence degree of facial situation awareness.

2. The dynamic risk warning system based on intelligent situational awareness according to claim 1 is characterized in that: The situation awareness module includes a situation awareness data acquisition unit, a data preprocessing unit and a situation awareness anomaly recognition unit; The situation awareness data acquisition unit is used to monitor the situation awareness changes of the patient to be visited in real time within a preset time period and record the situation awareness data of the patient to be visited; The data preprocessing unit is used to convert the format of the situation awareness data; The situation awareness anomaly identification unit is used to determine whether the situation awareness of the patient to be visited is equal to a preset situation awareness value by real-time monitoring of the fluctuation of the situation awareness data within a preset time period. If not, it is recorded as situation awareness anomaly data and fed back to the preset personnel, otherwise it is stored in the preset database.

3. The dynamic risk warning system based on intelligent situational awareness according to claim 2 is characterized in that: The situation awareness abnormality data is recorded and fed back to the preset personnel, and then the situation awareness abnormality data is classified according to the acquired situation awareness abnormality degree. The specific steps for acquiring the situation awareness abnormality degree are: Obtaining an abnormal deviation degree according to the degree of deviation between the situation awareness abnormal data and the situation awareness data; Obtain the abnormal duration based on the time difference between the abnormal start time and the abnormal end time of the situation awareness abnormal data; The abnormal degree of situation awareness is obtained by combining the acquired abnormal deviation, abnormal duration and the action behavior deviation of the patient to be visited within a preset time period.

4. The dynamic risk warning system based on intelligent situational awareness according to claim 1 is characterized in that: The step of inputting the acquired correlation index and situation awareness data into the risk warning model also includes preprocessing the acquired situation awareness data, wherein the preprocessing includes data deduplication, data normalization and data smoothing.

5. The dynamic risk warning system based on intelligent situational awareness according to claim 3 is characterized by: The step of inputting the acquired correlation index and situation awareness data into the risk warning model further includes obtaining the risk warning index through the risk warning model. The specific steps of obtaining the risk warning index include: The risk probability value is obtained based on the situation awareness abnormal data and the situation awareness data, and the risk warning indicator is obtained by combining the obtained correlation index and the situation awareness abnormality degree.

6. The dynamic risk warning system based on intelligent situational awareness according to claim 5 is characterized in that: The expression of the risk warning index obtained by the risk warning model is: ； In the formula, q is the number of the preset collection area, , Q is the total number of preset collection areas, e is a natural constant, represents the correlation index of the hth patient to be visited within the same timestamp, represents the reference correlation index, represents the abnormal situation awareness level of the situation awareness abnormality data of the hth patient to be visited within the preset time period, Indicates the abnormality level of situation awareness reference, represents the risk probability value of the hth patient to be visited in the qth preset collection area, Represents the reference risk probability value.

7. The dynamic risk warning system based on intelligent situational awareness according to claim 1 is characterized in that: The risk level is determined by mapping the obtained risk warning indicator with a preset threshold range; The preset threshold interval is set by a risk level classification standard, and the risk level classification standard is represented by a result of summing and averaging the risk level data in a preset database.

8. The dynamic risk warning system based on intelligent situational awareness according to claim 1 is characterized in that: The corresponding risk level is generated according to the risk warning indicator, and then the monitoring and warning rainbow map is generated. The specific generation steps include: The real-time situational awareness data and the corresponding risk levels are mapped to the basic framework of the monitoring and early warning rainbow map, and are rendered using preset colors and color levels to generate a monitoring and early warning rainbow map. The monitoring and early warning rainbow map represents the real-time changes in the risk levels corresponding to the situational awareness data through a graphic visualization containing preset colors and color levels.

Images