Information digital management system and method based on intensive care medicine department

Abstract

The invention relates to the technical field of medical informatization, and provides an information digital management system and method based on an intensive care medicine department, and the system comprises a data collection module, a data processing module, and a visualization and alarm module. The data acquisition module obtains a patient monitoring data set, and each patient corresponds to a monitoring data sequence. The data processing module executes the following operations on each patient: determining a target monitoring data sequence and four associated monitoring periods including a first monitoring period, a second monitoring period, a third monitoring period and a fourth monitoring period group, and generating a corresponding monitoring subset; generating an early warning signal through a disease deterioration early warning model; and if the signal is abnormal, analyzing an abnormal reason and an influence range, and generating a structured disease condition comprehensive report. And the visualization and alarm module dynamically displays the state of the patient according to the report, and performs graded alarm on the critical area. The system realizes accurate early warning and monitoring through multi-cycle data fusion analysis and three-dimensional visualization.

Description

Technical Field

[0001] The present invention relates to the field of medical informatization technology. More specifically, the present invention relates to an information digital management system and method based on the intensive care unit (ICU). Background Art

[0002] In the modern medical system, the intensive care unit (ICU) undertakes the important task of centralized management and treatment of critically ill patients. With the continuous development of medical technology, the patient monitoring devices in the ICU are becoming increasingly advanced, capable of real-time collection of a large amount of vital sign data, such as heart rate, blood oxygen, blood pressure, etc. These data are crucial for doctors to evaluate the patient's condition and formulate treatment plans. However, the current information management in the ICU mostly relies on traditional electronic medical record systems and manual records, with low levels of data integration, analysis, and visualization. On the one hand, medical staff need to spend a lot of time and energy on data collation and search, reducing work efficiency; on the other hand, it is difficult to monitor and warn of the patient's condition changes in real-time and accurately, and it is easy to miss the best treatment opportunity. In addition, the existing systems also have deficiencies in disease analysis and report generation, lacking structured and multi-dimensional displays, which is not conducive to information sharing and collaboration among the medical team.

[0003] There are at least the following problems or defects in the prior art: First, the traditional information management system cannot effectively integrate multi-source heterogeneous patient monitoring data, resulting in scattered data and difficult to manage uniformly; second, there is a lack of a real-time warning mechanism for changes in the patient's condition and it is impossible to detect potential risks of condition deterioration in a timely manner; third, the methods of disease analysis and report generation are relatively single and cannot meet the complex and changeable clinical needs; finally, the existing visualization display effects are limited and cannot intuitively present the real-time status of patients and the intensive care area, affecting the efficiency and accuracy of medical decision-making. Summary of the Invention

[0004] The present invention provides an information digital management system and method based on the intensive care unit (ICU).

[0005] In the first aspect of the present invention, an information digital management system based on the intensive care unit (ICU) is provided, including:

[0006] A data acquisition module for obtaining the patient monitoring data sets corresponding to each patient in the target intensive care unit (ICU), where each patient has a corresponding patient monitoring data sequence;

[0007] A data processing module for performing the following operations on each patient among all patients:

[0008] Determining the patient monitoring data sequence corresponding to the patient as the target monitoring data sequence;

[0009] Determine four monitoring cycle groups corresponding to the patient for the current monitoring cycle;

[0010] Generate a monitoring subset group corresponding to the monitoring cycle group according to the target monitoring data sequence;

[0011] Generate a deterioration warning signal for the patient according to the monitoring subset group by using a pre-trained deterioration warning model;

[0012] In response to the deterioration warning signal indicating that the patient's condition is abnormal, determine the abnormal cause analysis result and the impact range evaluation result corresponding to the patient;

[0013] Generate a structured comprehensive condition report in a predetermined format according to the abnormal cause analysis result, the target monitoring data sequence, and the impact range evaluation result;

[0014] A visualization and alarm module for performing dynamic visualization display for each patient according to the comprehensive condition report and performing hierarchical alarm processing on associated intensive care areas.

[0015] As a further improvement of this application, the four monitoring cycle groups include a first monitoring cycle, a second monitoring cycle, a third monitoring cycle, and a fourth monitoring cycle; the first monitoring cycle includes the second monitoring cycle, the time range corresponding to the first monitoring cycle is greater than that of the second monitoring cycle, the fourth monitoring cycle includes the third monitoring cycle, and the central monitoring cycle corresponding to the third monitoring cycle is in the same cycle as the current monitoring cycle.

[0016] As a further improvement of this application, the dynamic visualization display includes:

[0017] Determine the condition impact level of each sub-region in the target intensive care unit according to each comprehensive condition report;

[0018] Based on the condition impact level and the comprehensive condition report, assign a visualization strategy to each patient and sub-region, and the visualization strategy includes transparency level mapping and color coding mapping;

[0019] Render the abnormal status of the patient and the sub-region in real time in the virtual ward three-dimensional model according to the visualization strategy.

[0020] As a further improvement of this application, the steps for determining the monitoring cycle configuration corresponding to the patient include:

[0021] Match the initial monitoring cycle configuration from a preset association table according to the patient's clinical classification and the sub-region where the patient is located;

[0022] Obtain the historical data of the patient's vital sign fluctuations and calculate its periodic characteristics;

[0023] Based on the periodic characteristics, dynamically adjust the initial monitoring cycle configuration to generate the first adjusted cycle, the second adjusted cycle, the third adjusted cycle, and the fourth adjusted cycle that are adapted;

[0024] Combined with the current monitoring cycle, generate the final first monitoring cycle, the second monitoring cycle, the third monitoring cycle, and the fourth monitoring cycle.

[0025] As a further improvement of this application, the steps of generating a comprehensive disease report include:

[0026] Generate a multi-dimensional trend graph corresponding to the target monitoring data sequence, including the time series changes of heart rate, blood oxygen, and blood pressure;

[0027] Obtain the timestamps, intervention measures, and associated video clips of the patient's abnormal treatment events;

[0028] Generate an encrypted access link for the video clip;

[0029] Integrate the abnormal cause analysis results, the multi-dimensional trend graph, the impact range assessment results, the abnormal treatment events, and the encrypted link into a structured comprehensive disease report according to the clinical report template.

[0030] As a further improvement of this application, it also includes:

[0031] In response to the interaction operation of the target patient in the virtual ward model triggered by the user, pop up an information category selection interface including vital signs, treatment records, and video playback;

[0032] Extract the corresponding data from the comprehensive disease report according to the category selected by the user;

[0033] Overlay and display the data at the bed coordinate position of the target patient in the virtual ward model.

[0034] As a further improvement of this application, the operation logic of the disease deterioration warning model includes:

[0035] The monitoring subset group includes the first monitoring subset, the second monitoring subset, the third monitoring subset, and the fourth monitoring subset;

[0036] Extract the data part in the first monitoring subset that does not overlap with the second monitoring subset as an independent feature analysis set;

[0037] Analyze the independent feature analysis set through the first feature extraction module to generate a first feature vector reflecting long-term cycle fluctuations;

[0038] Fuse the first feature vector and the original data of the first monitoring subset through the second feature extraction module to generate a second feature vector;

[0039] Input the second monitoring subset into the second feature extraction module to generate a third feature vector;

[0040] Perform horizontal fusion on the first feature vector and the second feature vector to generate a first fused feature;

[0041] Perform vertical fusion on the first feature vector and the third feature vector to generate a second fused feature;

[0042] Input the first fused feature and the second fused feature into the temporal attention module to predict the clinical index parameters for the target's future monitoring period;

[0043] Perform a similarity comparison between the predicted parameters and the historical data of the same period in the third monitoring subset. If the difference exceeds the threshold, trigger an early warning;

[0044] Combine the cross-period consistency verification results of the fourth monitoring subset to generate a weighted early warning signal for disease deterioration, where the temporal attention module uses a deep learning algorithm based on the attention mechanism.

[0045] As a further improvement of this application, the processing logic when the similarity comparison does not meet the threshold includes:

[0046] Screen a historical data subset from the third monitoring subset that matches the target's future cycle date;

[0047] Extract the historical data of the same period of all associated subsets from the fourth monitoring subset;

[0048] Compare the statistical distribution characteristics of the predicted parameters and the historical data. If the distribution deviation exceeds the preset range, trigger an artificial review process;

[0049] Update the model weights according to the review results and regenerate the early warning signal.

[0050] In the second aspect of the present invention, an information digital management method based on the intensive care unit is provided, which is characterized by including:

[0051] Obtain the patient monitoring data sets corresponding to each patient in the target intensive care unit, where each patient has a corresponding patient monitoring data sequence;

[0052] For each patient among all patients, perform the following generation steps:

[0053] Determine the patient monitoring data sequence corresponding to the patient as the target monitoring data sequence;

[0054] Determine the first monitoring cycle, the second monitoring cycle, the third monitoring cycle, and the fourth monitoring cycle corresponding to the patient for the current monitoring period;

[0055] Generate a first monitoring subset corresponding to a first monitoring period, a second monitoring subset corresponding to a second monitoring period, a third monitoring subset corresponding to a third monitoring period, and a fourth monitoring subset corresponding to a fourth monitoring period according to the target monitoring data sequence;

[0056] Generate a deterioration warning signal for the patient by using a pre-trained deterioration warning model according to the first monitoring subset, the second monitoring subset, the third monitoring subset, and the fourth monitoring subset;

[0057] In response to the deterioration warning signal indicating that the patient's condition is abnormal, determine the abnormal cause analysis result and the impact scope evaluation result corresponding to the patient;

[0058] Generate a structured comprehensive condition report in a predetermined format according to the abnormal cause analysis result, the target monitoring data sequence, and the impact scope evaluation result;

[0059] Execute dynamic visualization display for each patient according to the obtained comprehensive condition reports, and perform hierarchical warning processing on associated critical areas.

[0060] As a further improvement of this application, the first monitoring period includes the second monitoring period, the time range corresponding to the first monitoring period is greater than that of the second monitoring period, the fourth monitoring period includes the third monitoring period, and the central monitoring period corresponding to the third monitoring period is in the same period as the current monitoring period.

[0061] According to the above embodiments of the present invention, it has at least the following beneficial effects: The present invention can realize precise monitoring and warning of the patient's condition through the synergistic effect of the first monitoring period, the second monitoring period, the third monitoring period, and the fourth monitoring period based on the multi-period data fusion analysis technology. Through the temporal attention mechanism and feature fusion algorithm of the deterioration warning model, the prediction accuracy of the deterioration of the condition can be improved. At the same time, combined with the abnormal cause analysis and the impact scope evaluation, it can provide comprehensive and reliable data support for clinical decision-making.

[0062] The present invention can also intuitively present the change trend of the patient's condition and the risk level of the critical area by means of dynamic visualization display and hierarchical warning mechanism. Through the real-time rendering of the virtual ward three-dimensional model and the interactive data display, the monitoring efficiency of medical staff can be optimized. At the same time, based on the structured report of encrypted video clips and multi-dimensional trend maps, the response speed of the medical team to critical situations can be improved, and finally the overall management level and treatment effect of the intensive care unit can be improved. Description of the Drawings

[0063] By reading the following detailed description with reference to the accompanying drawings, the above and other objects, features, and advantages of the exemplary embodiments of the present invention will become readily understandable. In the drawings, several embodiments of the present invention are shown by way of example and not limitation, wherein:

[0064] Figure 1 FIG. 4 is a schematic structural diagram of an information digital management system based on the intensive care unit according to an embodiment of the present invention;

[0065] Figure 2 FIG. 5 is a schematic flowchart of an information digital management method based on the intensive care unit according to an embodiment of the present invention;

[0066] Figure 3 FIG. 6 is a schematic structural diagram of an electronic device according to an embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0067] The principles and spirit of the present invention will now be described with reference to several exemplary embodiments. It should be understood that these embodiments are provided only to enable those skilled in the art to better understand and then implement the present invention, and not to limit the scope of the present invention in any way. On the contrary, these embodiments are provided to make the present invention more thorough and complete, and to convey the scope of the present invention fully to those skilled in the art.

[0068] Those skilled in the art know that the embodiments of the present invention can be implemented as a system, apparatus, device, method, or computer program product. Therefore, the present invention can be specifically implemented in the following forms: complete hardware, complete software (including firmware, resident software, microcode, etc.), or a combination of hardware and software.

[0069] It should be noted that any number of elements in the drawings is for illustration rather than limitation, and any naming is only for distinction and does not have any limiting meaning.

[0070] The following references Figure 1 , Figure 1 FIG. 4 is a schematic structural diagram of an information digital management system based on the intensive care unit according to an embodiment of the present invention. As Figure 1 shown, an information digital management system based on the intensive care unit includes:

[0071] A data acquisition module 101, configured to obtain a patient monitoring data set corresponding to each patient in the target intensive care unit, wherein each patient has a corresponding patient monitoring data sequence;

[0072] A data processing module 102, configured to perform the following generation steps for each patient among all patients:

[0073] Determine the patient monitoring data sequence corresponding to the patient as the target monitoring data sequence;

[0074] Determine the first monitoring period, second monitoring period, third monitoring period, and fourth monitoring period for the current monitoring period corresponding to the patient, where the first monitoring period includes the second monitoring period, the time range corresponding to the first monitoring period is greater than that of the second monitoring period, the fourth monitoring period includes the third monitoring period, and the central monitoring period corresponding to the third monitoring period is within the same week as the current monitoring period;

[0075] Generate a first monitoring subset corresponding to the first monitoring period, a second monitoring subset corresponding to the second monitoring period, a third monitoring subset corresponding to the third monitoring period, and a fourth monitoring subset corresponding to the fourth monitoring period according to the target monitoring data sequence;

[0076] Generate a deterioration warning signal for the patient by using the pre-trained deterioration warning model based on the first monitoring subset, second monitoring subset, third monitoring subset, and fourth monitoring subset;

[0077] In response to the deterioration warning signal indicating that the patient's condition is abnormal, determine the abnormal cause analysis result and the impact range evaluation result corresponding to the patient;

[0078] Generate a structured comprehensive condition report in a predetermined format according to the abnormal cause analysis result, the target monitoring data sequence, and the impact range evaluation result;

[0079] The visualization and alarm module 103 is used to perform dynamic visualization display for each patient based on the obtained comprehensive condition reports for each patient, and perform hierarchical alarm processing on the associated critical areas.

[0080] It should be noted that an information digital management system based on the intensive care unit proposed by the present invention aims to achieve precise monitoring and management of intensive care unit patients through functional modules such as data collection, processing, visualization, and alarm. Among them, the data collection module refers to the component used to obtain the vital signs and related monitoring data of each patient in the intensive care unit. These data include, but are not limited to, continuous monitoring signals such as heart rate, blood oxygen, and blood pressure, as well as the treatment records and imaging data of the patient. These data are organized into a patient monitoring data set, and each patient corresponds to a data sequence for subsequent analysis and processing. The data processing module is the core part of analyzing and processing the collected data. By processing the patient monitoring data sequence, subsets for different monitoring periods are generated, and a pre-trained deterioration warning model is used to predict the changes in the patient's condition, so as to achieve real-time monitoring and warning of the patient's condition.

[0081] Specifically, the first monitoring period, the second monitoring period, the third monitoring period, and the fourth monitoring period are monitoring time segments divided according to different time ranges and monitoring requirements. The first monitoring period is usually a relatively long time range, such as one week or one month, for observing the overall trend of the patient's condition changes; the second monitoring period is a shorter time segment included within the first monitoring period, such as one day or several days, for capturing short-term fluctuations in the patient's condition; the third monitoring period is a specific time segment centered around the current monitoring period, usually a specific time point within one week, for analyzing the patient's condition changes near that time point; the fourth monitoring period includes multiple third monitoring periods for cross-period comparative analysis. The division of these monitoring periods can be dynamically adjusted according to the patient's clinical characteristics and the severity of the condition to better meet the needs of different patients. For example, for patients with relatively stable conditions, the monitoring period can be appropriately extended; while for patients with rapidly changing conditions, the monitoring period needs to be shortened to detect condition changes more promptly.

[0082] Exemplarily, for the cross-period monitoring requirements of critically ill patients, the system adopts a master-slave time window design to construct the fourth monitoring period. The main period is set to 72 hours, internally divided into 9 overlapping 8-hour sub-periods, with a 2-hour overlap area set between adjacent sub-periods. Data collection uses a sliding window mechanism, with a sampling interval of one minute, and seamless splicing of time series data is achieved through a circular buffer. For example, in the scenario of monitoring sepsis patients, when the central venous pressure fluctuates by more than ±5 mmHg, the intensive sampling mode of the fourth monitoring period is automatically activated, shortening the sub-period to 4 hours and starting video monitoring for synchronous marking.

[0083] Preferably, in the process of generating a warning signal for deteriorating condition, a model construction method based on deep learning can be adopted. Specifically, first, features are extracted from a large amount of historical patient monitoring data, including the time-series change features of vital signs, the timestamp features of treatment events, etc., as the input parameters of the model. Then, the model is trained using the labeled samples of deteriorating condition so that the model can learn the feature patterns of the patient's deteriorating condition. In actual application, the real-time collected patient monitoring data is input into the trained model, and the model will output the probability or warning signal of the patient's deteriorating condition according to the features of the input data. In addition, during the data processing process, the collected data can also be preprocessed, such as filtering, denoising, normalization, etc., to improve the data quality and the accuracy of the model. For example, for heart rate data, high-frequency noise can be removed through a filtering algorithm to make it smoother for subsequent analysis and processing.

[0084] In some embodiments, the dynamic visual display includes:

[0085] Determine the disease impact level of each sub-region in the target intensive care unit according to the comprehensive disease reports for each case.

[0086] Based on the disease impact level and the comprehensive disease reports, allocate visualization strategies for each patient and sub-region. The visualization strategies include transparency level mapping and color-coding mapping.

[0087] According to the visualization strategies, render the abnormal states of patients and sub-regions in real time in the three-dimensional model of the virtual ward.

[0088] It should be noted that the dynamic visualization display function mentioned in the present invention is an important part of the system. It visualizes the key information in the comprehensive disease reports, enabling medical staff to more intuitively understand the disease distribution and changes of patients in the intensive care unit. Among them, the disease impact level refers to the level divided according to the severity of the patient's condition, the trend of the disease change, and the potential impact on the surrounding environment (such as other patients in the same ward), usually divided into different levels such as mild, moderate, and severe. The visualization strategies include transparency level mapping and color-coding mapping, which are used to display the abnormal states of patients and sub-regions in the three-dimensional model of the virtual ward with different visual effects, so that medical staff can quickly identify the problems.

[0089] Specifically, the three-dimensional model of the virtual ward is a digital virtual environment that simulates the actual layout and structure of the intensive care unit ward, including hospital beds, medical equipment, ward areas, etc. In this model, the position and status of each patient can be displayed through the visualization strategies. Transparency level mapping means adjusting the transparency of the patient or sub-region in the virtual model according to the disease impact level. The more severe the condition, the lower the transparency, thus highlighting the areas with more severe conditions. Color-coding mapping uses different colors to represent different disease states or impact levels. For example, red represents severe abnormality, yellow represents moderate abnormality, and green represents normal state. The settings of these parameters can be adjusted according to actual needs to ensure that the visualization effect can clearly convey the key information.

[0090] The definition of the disease impact level can be comprehensively evaluated through the SOFA scoring system in combination with vital sign parameters. Exemplarily, the SOFA scoring system scores the organ dysfunction of patients from six aspects: consciousness status, circulatory system, respiratory system, liver function, coagulation function, and renal function. Each aspect has a full score of 4 points, and the total score is 24 points. The higher the score, the more severe the condition. In addition, we also consider the degree of abnormality of vital sign parameters (such as heart rate, blood pressure, blood oxygen saturation, etc.) and combine it with the SOFA score. For example, if the patient's heart rate continuously exceeds 120 beats per minute or is lower than 40 beats per minute, 2 additional points will be added; if the blood pressure continuously drops below 90 / 60 mmHg, 3 additional points will be added. Through this comprehensive scoring method, the disease impact level is divided into four levels: mild (0 - 4 points), moderate (5 - 9 points), severe (10 - 14 points), and extremely severe (15 points and above), providing a basis for subsequent dynamic visualization display and grading alarm.

[0091] Preferably, in the process of implementing the dynamic visualization display, the system can further optimize the data processing and display methods. For example, when generating a comprehensive disease report, the system can automatically extract key data, such as the abnormal values of the patient's vital signs, the time points of treatment events, etc., and associate them with the corresponding positions in the three-dimensional model of the virtual ward. When medical staff view a certain patient or sub-region in the virtual model, the system can render and display the relevant abnormal states in real time, including changes in transparency and color. In addition, the system can also provide an interactive function, allowing medical staff to select and view the detailed information of specific patients or sub-regions, such as the real-time trend chart of vital signs, treatment records, etc., through mouse clicks or gesture operations, thereby further enhancing the readability and practicality of the information.

[0092] For the selection of the graphics engine for the three-dimensional model rendering strategy, the Unity3D engine can be selected. Through the ShaderLab shader programming language of Unity3D, a custom shader program is written to achieve the real-time rendering of the three-dimensional model of the virtual ward. For example, in order to render the abnormal states of patients and their sub-regions, strategies of transparency level mapping and color coding mapping can be adopted. Different transparency values and color values are assigned to each patient and sub-region according to the disease impact level. In the shader program, by controlling the transparency and color attributes of pixels, the dynamic rendering of the model is achieved. At the same time, the Unity3D engine also supports functions such as scene management, lighting calculation, and texture mapping, which can ensure the rendering performance and effect in complex scenarios (such as large intensive care departments), providing a smooth visualization experience for medical staff.

[0093] In some embodiments, the steps for determining the monitoring cycle configuration corresponding to a patient include:

[0094] Match the initial monitoring cycle configuration from the preset association table according to the clinical classification and sub-region of the patient;

[0095] Obtain the historical data of the patient's vital sign fluctuations and calculate its periodic characteristics;

[0096] Based on the periodic characteristics, dynamically adjust the initial monitoring cycle configuration to generate the adapted first adjustment cycle, second adjustment cycle, third adjustment cycle and fourth adjustment cycle;

[0097] Combine the current monitoring cycle to generate the final first monitoring cycle, second monitoring cycle, third monitoring cycle and fourth monitoring cycle.

[0098] It should be noted that the steps of determining the monitoring cycle configuration corresponding to the patient in the present invention are to better adapt to the individual differences and disease conditions of different patients, so as to achieve precise monitoring. Among them, the clinical classification refers to the classification based on the patient's disease type, disease severity, etc., which is an important basis for determining the initial monitoring cycle configuration. The historical data of vital sign fluctuations refers to the changes in vital sign data such as heart rate, blood pressure, and blood oxygen recorded during the patient's previous monitoring. These data reflect the periodic laws of the patient's physiological characteristics and disease conditions. By analyzing the periodic characteristics of these data, the initial monitoring cycle configuration can be dynamically adjusted to generate a monitoring cycle group that better conforms to the patient's current condition.

[0099] Specifically, the initial monitoring cycle configuration is matched from the preset association table, which pre-sets different monitoring cycle templates according to the clinical classification and sub-region of the patient. For example, for patients with mild conditions, the initial monitoring cycle may be longer; while for critically ill patients, the initial monitoring cycle is shorter. The periodic characteristics refer to the regular changes of the patient's vital sign data in the time series, such as the circadian rhythm of the heart rate and the fluctuation cycle of the blood pressure. By calculating these characteristics, the initial monitoring cycle can be adjusted. For example, if the vital sign fluctuations of the patient have an obvious circadian rhythm, the monitoring cycle can be set to be dynamically adjusted in units of 24 hours. The finally generated first adjustment cycle, second adjustment cycle, third adjustment cycle and fourth adjustment cycle will be combined with the current monitoring cycle to form the final monitoring cycle group for subsequent disease monitoring and analysis.

[0100] Preferably, in the process of determining the monitoring cycle configuration, the following steps can be adopted for optimization. First, match the initial monitoring cycle configuration from the preset association table according to the patient's clinical classification and the sub-region where the patient is located. For example, for a patient with severe respiratory failure, the initial monitoring cycle may be set to monitor the blood oxygen saturation once per hour. Next, obtain the historical data of the patient's vital sign fluctuations over a past period of time, and calculate its periodic characteristics through a time series analysis algorithm. For example, it is found that the patient's blood oxygen saturation fluctuates greatly at night, with a cycle of about 8 hours. Then, dynamically adjust the initial monitoring cycle according to these periodic characteristics. For example, shorten the monitoring cycle at night to once every half hour to better capture the changes in the condition. Finally, combine the current monitoring cycle to generate the final monitoring cycle group, ensuring that the monitoring cycle can accurately reflect the law of the patient's condition changes, and providing more accurate data support for subsequent early warning of condition deterioration and generation of comprehensive reports.

[0101] When calculating the periodic characteristics of the historical data of vital sign fluctuations, time-frequency analysis methods such as Fourier transform and wavelet analysis can be adopted. The Fourier transform is used to convert the time-domain signal into a frequency-domain signal, so as to identify the periodic frequency components in the signal. For example, for heart rate data, its spectrogram can be obtained through the Fourier transform, showing the periodic frequencies of heart rate fluctuations, such as the circadian rhythm (about 1 time / 24 hours) and the respiration-related frequency (about 1 time / several seconds). Wavelet analysis is suitable for the analysis of non-stationary signals. By decomposing the signal at different scales, it can capture the periodic changes of the signal at different time scales. For example, for blood pressure data, wavelet analysis can reveal the periodic fluctuation characteristics of blood pressure in the short term (from several minutes to several hours) and the long term (from several days to several weeks). Through these time-frequency analysis methods, we can accurately calculate the periodic characteristics of vital sign data, providing a basis for the subsequent dynamic adjustment of the monitoring cycle.

[0102] In some embodiments, the steps of generating a comprehensive condition report include:

[0103] Generate a multi-dimensional trend map corresponding to the target monitoring data sequence, including the time series changes of heart rate, blood oxygen, and blood pressure;

[0104] Obtain the timestamps, intervention measures, and associated video clips of the patient's abnormal treatment events;

[0105] Generate an encrypted access link to the video clip;

[0106] Integrate the abnormal cause analysis results, multi-dimensional trend map, impact range assessment results, abnormal treatment events, and encrypted link into a structured comprehensive condition report according to the clinical report template.

[0107] It should be noted that the step of generating a comprehensive condition report in the present invention is to integrate information such as the patient's monitoring data, the analysis results of abnormal causes, and the evaluation results of the affected range to form a structured report, so that medical staff can quickly and comprehensively understand the patient's condition. Among them, the multi-dimensional trend graph refers to visualizing the patient's vital sign data (such as heart rate, blood oxygen, blood pressure, etc.) in the form of time series changes, so as to intuitively present the change trend of the patient's condition. The time stamps, intervention measures, and associated video clips of abnormal treatment events refer to the special situations and their handling measures that occur during the patient's treatment process. These information are of great reference value for analyzing the changes in the patient's condition. The encrypted access link is to protect the patient's privacy and data security, ensuring that only authorized personnel can access the relevant video clips.

[0108] Specifically, the multi-dimensional trend graph is generated by analyzing the target monitoring data sequence, extracting the time series data of vital signs such as heart rate, blood oxygen, and blood pressure, and plotting them in the same coordinate system to form a multi-dimensional trend graph. For example, the abscissa represents time, and the ordinate represents the values of heart rate, blood oxygen, and blood pressure respectively. By distinguishing with different colors or lines, the change trends of each vital sign can be clearly displayed. An abnormal treatment event refers to a special situation that occurs during the patient's treatment process, such as sudden arrhythmia, apnea, etc. Its time stamp records the specific time point when the event occurs, and the intervention measure is the treatment method taken by medical staff for this event, such as drug injection, defibrillation, etc. The associated video clip is the real-time video recorded when the event occurs, which is used for auxiliary analysis and review. Integrating these information into the structured comprehensive condition report can provide comprehensive condition reference for medical staff.

[0109] Preferably, in the process of generating the comprehensive condition report, the integration and display methods of information can be further optimized. For example, when generating the multi-dimensional trend graph, a data smoothing algorithm can be used to process the original monitoring data to reduce noise interference and make the trend graph clearer and more accurate. At the same time, for the record of abnormal treatment events, the types, severity, and treatment results of the events can be marked in detail, and key frames can be extracted from the associated video clips so that medical staff can quickly locate the critical moment when the event occurs. When integrating the report, according to the requirements of the clinical report template, information such as the analysis results of abnormal causes, the multi-dimensional trend graph, the evaluation results of the affected range, and abnormal treatment events can be reasonably arranged to form a comprehensive condition report with rich content and clear structure. In addition, an index and keyword search function can be added to the report to facilitate medical staff to quickly find the information they need.

[0110] The encrypted access link can adopt a temporary access key generation scheme based on JWT tokens. JWT (JSON Web Token) is an open standard (RFC 7519) that defines a compact and self - contained way to transmit security information as a JSON object between parties. Each encrypted access link contains a JWT token, which includes the following information: the unique identifier of the accessed video clip, the identity information of the visitor, the access validity period (e.g., expires after 1 hour), and the signature key (used to verify the authenticity of the token). The signature key is generated using the HMAC SHA256 algorithm to ensure the security of the token. For example, when medical staff needs to access the associated video clip of a certain abnormal treatment event, the system generates an encrypted access link with a JWT token. After the medical staff clicks the link, they can view the video clip within the validity period. After the validity period expires, the link will become invalid and the video content cannot be accessed, thus protecting the privacy and data security of patients.

[0111] In some embodiments, it further includes:

[0112] In response to an interaction operation on the target patient in the virtual ward model triggered by the user, a information category selection interface including vital signs, treatment records, and video playback is popped up;

[0113] According to the category selected by the user, corresponding data is extracted from the comprehensive condition report;

[0114] The data is superimposed and displayed at the bed coordinate position of the target patient in the virtual ward model.

[0115] It should be noted that the interaction operation function mentioned in the present invention is to enhance the interaction between medical staff and the system, enabling medical staff to more conveniently obtain key information of patients. When the user performs an interaction operation on the target patient in the virtual ward model, the system pops up an information category selection interface, which allows the user to select different categories of information such as vital signs, treatment records, and video playback. The realization of this function is based on the structured storage and fast retrieval capabilities of the system for the comprehensive condition report, ensuring that users can quickly obtain relevant information according to their needs. By superimposing and displaying the data at the bed coordinate position of the target patient in the virtual ward model, medical staff can quickly understand the detailed situation of the patient in an intuitive virtual environment, thereby improving work efficiency and the accuracy of decision - making.

[0116] Specifically, the interaction operation refers to the action of the user selecting or querying the target patient in the virtual ward model through mouse clicks, touch screen operations, or other input devices. When the user triggers the interaction operation, an information category selection interface will pop up, which provides multiple options for the user to choose from. For example, the vital signs option can display real-time data such as the patient's heart rate, blood pressure, and blood oxygen; the treatment record option can show the patient's medication records, surgical records, etc.; and the video playback option can play monitoring video clips related to the patient. These information categories are classified according to the data content stored in the comprehensive condition report to ensure that the user can quickly locate the required information. The bed coordinate position of the target patient refers to the specific position information of the patient's bed in the virtual ward model. The system realizes the visual display of information by overlaying the data on this position, enabling medical staff to intuitively view the patient information in the virtual environment.

[0117] Preferably, when implementing the interaction operation function, the user experience and the system's response speed can be further optimized. For example, the system can pre-index and classify the data in the comprehensive condition report in advance so that the required information can be quickly retrieved when the user triggers the interaction operation. For the display of vital signs data, the system can obtain the latest data from the monitoring device in real time and display it in the information category selection interface in the form of a chart, while providing a comparison function for historical data to help medical staff quickly understand the changing trend of the patient's vital signs. For the treatment record data, the system can sort it according to time sequence or event type to facilitate the user to quickly find specific treatment events. For the video playback function, the system can extract and annotate the key frames of the video clips, and the user can quickly locate the important moments in the video by clicking on the key frames. In addition, the system can provide a personalized interface layout and data display method according to the user's operation habits and preferences to further improve the usability and practicality of the system.

[0118] In some embodiments, the operation logic of the condition deterioration warning model includes:

[0119] Extract the data part in the first monitoring subset that does not overlap with the second monitoring subset as an independent feature analysis set;

[0120] Analyze the independent feature analysis set through the first feature extraction module to generate a first feature vector reflecting long-term fluctuations;

[0121] Fuse the first feature vector with the original data of the first monitoring subset through the second feature extraction module to generate a second feature vector;

[0122] Input the second monitoring subset into the second feature extraction module to generate a third feature vector;

[0123] Horizontally fuse the first feature vector and the second feature vector to generate a first fusion feature;

[0124] Vertically fuse the first eigenvector and the third eigenvector to generate a second fused feature;

[0125] Input the first fused feature and the second fused feature into the temporal attention module to predict the clinical index parameters of the target in the future monitoring period;

[0126] Compare the predicted parameters with the historical data of the same period in the third monitoring subset. If the difference exceeds the threshold, a warning is triggered;

[0127] Combine the cross-period consistency verification results of the fourth monitoring subset to generate a weighted warning signal for disease deterioration, where the temporal attention module uses a deep learning algorithm based on the attention mechanism.

[0128] It should be noted that the operation logic of the disease deterioration warning model mentioned in the present invention is to predict in advance the possible deterioration of the patient's condition through multi-dimensional analysis of the patient's monitoring data, so that medical staff can take intervention measures in time. Among them, the independent feature analysis set refers to the long-period data part extracted from the monitoring data that does not overlap with the short-term monitoring period, and these data are used to analyze the long-term change trend of the patient's condition. The first feature extraction module and the second feature extraction module are algorithm modules for extracting key features from the monitoring data, and these features can reflect the change law of the patient's condition. The temporal attention module is a deep learning-based algorithm that can analyze time series data and predict the clinical index parameters of the future monitoring period. By comparing the predicted parameters with the historical data, the system can judge whether there is an abnormal change in the patient's condition and generate a warning signal.

[0129] Specifically, the first monitoring subset refers to the patient monitoring data collected in the first monitoring period, which usually contains the disease change information of a relatively long period; the second monitoring subset refers to the data collected in the second monitoring period, which reflects the disease change in a relatively short period. The first eigenvector and the second eigenvector are features extracted from these monitoring subsets through the feature extraction module for subsequent fusion analysis. The third monitoring subset and the fourth monitoring subset are used to generate the third eigenvector and cross-period consistency verification respectively. In the feature fusion process, horizontal fusion refers to combining different eigenvectors in the same monitoring period to obtain a more comprehensive feature representation; vertical fusion refers to combining eigenvectors in different monitoring periods to analyze the long-term change trend of the disease. Finally, the temporal attention module analyzes the fused features, predicts the clinical index parameters of the future monitoring period, and compares them with the historical data for similarity to determine whether to trigger a warning.

[0130] The sequential attention module can be based on the Transformer architecture. The Transformer architecture captures the dependencies between different time points in sequential data through the self-attention mechanism. The input sequence of feature vectors is passed through the multi-head self-attention mechanism to calculate the attention weights in multiple different subspaces, thereby capturing richer feature information. At the same time, to better handle the long-term dependencies of sequential data, an LSTM (long short-term memory network) structure can be combined with the Transformer architecture to form an LSTM+Attention hybrid model. LSTM is used to initially process sequential data, extract short-term and long-term features of the time series, and then input these features into the Attention module to further emphasize key features and ignore unimportant features. For example, when processing electrocardiogram (ECG) data, LSTM can capture the waveform features of the ECG signal, while the Attention mechanism can identify the feature segments related to arrhythmia, thereby improving the accuracy of early warning of disease deterioration.

[0131] Preferably, when constructing the early warning model for disease deterioration, the following steps can be adopted. First, extract the data part that does not overlap with the second monitoring subset from the first monitoring subset to form an independent feature analysis set. Then, analyze the independent feature analysis set through the first feature extraction module to extract the first feature vector reflecting the long-period fluctuations. Next, input the first feature vector and the original data of the first monitoring subset into the second feature extraction module to generate the second feature vector. At the same time, input the second monitoring subset into the second feature extraction module to generate the third feature vector. After that, horizontally fuse the first feature vector and the second feature vector to generate the first fused feature; vertically fuse the first feature vector and the third feature vector to generate the second fused feature. Finally, input the first fused feature and the second fused feature into the sequential attention module, which adopts a deep learning algorithm based on the attention mechanism and can automatically learn the key features in the time series data and predict the clinical index parameters of the target future monitoring period. By comparing the predicted parameters with the historical same-period data of the third monitoring subset for similarity, if the difference exceeds the preset threshold, an early warning signal is triggered. In addition, the cross-period consistency verification results of the fourth monitoring subset can be combined to weight the early warning signal to improve the accuracy and reliability of the early warning.

[0132] Cross - cycle consistency verification uses the Kolmogorov - Smirnov test (KS test) to evaluate the distribution consistency between the predicted parameters and historical data for the same period. The KS test calculates the maximum vertical distance (D - value) between the cumulative distribution functions (CDFs) of two samples by comparing them, and determines whether to accept the null hypothesis (that is, the two samples come from the same distribution) based on the sample size and significance level (usually taken as 0.05). If the p - value of the KS test result for the predicted parameters and historical data for the same period is greater than 0.05, it is considered that the two distributions are consistent and the verification passes; otherwise, the verification fails. For example, when verifying the blood pressure prediction parameters during the third monitoring cycle of a patient, the predicted blood pressure values are subjected to the KS test with the historical blood pressure data for the same period. If the p - value is 0.03 (less than 0.05), it is considered that the predicted parameters and the historical data have inconsistent distributions, triggering the manual review process. After manual review, the model weights are adjusted according to the review results, and prediction and verification are carried out again until the verification passes.

[0133] In some embodiments, the processing logic when the similarity comparison does not meet the threshold includes:

[0134] Screen the historical data subset that matches the target future cycle date from the third monitoring subset;

[0135] Extract the contemporaneous historical data of all associated subsets from the fourth monitoring subset;

[0136] Compare the statistical distribution characteristics of the predicted parameters and historical data. If the distribution deviation exceeds the preset range, trigger the manual review process;

[0137] Update the model weights according to the review results and regenerate the warning signal.

[0138] It should be noted that the processing logic when the similarity comparison does not meet the threshold in the present invention is to further verify the accuracy of the model prediction results in the case where the disease deterioration warning model fails to trigger a warning, ensuring that the potential disease deterioration risk of patients is not missed due to model misjudgment. Among them, the similarity comparison refers to comparing and analyzing the clinical index parameters of the future monitoring cycle predicted by the model with the historical data for the same period to determine whether the difference between the two is within the allowable range. If the difference exceeds the preset threshold, a warning is triggered; when the similarity comparison does not meet the threshold, that is, the prediction result is relatively close to the historical data, but there may still be potential risks. At this time, further verification and processing are required.

[0139] Specifically, the third monitoring subset refers to the patient monitoring data collected during the third monitoring cycle, which is used for similarity comparison with the clinical index parameters predicted by the model for the future monitoring cycle. The fourth monitoring subset refers to the patient monitoring data collected during the fourth monitoring cycle, which is used for cross-cycle consistency verification to ensure the stability and reliability of the model prediction results. The historical data subset is the data screened from the third monitoring subset that matches the target future cycle date and is used for further comparison and analysis. Statistical distribution characteristics refer to the statistical characteristics of the data, such as the mean, variance, and distribution pattern. By comparing the predicted parameters with the statistical distribution characteristics of the historical data, it can be determined whether they have similar distribution laws. If the distribution deviation exceeds the preset range, it may indicate potential changes in the patient's condition and trigger the manual review process.

[0140] Preferably, during the processing when the similarity comparison does not meet the threshold, the following steps can be further refined. First, screen out the historical data subset that matches the target future cycle date from the third monitoring subset to ensure the comparability of the data for comparison. Then, extract the contemporaneous historical data of all associated subsets from the fourth monitoring subset for cross-cycle consistency verification. Next, perform a statistical distribution characteristic comparison between the clinical index parameters predicted by the model for the future monitoring cycle and the screened historical data subset, such as calculating statistical indicators such as the mean difference and variance difference between the two. If these differences exceed the preset range, it may indicate a deviation between the model prediction result and the actual condition, and at this time, trigger the manual review process. During the manual review process, medical staff can evaluate and adjust the model's prediction result based on the patient's specific situation and clinical experience. Finally, according to the result of the manual review, update the model weights and regenerate the warning signal to improve the accuracy and adaptability of the model.

[0141] Combined with specific examples, the following is a complete implementation case of sepsis warning for diabetic patients:

[0142] Patient Li, male, 55 years old, with a 10-year history of type 2 diabetes, was admitted to the hospital due to symptoms such as fever, cough, and dyspnea, and was diagnosed with diabetes complicated with sepsis. After admission, he was included in the intensive care unit monitoring and management system.

[0143] Monitoring cycle configuration: Set 72 hours as the main monitoring cycle (the first monitoring cycle), including a secondary monitoring cycle of 48 hours (the second monitoring cycle), the third monitoring cycle is 24 hours, and the fourth monitoring cycle includes 3 third monitoring cycles, with an interval of 8 hours between each cycle.

[0144] During the main monitoring cycle, pay attention to indicators such as the patient's blood glucose, lactate, heart rate, blood pressure, and oxygen saturation.

[0145] Cross-cycle feature fusion: In the fourth monitoring cycle, cross-cycle feature fusion is performed on blood glucose and lactate indicators.

[0146] First, calculate statistical features such as the mean, standard deviation, maximum value, and minimum value of blood glucose and lactate in each third monitoring cycle. Then, normalize these features so that their value ranges are within [0, 1]. Next, adopt a weighted fusion strategy, assign weights according to the importance of the features (such as the weight of the blood glucose mean is 0.4, and the weight of the lactate maximum value is 0.3), and calculate the fused feature value.

[0147] Finally, input the fused feature value into the disease deterioration warning model as one of the bases for judging disease deterioration.

[0148] Example of orange warning rendering in 3D visualization: In the 3D model of the virtual ward, according to the scoring results of the patient's disease impact level, if the patient's disease impact level reaches moderate (5 - 9 points), then apply orange warning rendering to the patient's bed and its sub-region. Through shader programming of the Unity3D engine, set the diffuse color of the bed model to RGB(255, 165, 0) (orange) and adjust its transparency to 0.7 to highlight the abnormal area. At the same time, apply color-coded mapping in the surrounding area of the patient, set the ground color of the sub-region to orange, and gradually reduce the color saturation according to the decreasing degree of disease impact to form an orange warning area spreading outward from the patient's bed, enabling medical staff to intuitively identify potential risk areas.

[0149] The above embodiments of the present invention have the following beneficial effects: The digital management system for critical care medicine can be based on a multi-cycle data collaborative analysis mechanism, and through the dynamic configuration and data fusion of the first monitoring cycle, the second monitoring cycle, the third monitoring cycle, and the fourth monitoring cycle, achieve accurate monitoring and early warning of the patient's condition. The system can utilize the temporal attention mechanism and feature fusion algorithm in the disease deterioration warning model, combined with the cross-validation of long-cycle fluctuation features and real-time monitoring data, to significantly improve the accuracy of disease prediction. Through the functions of automatic analysis of abnormal causes and intelligent assessment of the influence range, it can provide data support for clinical decision-making. At the same time, the structured comprehensive disease report can integrate multi-dimensional trend graphs, treatment event records, and encrypted video segments to form a complete diagnosis and treatment basis.

[0150] The system can rely on 3D visualization technology to achieve dynamic rendering of sub-regions and patients according to the disease impact level, and intuitively present the risk distribution through transparency level mapping and color coding strategies. The interactive operation function can support medical staff to quickly retrieve key information such as patients' vital signs, treatment records, and video playback, optimizing the clinical work efficiency. The hierarchical alarm mechanism can combine the cross-cycle consistency verification results to achieve intelligent grading and early warning of risks in the intensive care area, while the model self-learning function can automatically trigger the manual review process when the early warning error exceeds the standard, continuously optimizing the early warning accuracy, and ultimately improving the overall management efficiency and treatment quality of the intensive care unit.

[0151] As Figure 2 shown, an information digital management method based on the intensive care unit in some embodiments includes:

[0152] Obtain the patient monitoring data sets corresponding to each patient in the target intensive care unit, where each patient has a corresponding patient monitoring data sequence;

[0153] For each patient among all patients, perform the following generation steps:

[0154] Determine the patient monitoring data sequence corresponding to the patient as the target monitoring data sequence;

[0155] Determine the first monitoring cycle, the second monitoring cycle, the third monitoring cycle, and the fourth monitoring cycle corresponding to the patient for the current monitoring cycle, where the first monitoring cycle includes the second monitoring cycle, the time range corresponding to the first monitoring cycle is greater than that of the second monitoring cycle, the fourth monitoring cycle includes the third monitoring cycle, and the central monitoring cycle corresponding to the third monitoring cycle is within the same week as the current monitoring cycle;

[0156] Generate a first monitoring subset corresponding to the first monitoring cycle, a second monitoring subset corresponding to the second monitoring cycle, a third monitoring subset corresponding to the third monitoring cycle, and a fourth monitoring subset corresponding to the fourth monitoring cycle according to the target monitoring data sequence;

[0157] Generate a disease deterioration early warning signal for the patient by using the pre-trained disease deterioration early warning model according to the first monitoring subset, the second monitoring subset, the third monitoring subset, and the fourth monitoring subset;

[0158] In response to the disease deterioration early warning signal indicating that the patient's condition is abnormal, determine the abnormal cause analysis result and the impact range assessment result corresponding to the patient;

[0159] Generate a structured comprehensive disease report in a predetermined format according to the abnormal cause analysis result, the target monitoring data sequence, and the impact range assessment result;

[0160] Based on the obtained comprehensive condition reports for each case, perform dynamic visual displays for each patient and conduct hierarchical alarm processing for associated critical areas.

[0161] It can be understood that the various steps described in this information digital management method based on the intensive care unit correspond to the respective modules in the information digital management system based on the intensive care unit described in the reference Figure 1 Therefore, the modules, features, and beneficial effects described above for the information digital management system based on the intensive care unit also apply to the information digital management method based on the intensive care unit and the operations included therein, and will not be elaborated herein.

[0162] As Figure 3 shown, the electronic device 300 may include a processing device (such as a central processing unit, a graphics processing unit, etc.) 301, which can perform various appropriate actions and processes according to the program stored in the read-only memory (ROM) 302 or the program loaded from the storage device 308 into the random access memory (RAM) 303. In the RAM 303, various programs and data required for the operation of the electronic device 300 are also stored. The processing device 301, the ROM 302, and the RAM 303 are connected to each other through the bus 304. The input / output (I / O) interface 305 is also connected to the bus 304.

[0163] Generally, the following devices may be connected to the I / O interface 305: an input device 306 including, for example, a touch screen, a touchpad, a keyboard, a mouse, a camera, a microphone, an accelerometer, a gyroscope, etc.; an output device 307 including, for example, a liquid crystal display (LCD), a speaker, a vibrator, etc.; a storage device 308 including, for example, a magnetic tape, a hard disk, etc.; and a communication device 309. The communication device 309 can allow the electronic device 300 to communicate with other devices wirelessly or wiredly to exchange data. Although Figure 3 the electronic device 300 with various devices is shown, it should be understood that it is not required to implement or have all the shown devices. Instead, more or fewer devices may be implemented or had. Figure 3 Each block shown in

[0164] Further, the storage medium of the embodiments of the present application stores program instructions capable of implementing all the above methods. Among them, the program instructions can be stored in the above storage medium in the form of a software product, including several instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) or a processor to execute all or part of the steps of the methods described in the various embodiments of the present application. The foregoing storage medium includes: various media that can store program codes such as USB flash drives, mobile hard disks, read-only memories (ROMs), random access memories (RAMs), magnetic disks, or optical discs, or terminal devices such as computers, servers, mobile phones, and tablets.

[0165] The above description is only some preferred embodiments of the present invention and an explanation of the applied technical principles. Those skilled in the art should understand that the scope of the invention involved in the embodiments of the present invention is not limited to the technical solutions formed by the specific combination of the above technical features, and should also cover other technical solutions formed by any combination of the above technical features or their equivalent features without departing from the above inventive concept. For example, the technical solutions formed by mutually replacing the above features with the technical features (but not limited to) having similar functions disclosed in the embodiments of the present invention.

Claims

1. An information digital management system based on the intensive care unit, characterized in that, Including: A data acquisition module that obtains a patient monitoring data set corresponding to each patient in the target intensive care unit, and each patient has a corresponding patient monitoring data sequence; A data processing module that performs the following for each patient: Determine the patient monitoring data sequence corresponding to the patient as the target monitoring data sequence; Determine four monitoring cycle groups corresponding to the patient for the current monitoring cycle; Generate monitoring subset groups corresponding to the monitoring cycle groups respectively according to the target monitoring data sequence; Generate a deterioration warning signal for the patient by using a pre-trained deterioration warning model according to the monitoring subset groups; In response to the deterioration warning signal indicating that the patient's condition is abnormal, determine the abnormal cause analysis result and the impact range evaluation result corresponding to the patient; Generate a structured comprehensive condition report according to the abnormal cause analysis result, the target monitoring data sequence and the impact range evaluation result; A visualization and alarm module for performing dynamic visualization display for each patient according to the comprehensive condition report and performing hierarchical alarm processing on the associated intensive care areas.

2. The system according to claim 1, wherein The four monitoring cycle groups include a first monitoring cycle, a second monitoring cycle, a third monitoring cycle, and a fourth monitoring cycle; The first monitoring cycle includes the second monitoring cycle, the time range corresponding to the first monitoring cycle is greater than that of the second monitoring cycle, the fourth monitoring cycle includes the third monitoring cycle, and the central monitoring cycle corresponding to the third monitoring cycle is in the same cycle as the current monitoring cycle.

3. The system according to claim 2, wherein The dynamic visualization display includes: Determine the disease impact level of each sub-area in the target intensive care unit according to each comprehensive condition report; Based on the disease impact level and the comprehensive condition report, assign a visualization strategy to each patient and sub-area, and the visualization strategy includes transparency level mapping and color coding mapping; According to the visualization strategy, the abnormal states of the patient and the sub-area are rendered in real time in the virtual ward three-dimensional model.

4. The system according to claim 2, wherein The steps for determining the monitoring cycle configuration corresponding to the patient include: Match the initial monitoring cycle configuration from the preset association table according to the patient's clinical classification and the sub-area where the patient is located; Obtain the historical data of the patient's vital sign fluctuations and calculate its periodic characteristics; Based on the periodic characteristics, dynamically adjust the initial monitoring cycle configuration to generate an adapted first adjustment cycle, a second adjustment cycle, a third adjustment cycle, and a fourth adjustment cycle; Combine the current monitoring cycle to generate the final first monitoring cycle, second monitoring cycle, third monitoring cycle, and fourth monitoring cycle.

5. The system according to claim 3, wherein The steps for generating the comprehensive condition report include: Generate a multi-dimensional trend graph corresponding to the target monitoring data sequence, including the time series changes of heart rate, blood oxygen, and blood pressure; Obtain the time stamps, intervention measures, and associated video clips of the patient's abnormal treatment events; Generate an encrypted access link for the video clip; Integrate the abnormal cause analysis result, the multi-dimensional trend graph, the impact range evaluation result, the abnormal treatment event, and the encrypted link into a structured comprehensive condition report according to the clinical report template.

6. The system according to claim 4, wherein It also includes: In response to the interactive operation of the target patient in the virtual ward model triggered by the user, pop up an information category selection interface including vital signs, treatment records, and video playback; Extract the corresponding data from the comprehensive condition report according to the category selected by the user; Overlay and display data at the bed coordinate position of the target patient in the virtual ward model.

7. The system according to claim 2, wherein The operation logic of the deterioration warning model includes: The monitoring subset group includes a first monitoring subset, a second monitoring subset, a third monitoring subset, and a fourth monitoring subset; Extract the data part in the first monitoring subset that does not overlap with the second monitoring subset as the independent feature analysis set; Analyze the independent feature analysis set through the first feature extraction module to generate a first feature vector reflecting long-term cycle fluctuations; Fuse the first feature vector with the original data of the first monitoring subset through the second feature extraction module to generate a second feature vector; Input the second monitoring subset into the second feature extraction module to generate a third feature vector; Horizontally fuse the first feature vector and the second feature vector to generate a first fusion feature; Vertically fuse the first feature vector and the third feature vector to generate a second fusion feature; Input the first fusion feature and the second fusion feature into the time series attention module to predict the clinical index parameters of the target future monitoring cycle; Compare the predicted parameters with the historical same-cycle data of the third monitoring subset. If the difference exceeds the threshold, trigger a warning; Combine the cross-cycle consistency verification results of the fourth monitoring subset to generate a weighted deterioration warning signal, where the time series attention module uses a deep learning algorithm based on the attention mechanism.

8. The system according to claim 7, wherein The processing logic when the similarity comparison does not meet the threshold includes: Screen the historical data subset in the third monitoring subset that matches the target future cycle date; Extract the same-cycle historical data of all associated subsets from the fourth monitoring subset; Compare the statistical distribution characteristics of the predicted parameters and the historical data. If the distribution deviation exceeds the preset range, trigger an artificial review process; Update the model weights according to the review results and regenerate the warning signal.

9. An information digital management method based on the intensive care unit, characterized in that, Include: Obtain the patient monitoring data sets corresponding to each patient in the target intensive care unit. Among them, each patient has a corresponding patient monitoring data sequence; For each patient among all patients, perform the following generation steps: Determine the patient monitoring data sequence corresponding to the patient as the target monitoring data sequence; Determine the first monitoring cycle, second monitoring cycle, third monitoring cycle, and fourth monitoring cycle corresponding to the patient for the current monitoring cycle; Generate a first monitoring subset corresponding to the first monitoring cycle, a second monitoring subset corresponding to the second monitoring cycle, a third monitoring subset corresponding to the third monitoring cycle, and a fourth monitoring subset corresponding to the fourth monitoring cycle according to the target monitoring data sequence; Generate a deterioration warning signal for the patient by using the pre-trained deterioration warning model according to the first monitoring subset, the second monitoring subset, the third monitoring subset, and the fourth monitoring subset; In response to the deterioration warning signal indicating that the patient's condition is abnormal, determine the abnormal cause analysis result and the impact range assessment result corresponding to the patient; Generate a structured comprehensive condition report in a predetermined format according to the abnormal cause analysis result, the target monitoring data sequence, and the impact range assessment result; Perform dynamic visualization display for each patient according to the obtained comprehensive condition reports for each patient, and perform hierarchical alarm processing on the associated intensive care areas.

10. The method according to claim 9, wherein The first monitoring period includes the second monitoring period, the time range corresponding to the first monitoring period is greater than that of the second monitoring period, the fourth monitoring period includes the third monitoring period, and the central monitoring period corresponding to the third monitoring period is within the same period as the current monitoring period.

Images