An intelligent assessment and early warning system for emergency internal medicine patients

Abstract

This invention discloses an intelligent assessment and early warning system for emergency internal medicine conditions, belonging to the field of emergency internal medicine technology. This system, through a comprehensive assessment and early warning module and a blood glucose fluctuation and fundus disease analysis module, performs quantitative assessment based on multiple dimensions of pathological characteristics. The fusion of these data forms a comprehensive fundus disease risk factor, Rrisk, providing a comprehensive quantitative assessment of the fundus disease risk in diabetic patients. The system classifies patients' fundus disease risks according to set risk levels and, combined with the comprehensive fundus disease risk factor Rrisk, provides intervention suggestions at different levels. When a patient is at a high-risk level, the system can promptly trigger relevant early warnings, reminding medical staff to take targeted treatment or further examinations to prevent the condition from progressing to serious consequences such as blindness. This multi-dimensional comprehensive assessment not only improves the diagnostic accuracy of fundus diseases but also provides strong decision support for clinical treatment.

Description

Technical Field

[0001] This invention relates to the field of emergency internal medicine technology, specifically to an intelligent assessment and early warning system for emergency internal medicine conditions. Background Technology

[0002] Emergency medicine is a crucial department in the medical field, responsible for handling various acute conditions and sudden illnesses, especially in the management of acute complications of chronic diseases such as diabetes. With an aging population and the prevalence of chronic diseases, the number of diabetic patients is increasing year by year, and acute complications caused by diabetes, particularly diabetic retinopathy, are a key focus of diagnosis and treatment in emergency medicine. Diabetic retinopathy poses a significant threat to patients' eye health; its pathogenesis is complex and closely related to multiple factors, including fluctuations in blood glucose levels, disease duration, and retinal microvascular damage. Therefore, early detection and accurate assessment of diabetic retinopathy have become one of the core issues that urgently need to be addressed in emergency medicine.

[0003] Currently, the assessment of diabetic retinopathy in emergency medicine largely relies on traditional diagnostic methods, such as fundus examination and blood glucose monitoring. However, these methods have several limitations. First, while fundus examination is an important tool for diagnosing retinopathy, it requires specialized experience from physicians, has a long examination cycle, and lacks real-time accuracy, making it difficult to meet the rapid treatment needs of emergency departments. Second, the relationship between blood glucose fluctuations and fundus lesions is complex, and traditional assessment methods cannot accurately quantify the long-term impact of blood glucose fluctuations on fundus lesions. Furthermore, existing assessment systems often overlook the non-linear effects of diabetes duration on fundus lesions and lack comprehensive, data-driven assessment tools. Therefore, existing assessment methods have significant deficiencies in accuracy, real-time performance, and personalized management, failing to effectively support the rapid and accurate decision-making needs of emergency medicine departments. Summary of the Invention

[0004] To address the shortcomings of existing technologies, this invention provides an intelligent assessment and early warning system for emergency internal medicine conditions, which solves the problems mentioned in the background art.

[0005] To achieve the above objectives, the present invention is implemented through the following technical solutions: including a disease course extraction module, a feature extraction module, a blood glucose fluctuation and fundus disease analysis module, a comprehensive assessment and early warning module, and a depth assessment and early warning module;

[0006] The disease course extraction module extracts disease course data from the electronic health records (EHRs) of diabetic patients by constructing a disease assessment and early warning platform. At the same time, it constructs a time-series database and connects the time-series database with the disease assessment and early warning platform to store the collected disease course data in the time-series database.

[0007] The feature extraction module extracts disease course data from the time-series database for preprocessing to obtain a pathological dataset, and then extracts features from the pathological dataset to obtain pathological feature vectors.

[0008] The blood glucose fluctuation and fundus disease analysis module constructs a disease course algorithm model and a fundus disease risk algorithm model, inputs the obtained pathological feature vectors into the disease course algorithm model and the fundus disease risk algorithm model, and calculates and outputs the disease course influencing factor Rdt, the risk contribution value of retinal microvascular damage Rwx and the risk contribution value of blood glucose fluctuation Rbv respectively.

[0009] The comprehensive assessment and early warning module calculates and outputs the comprehensive fundus disease risk factor Rrisk by combining the output results of the disease course algorithm model and the fundus disease risk algorithm model. Based on the output result of the comprehensive fundus disease risk factor Rrisk, it makes a preliminary judgment on the risk of diabetic fundus disease and classifies the fundus disease of diabetic patients into different risk levels based on the preliminary judgment result.

[0010] The risk levels include Level 1 risk, Level 2 risk, and Level 3 risk.

[0011] The depth assessment and early warning module triggers depth analysis when a diabetic patient is identified as being at level 2 or 3 risk. The depth analysis calculates depth based on the comprehensive fundus lesion risk factor Rrisk and outputs a depth risk factor Rdeep. Based on the output of the depth risk factor Rdeep, the module determines the degree of fundus lesions and generates relevant early warning measures based on the depth determination results.

[0012] Preferably, the disease course extraction module includes a disease course extraction unit and a data storage unit;

[0013] The disease course extraction unit extracts the disease course data of diabetic patients by setting up an API application interface to connect the disease assessment and early warning platform to the hospital's electronic health record (EHR).

[0014] The disease progress data includes patient information, duration of diabetes (Dt), blood glucose monitoring data (B), and fundus examination images;

[0015] The data storage unit constructs a time-series database and sets up write and output ports. The write port stores the disease progress data into the time-series database according to the extraction timestamp. At the same time, it generates a unique character ID based on the patient information using encoding technology. Each unique character ID corresponds to the disease progress data of the relevant diabetic patient.

[0016] Preferably, the feature extraction module includes a preprocessing unit and a feature extraction unit;

[0017] The preprocessing unit extracts the disease course data through the write port and preprocesses the disease course data. The preprocessing methods include image preprocessing and data preprocessing.

[0018] The image preprocessing involves using AI image processing technology on fundus examination images in the disease course data to intelligently identify the number of retinal microbleeds (Nmic), the area of ​​retinal lesions (Aretina), the area of ​​retinal edema (Aedema), and the total area of ​​the retina (Azqymj) in the fundus examination images.

[0019] The data preprocessing involves cleaning and normalizing the diabetes duration Dt and blood glucose monitoring B in the disease course data, combined with the obtained retinal microbleeds number Nmic, retinal lesion area Aretina, retinal edema area Aedema, and total retinal area Azqymj. This process normalizes all data in the disease course data to the range of [0,1], removes outliers, missing data, and eliminates the influence of dimensions.

[0020] The pathological dataset is generated by summarizing the preprocessed data, including the number of retinal microbleeds (Nmic), the area of ​​retinal lesions (Aretina), the area of ​​retinal edema (Aedema), the total area of ​​the retina (Azqymj), the duration of diabetes (Dt), and the blood glucose monitoring data (B).

[0021] The feature extraction unit extracts pathological feature vectors based on the acquired pathological dataset.

[0022] The pathological feature vector includes the duration of diabetes Dt, blood glucose monitoring B, blood glucose fluctuation index Bv, retinal microvascular damage index Rmi, and retinal edema index Rsw.

[0023] The blood glucose fluctuation index Bv is extracted by calculating the standard deviation based on the blood glucose monitoring value B detected each time. The specific extraction formula is as follows: Where n represents the total number of measurements, and Bi represents the blood glucose value of the i-th measurement. This represents the average blood glucose measurement.

[0024] The retinal microvascular damage index Rmi is calculated and extracted based on the number of retinal microbleeds Nmic and the area of ​​retinal lesions Aretina. The specific extraction formula is as follows: ;

[0025] The retinal edema index Rsw is calculated and extracted based on the area of ​​the retinal edema region Aedema and the total retinal area Azqymj. The specific extraction formula is as follows: .

[0026] Preferably, the blood glucose fluctuation and fundus disease analysis module includes a disease course and lesion analysis unit, a microvascular damage analysis unit, and a blood glucose analysis unit;

[0027] The disease course and lesion analysis unit extracts the duration of diabetes Dt from the pathological feature vector, inputs it into the disease course algorithm model, calculates and outputs the disease course impact factor Rdt, and analyzes the impact of the disease course on fundus lesions.

[0028] The disease course impact factor Rdt is calculated and output using the following disease course algorithm model;

[0029] ;

[0030] In the formula, tanh represents the hyperbolic tangent function. The parameter represents the effect of regulating disease course on fundus lesions, and e represents an exponential function. Represents the exponentially decaying term. This indicates the sensitivity of the disease course to changes in risk. This represents the adjustment coefficient for the exponential decay term. This represents the speed control coefficient for the exponential decay term.

[0031] Preferably, the microvascular injury analysis unit constructs a fundus disease risk algorithm model, which includes a microvascular injury risk algorithm model and a blood glucose fluctuation risk algorithm model.

[0032] Simultaneously, the retinal microvascular injury index Rmi is extracted from the pathological feature vector and input into the microvascular injury risk algorithm model to calculate and output the risk contribution value Rwx of retinal microvascular injury, thereby quantifying the degree of retinal microvascular injury.

[0033] The risk contribution value Rwx of retinal microvascular injury is calculated and output by the following microvascular injury risk algorithm model;

[0034] ;

[0035] In the formula, ln represents the natural logarithm. Represents the logarithmic transformation term. Rmi represents the adjustment coefficient of the logarithmic transformation term. 3 The third term represents the retinal microvascular damage index. This represents the adjustment coefficient for the cubic term.

[0036] Preferably, the blood glucose analysis unit extracts the blood glucose fluctuation index Bv from the pathological feature vector, inputs it into the blood glucose fluctuation risk algorithm model, calculates and outputs the blood glucose fluctuation risk contribution value Rbv, and quantifies the impact of blood glucose fluctuation on fundus lesions.

[0037] The blood glucose fluctuation risk contribution value Rbv is calculated and output using the following blood glucose fluctuation risk algorithm model;

[0038] ;

[0039] In the formula, e represents an exponential function. The index term representing the glycemic fluctuation index, The coefficient representing the increase control of the exponential term in the blood glucose fluctuation index; sin represents the sine function. The sensitivity control coefficient representing the glycemic variability index. This represents the periodic control coefficient of the blood glucose fluctuation index.

[0040] Preferably, the comprehensive assessment and early warning module includes a comprehensive summary unit and a graded assessment unit;

[0041] The comprehensive aggregation unit calculates and outputs a comprehensive fundus disease risk factor Rrisk by combining blood glucose fluctuation with the disease course influencing factor Rdt, the risk contribution value of retinal microvascular damage Rwx, and the risk contribution value of blood glucose fluctuation Rbv output by the fundus disease analysis module, and the retinal edema index Rsw, thereby comprehensively quantifying the risk of fundus diseases.

[0042] The comprehensive fundus disease risk factor Rrisk is calculated and output using the following algorithm formula;

[0043] ;

[0044] In the formula, This represents the adjustment coefficient of the retinal edema index.

[0045] Preferably, the grading assessment unit uses the user to set the risk range of fundus lesions for diabetic patients [0.3, 0.6] based on the risk of fundus lesions, and then makes a preliminary judgment on the current risk of fundus lesions of diabetic patients based on the output of the comprehensive fundus lesion risk factor Rrisk, and classifies the lesion level of diabetic patients. The specific judgment content is as follows;

[0046] When the comprehensive risk factor for fundus diseases, Rrisk, is <0.3, it is marked as a level 1 risk.

[0047] When 0.3 ≤ comprehensive fundus disease risk factor Rrisk < 0.6, it is marked as level 2 risk;

[0048] When the comprehensive risk factor for fundus diseases, Rrisk, is ≥0.6, it is marked as level three risk;

[0049] Among them, the risk level of the lesion is from low to high, with level 1 risk < level 2 risk < level 3 risk.

[0050] Preferably, the deep assessment and early warning module includes a deep analysis unit and a deep assessment unit;

[0051] The deep analysis unit triggers deep analysis after initial assessment of level 2 and level 3 risk. This deep analysis extracts the average glycated hemoglobin (HbA1c) value of diabetic patients over 3 months from their electronic health records (EHR) and calculates the glycemic control index (Dci). The specific algorithm formula is as follows: HbA1c represents the average glycated hemoglobin value, 6.5 represents the diagnostic criteria for diabetes, and 8.5 represents abnormal blood glucose control.

[0052] Then, based on the blood glucose control index Dci and the preliminary comprehensive fundus disease risk factor Rrisk, the deep risk factor Rdeep is output for in-depth analysis of the fundus disease risk in diabetic patients.

[0053] The deep risk factor Rdeep is calculated and output using the following algorithm formula;

[0054] ;

[0055] In the formula, This represents the adjustment factor for the glycemic control index. This represents the second adjustment coefficient of the retinal edema index.

[0056] Preferably, the depth assessment unit extracts historical diabetes data from the hospital's electronic health record (EHR) via an API application interface, combines it with clinical examination results and treatment results, allows the user to set the lesion severity range [0.4, 0.7], and then outputs the obtained depth risk factor Rdeep to determine the risk of fundus lesions in diabetic patients. The specific determination content is as follows;

[0057] When the deep risk factor Rdeep < 0.4, it is recommended to check regularly.

[0058] When 0.4 ≤ deep risk factor Rdeep < 0.7, a first warning message is generated to prompt medical staff to adjust the medication for diabetic patients.

[0059] When the deep risk factor Rdeep ≥ 0.7, a second early warning message is generated to prompt medical staff to intervene and treat the patient.

[0060] This invention provides an intelligent assessment and early warning system for emergency internal medicine patients. It has the following beneficial effects:

[0061] (1) This system achieves intelligent acquisition and analysis of diabetic patients' disease course data, blood glucose monitoring, and fundus examination images by constructing a disease course extraction module, a feature extraction module, and a blood glucose fluctuation and fundus disease analysis module. The disease course extraction module extracts diabetic patients' disease course data by accessing the hospital's electronic health record (EHR) and stores it in a time-series database, providing a basis for subsequent pathological dataset extraction and pathological feature vector generation. Based on the pathological feature vector, the system comprehensively assesses the risk of fundus lesions in patients and identifies high-risk patients in the early stages, providing timely warnings. Through comprehensive analysis of the disease course influencing factor Rdt, the risk contribution value of retinal microvascular damage Rwx, and the risk contribution value of blood glucose fluctuation Rbv, the system can provide targeted warnings, providing a basis for early intervention of diabetic fundus lesions and significantly improving the early diagnosis and intervention effect of fundus lesions.

[0062] (2) This system uses a comprehensive assessment and early warning module and a blood glucose fluctuation and fundus disease analysis module to quantitatively assess multiple dimensions of pathological characteristics, including duration of diabetes, blood glucose fluctuation, retinal microvascular damage, and retinal edema. The fusion of these data forms a comprehensive fundus disease risk factor, Rrisk, providing a comprehensive quantitative assessment of the fundus disease risk in diabetic patients. The system classifies patients' fundus disease risks according to set risk levels and provides different levels of intervention recommendations based on the comprehensive fundus disease risk factor Rrisk. When a patient is at a high risk level, the system can promptly trigger relevant early warnings, reminding medical staff to take targeted treatments or further examinations to avoid serious consequences such as blindness. This multi-dimensional comprehensive assessment not only improves the diagnostic accuracy of fundus diseases but also provides strong decision support for clinical treatment, ensuring that patients receive the best treatment and management.

[0063] (3) This system provides a personalized risk assessment mechanism by combining the patient's blood glucose control status with the fundus disease risk factor Rrisk through a deep assessment and early warning module. When the patient's comprehensive fundus disease risk factor Rrisk exceeds the set risk threshold, the system can output the deep risk factor Rdeep through deep analysis and calculation, and take corresponding interventions according to the patient's specific condition. Especially when the fundus disease risk level is level II or III, the deep assessment module will combine the patient's glycated hemoglobin value (HbA1c) and blood glucose control status for in-depth analysis to accurately assess the progression of diabetic retinopathy and trigger necessary clinical interventions, such as increasing the frequency of fundus examinations or adjusting the treatment plan. Through this personalized intervention mechanism, the system can effectively reduce the aggravation of fundus disease caused by the lack of timely intervention, improve treatment effectiveness and the patient's quality of life. Attached Figure Description

[0064] Figure 1This is a schematic diagram of the process of an intelligent assessment and early warning system for emergency internal medicine patients according to the present invention. Detailed Implementation

[0065] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0066] Example 1

[0067] Please see Figure 1 This invention provides an intelligent assessment and early warning system for emergency internal medicine conditions. To achieve the above objectives, this invention is implemented through the following technical solutions: including a disease course extraction module, a feature extraction module, a blood glucose fluctuation and fundus disease analysis module, a comprehensive assessment and early warning module, and a depth assessment and early warning module.

[0068] The disease course extraction module extracts disease course data from the electronic health records (EHRs) of diabetic patients by building a disease assessment and early warning platform. At the same time, it builds a time-series database and connects the time-series database with the disease assessment and early warning platform to store the collected disease course data in the time-series database.

[0069] The feature extraction module extracts disease course data from the time series database for preprocessing to obtain a pathological dataset, and then extracts features from the pathological dataset to obtain pathological feature vectors.

[0070] The blood glucose fluctuation and fundus disease analysis module constructs a disease course algorithm model and a fundus disease risk algorithm model. The obtained pathological feature vectors are input into the disease course algorithm model and the fundus disease risk algorithm model, and the disease course influencing factor Rdt, the risk contribution value of retinal microvascular damage Rwx and the blood glucose fluctuation risk contribution value Rbv are calculated and output respectively.

[0071] The comprehensive assessment and early warning module calculates the comprehensive fundus disease risk factor Rrisk by combining the outputs of the disease course algorithm model and the fundus disease risk algorithm model. Based on the output of the comprehensive fundus disease risk factor Rrisk, it makes a preliminary judgment on the risk of diabetic fundus disease and classifies the fundus disease of diabetic patients into different risk levels based on the preliminary judgment results.

[0072] Risk levels are categorized into Level 1, Level 2, and Level 3.

[0073] The depth assessment and early warning module triggers depth analysis when diabetic patients are identified as being at level 2 or 3 risk. The depth analysis calculates depth based on the comprehensive fundus lesion risk factor Rrisk and outputs a depth risk factor Rdeep. Based on the output of the depth risk factor Rdeep, the module determines the degree of fundus lesions and generates relevant early warning measures based on the depth determination results.

[0074] In this embodiment, the system's disease progression extraction module accesses the electronic health records (EHRs) of diabetic patients and stores the disease progression data in a time-series database, ensuring efficient data management and retrieval. The feature extraction module preprocesses the disease progression data in the time-series database to obtain a pathological dataset and extract pathological feature vectors, providing detailed pathological information support and ensuring data accuracy and comprehensiveness. The blood glucose fluctuation and fundus disease analysis module constructs a disease progression algorithm model and a fundus disease risk algorithm model to analyze the pathological feature vectors, calculating and outputting the disease progression influencing factor Rdt, the risk contribution value Rwx for retinal microvascular damage, and the risk contribution value Rbv for blood glucose fluctuation, quantifying the risk factors of fundus lesions from multiple dimensions and providing accurate data support for comprehensive assessment. The comprehensive assessment and early warning module comprehensively calculates various risk factors, outputting a comprehensive fundus disease risk factor Rrisk, and makes a preliminary judgment on the risk of fundus lesions in diabetic patients. Based on the comprehensive risk factor, it classifies the risk level, thereby achieving effective early warning of fundus disease risk. When a diabetic patient is assessed as having level 2 or 3 risk, the in-depth assessment and early warning module initiates in-depth analysis. By further calculating the in-depth risk factor Rdeep and combining it with information such as the patient's glycated hemoglobin level, it accurately assesses the severity of fundus lesions and generates corresponding early warning measures based on the assessment results, thereby providing more targeted interventions. Compared with existing technologies, this invention, through real-time acquisition and in-depth analysis of multidimensional data, overcomes the limitations of traditional methods for assessing the risk of diabetic retinopathy, enabling more comprehensive and dynamic monitoring of changes in the patient's disease course and its impact on fundus lesions. Existing technologies often rely on a single data source or static diagnostic methods, while this system, through intelligent data processing and multi-model combined analysis, can not only identify the risk of fundus lesions earlier but also provide personalized early warnings and interventions based on different risk levels, greatly improving the accuracy of early warnings and the timeliness of treatment, and promoting the intelligent and precise management of diabetic retinopathy.

[0075] Example 2

[0076] Specifically: the disease course extraction module includes a disease course extraction unit and a data storage unit;

[0077] The disease course extraction unit connects the disease assessment and early warning platform to the hospital's electronic health record (EHR) by setting up an API application interface to extract the disease course data of diabetic patients;

[0078] The disease progression data includes patient information, duration of diabetes (Dt), blood glucose monitoring data (B), and fundus examination images;

[0079] The data storage unit constructs a time-series database and sets up write and output ports. Through the output port, the disease progress data is stored in the time-series database according to the extraction timestamp. At the same time, based on the patient information, a unique character ID is generated using encoding technology. Each unique character ID corresponds to the disease progress data of the relevant diabetic patient.

[0080] In this embodiment, the system's disease progression extraction unit connects the disease assessment and early warning platform to the hospital's Electronic Health Record (EHR) system via an API interface, enabling automated extraction of diabetic patients' disease progression data. This data acquisition process avoids errors and delays from manual entry, ensuring the real-time nature and accuracy of the data. The data storage unit constructs a time-series database to store disease progression data in an orderly manner according to timestamps. This not only improves the efficiency of data access but also generates a unique character ID for each patient through a unique encoding technology, ensuring accurate matching and personalized management of disease progression data with patient information. This module has significant advantages over existing technologies, breaking through the limitations of traditional manual input and static database management, and improving the automation level of data processing. Compared with existing technologies, traditional disease progression data acquisition and storage methods often suffer from data redundancy, information asymmetry, and processing delays. The innovation of this system lies in the combination of a time-series database and unique ID encoding, which not only enables real-time and accurate extraction and storage of disease progression data but also allows for rapid retrieval and multi-dimensional analysis through structured data storage, providing a solid data foundation for subsequent data processing and intelligent assessment.

[0081] Example 3

[0082] Specifically: the feature extraction module includes a preprocessing unit and a feature extraction unit;

[0083] The preprocessing unit extracts disease course data through the write port and preprocesses the disease course data. The preprocessing methods include image preprocessing and data preprocessing.

[0084] Image preprocessing uses AI image processing technology to intelligently identify the number of retinal microhemorrhages (Nmic), the area of ​​retinal lesions (Aretina), the area of ​​retinal edema (Aedema), and the total retinal area (Azqymj) in fundus examination images from the disease course data.

[0085] Data preprocessing involves cleaning and normalizing the duration of diabetes (Dt) and blood glucose monitoring (B) in the disease course data, combined with the obtained retinal microbleeds (Nmic), retinal lesion area (Aretina), retinal edema area (Aedema), and total retinal area (Azqymj). All data in the disease course data are normalized to the range of [0,1], and outliers, missing data, and the influence of units are removed.

[0086] The pathological dataset is generated by summarizing the preprocessed data, including the number of retinal microbleeds (Nmic), the area of ​​retinal lesions (Aretina), the area of ​​retinal edema (Aedema), the total area of ​​the retina (Azqymj), the duration of diabetes (Dt), and the blood glucose monitoring data (B).

[0087] The feature extraction unit extracts pathological feature vectors based on the acquired pathological dataset.

[0088] The pathological feature vector includes the duration of diabetes (Dt), blood glucose monitoring data (B), blood glucose fluctuation index (Bv), retinal microvascular damage index (Rmi), and retinal edema index (Rsw).

[0089] The glycemic variability index (Bv) is calculated by taking the standard deviation of each measured blood glucose level (B). The specific extraction formula is as follows: Where n represents the total number of measurements, and Bi represents the blood glucose value of the i-th measurement. This represents the average blood glucose measurement.

[0090] The retinal microvascular damage index Rmi is calculated and extracted based on the number of retinal microbleeds (Nmic) and the area of ​​retinal lesions (Aretina). The specific extraction formula is as follows: ;

[0091] The retinal edema index Rsw is calculated and extracted based on the area of ​​the edematous retinal region (Aedema) and the total retinal area (Azqymj). The specific extraction formula is as follows: .

[0092] In this embodiment, the system performs image and data preprocessing on the disease progression data through a preprocessing unit. AI image processing technology intelligently identifies key indicators in fundus examination images, such as the number of retinal microhemorrhages, the area of ​​retinal lesions, and the area of ​​retinal edema. Data preprocessing further optimizes the disease progression data by removing outliers, filling in missing data, and eliminating the influence of dimensions through data cleaning and normalization, thus ensuring data consistency and comparability. This precise data processing method lays a solid foundation for subsequent analysis and feature extraction. The feature extraction unit extracts pathological feature vectors from the pathological dataset, providing high-quality input data for comprehensive assessment. In particular, the extraction of features such as the blood glucose fluctuation index, retinal microvascular damage index, and retinal edema index further strengthens the quantitative assessment of disease risk, making the risk prediction of diabetic retinopathy more scientific and accurate. Compared with existing technologies, traditional pathological data processing often relies on manual analysis, resulting in high errors and inefficiency. Through the innovative image processing and data cleaning methods of this module, the system can automatically extract multi-dimensional pathological features from complex disease progression data, significantly improving processing efficiency and accuracy. Meanwhile, the standardization and quantification of the feature extraction process makes the assessment of diabetic retinopathy risk more objective and avoids interference from human factors.

[0093] Example 4

[0094] Specifically: The blood glucose fluctuation and fundus disease analysis module includes a disease course and lesion analysis unit, a microvascular damage analysis unit, and a blood glucose analysis unit;

[0095] The disease course and lesion analysis unit extracts the duration of diabetes Dt from the pathological feature vector, inputs it into the disease course algorithm model, calculates and outputs the disease course impact factor Rdt, and analyzes the impact of disease course on fundus lesions.

[0096] The disease course impact factor Rdt is calculated and output using the following disease course algorithm model;

[0097] ;

[0098] In the formula, tanh represents the hyperbolic tangent function, used to simulate the rapid increase in lesion risk in the early stages of the disease, followed by a gradual slowdown in the rate of increase. When the disease duration is small, the risk increases rapidly; when the disease duration increases to a certain extent, the increase tends to stabilize. This parameter represents the effect of disease progression on fundus lesions, controlling the amplitude of the hyperbolic tangent function. The larger this value, the stronger the effect of disease progression on fundus lesions. e represents an exponential function. Represents the exponentially decaying term. This indicates the sensitivity of the disease course to changes in risk. This represents the adjustment coefficient for the exponential decay term. This represents the rate control coefficient for the exponential decay term;

[0099] The impact of the duration of diabetes on fundus lesions is not linear, but rather shows a gradually increasing risk as the duration of the disease increases. However, the impact may tend to stabilize after a long period of disease. The hyperbolic tangent function tanh is used to model and analyze the impact of the duration of the disease.

[0100] The microvascular injury analysis unit constructs a fundus disease risk algorithm model, which includes a microvascular injury risk algorithm model and a blood glucose fluctuation risk algorithm model.

[0101] Simultaneously, the retinal microvascular injury index Rmi is extracted from the pathological feature vector and input into the microvascular injury risk algorithm model to calculate and output the risk contribution value Rwx of retinal microvascular injury, thereby quantifying the degree of retinal microvascular injury.

[0102] The risk contribution value Rwx of retinal microvascular injury is calculated and output using the following microvascular injury risk algorithm model;

[0103] ;

[0104] In the formula, ln represents the natural logarithm. Represents the logarithmic transformation term. Rmi represents the adjustment coefficient of the logarithmic transformation term, which controls for a relatively gradual change in the risk of initial microvascular injury. 3 The cubic term of the retinal microvascular damage index indicates that the risk increases with the severity of microvascular damage. This represents the cubic adjustment coefficient, which controls the degree to which the risk increases sharply in cases of severe microvascular injury;

[0105] Retinal microvascular damage is an important indicator of diabetic retinopathy, and its impact has strong nonlinear characteristics. The effect of microvascular damage on fundus lesions is modeled by using a formula combining logarithmic transformation and polynomials.

[0106] The blood glucose analysis unit extracts the blood glucose fluctuation index Bv from the pathological feature vector, inputs it into the blood glucose fluctuation risk algorithm model, calculates and outputs the blood glucose fluctuation risk contribution value Rbv, and quantifies the impact of blood glucose fluctuation on fundus lesions.

[0107] The risk contribution value of blood glucose fluctuation, Rbv, is calculated and output using the following blood glucose fluctuation risk algorithm model;

[0108] ;

[0109] In the formula, e represents an exponential function. The index term representing the glycemic fluctuation index, The coefficient representing the increase control of the exponential term in the blood glucose fluctuation index; sin represents the sine function. The sensitivity control coefficient representing the glycemic variability index. It represents the periodic control coefficient of the blood glucose fluctuation index, used to control the impact of postprandial blood glucose fluctuations.

[0110] In this embodiment, the disease progression analysis unit of the system analyzes the impact of disease progression on fundus lesions through a disease progression algorithm model. It accurately simulates the rapid growth characteristics of fundus lesions in the early stages of the disease using the hyperbolic tangent function tanh, and then shows a stable risk growth pattern as the disease progresses. This nonlinear model not only reveals the progressive impact of diabetes progression on fundus lesions but also allows for fine-tuning of the degree of impact of disease progression on risk changes, providing a more accurate disease progression factor Rdt for risk assessment. This breaks through the linear assumptions of traditional methods and provides a more realistic analysis of the impact of disease progression. The microvascular damage analysis unit, through fundus disease risk algorithm models, especially microvascular damage risk algorithm models, deeply analyzes the impact of retinal microvascular damage. Using logarithmic transformation and polynomial combination formulas, it successfully models the complex nonlinear relationship between retinal microvascular damage and fundus lesions, thus deriving the risk contribution value Rwx of retinal microvascular damage. This method not only quantifies microvascular damage but also reveals the different impacts of different degrees of damage on fundus lesions, providing more detailed and accurate damage assessment results. The blood glucose analysis unit focuses on the impact of blood glucose fluctuations on retinopathy, quantifying the contribution of the blood glucose fluctuation index (Bv) to retinopathy through a blood glucose fluctuation risk algorithm model. This model uses a combination of exponential and sine functions to effectively simulate the periodic changes in blood glucose fluctuations and their long-term effects on retinopathy, particularly improving the accuracy of blood glucose fluctuation analysis in terms of sensitivity control, especially in postprandial blood glucose fluctuations. Compared to existing technologies, this module, by combining nonlinear modeling and multidimensional data analysis, provides a more comprehensive and accurate assessment of diabetic retinopathy.

[0111] Example 5

[0112] Specifically: The comprehensive assessment and early warning module includes a comprehensive summary unit and a graded assessment unit;

[0113] The comprehensive summary unit calculates and outputs the comprehensive fundus disease risk factor Rrisk by combining the disease course impact factor Rdt, the risk contribution value of retinal microvascular damage Rwx, and the risk contribution value of blood glucose fluctuations output by the fundus disease analysis module with the retinal edema index Rsw, thereby comprehensively quantifying the risk of fundus diseases.

[0114] The comprehensive risk factor Rrisk for fundus diseases is calculated and output using the following algorithm formula;

[0115] ;

[0116] In the formula, This represents the retinal edema index adjustment coefficient, used to control the ultimate impact of retinal edema on risk.

[0117] The grading assessment unit uses the user to set the risk range of fundus lesions for diabetic patients [0.3, 0.6] based on the risk of fundus lesions. Then, based on the output of the comprehensive fundus lesion risk factor Rrisk, it makes a preliminary judgment on the current risk of fundus lesions in diabetic patients and classifies the lesion grades of diabetic patients. The specific judgment content is as follows:

[0118] When the comprehensive risk factor for fundus disease Rrisk < 0.3, it is marked as a level 1 risk. These patients have a low risk of fundus disease and are advised to have regular blood glucose control and fundus examinations.

[0119] When the comprehensive fundus disease risk factor Rrisk < 0.6, it is marked as level 2 risk. These patients have a moderate risk of fundus disease, and it is recommended to further increase the frequency of fundus examinations and monitor blood glucose fluctuations.

[0120] When the comprehensive fundus disease risk factor Rrisk ≥ 0.6, it is marked as level 3 risk. These patients have a high risk of fundus disease, and the system recommends in-depth fundus examination and possible interventions.

[0121] Among them, the risk level of the lesion is from low to high, with level 1 risk < level 2 risk < level 3 risk.

[0122] In this embodiment, the system's comprehensive aggregation unit integrates the disease progression influencing factor Rdt, the risk contribution value of retinal microvascular damage Rwx, the risk contribution value of blood glucose fluctuation Rbv, and the retinal edema index Rsw output by the blood glucose fluctuation and fundus disease analysis module. A comprehensive algorithm is then used to calculate the comprehensive fundus disease risk factor Rrisk, achieving quantitative and refined analysis of fundus disease risk. This algorithm, through weighted integration of various risk factors, effectively reflects the comprehensive risk of fundus diseases in diabetic patients, significantly improving the accuracy and detail of risk assessment. The risk classification and assessment unit makes a preliminary judgment on the comprehensive fundus disease risk factor Rrisk based on a set risk range and classifies the fundus disease risk of diabetic patients into risk levels. By classifying patients into Level 1, Level 2, and Level 3 risk levels, the system can provide personalized medical advice based on different risk levels. For Level 1 risk patients, regular blood glucose control and fundus examinations are recommended; for Level 2 risk patients, increased frequency of fundus examinations and monitoring of blood glucose fluctuations are recommended; and for Level 3 risk patients, in-depth fundus examinations and possible interventions are suggested. This detailed risk-level classification can provide clinicians with precise decision support, ensuring early detection and timely intervention of fundus disease risks.

[0123] Example 6

[0124] Specifically: The in-depth assessment and early warning module includes an in-depth analysis unit and an in-depth assessment unit;

[0125] The in-depth analysis unit triggers in-depth analysis after the initial risk level is determined to be level 2 or 3. In-depth analysis extracts the average glycated hemoglobin (HbA1c) value of diabetic patients over the past 3 months from their electronic health records (EHR) and calculates the glycemic control index (Dci). The specific algorithm formula is as follows: HbA1c represents the average glycated hemoglobin value, the normalized value, 6.5 represents the diagnostic criteria for diabetes, and 8.5 represents abnormal glycemic control.

[0126] Then, based on the blood glucose control index Dci and the preliminary comprehensive fundus disease risk factor Rrisk, the deep risk factor Rdeep is output for in-depth analysis of the fundus disease risk in diabetic patients.

[0127] The deep risk factor Rdeep is calculated and output using the following algorithm formula;

[0128] ;

[0129] In the formula, This represents the adjustment factor for the glycemic control index. This represents the second adjustment coefficient of the retinal edema index.

[0130] The depth assessment unit extracts historical diabetes data from the hospital's electronic health record (EHR) via an API application interface, combines it with clinical examination and treatment results, allows users to set the lesion severity range [0.4, 0.7], and then outputs the obtained depth risk factor Rdeep to make a depth assessment of the risk of fundus lesions in diabetic patients. The specific assessment content is as follows:

[0131] When the depth risk factor Rdeep < 0.4, it indicates that the fundus lesions in diabetic patients are normal, and regular check-ups are recommended.

[0132] When 0.4 ≤ depth risk factor Rdeep < 0.7, it indicates that the retinopathy of diabetic patients is gradually worsening but has not yet reached the point of emergency intervention. At this time, the first warning message is generated to prompt medical staff to adjust the medication of diabetic patients.

[0133] When the depth risk factor Rdeep ≥ 0.7, it indicates that the fundus lesions of diabetic patients are more severe. At this time, a second warning message is generated to prompt medical staff to intervene and treat the condition, such as laser therapy and anti-VEGF drugs.

[0134] In this embodiment, after the initial risk assessment classifies the patient as having level two or three risk, the deep analysis unit triggers deep analysis to conduct an in-depth assessment of the patient's blood glucose control. By extracting the glycated hemoglobin (HbA1c) value from the electronic health record (EHR) and calculating the blood glucose control index (Dci), combined with the comprehensive fundus disease risk factor (Rrisk), a deep risk factor (Rdeep) is derived. This risk factor calculation further considers the patient's blood glucose control level and the specific condition of fundus lesions, thus providing a more comprehensive analytical basis for assessing fundus disease risk. The deep assessment unit then combines clinical examination results, historical data, and treatment effects to perform a deep determination of the deep risk factor. Based on the calculation results, corresponding early warning information is automatically generated. This module can not only promptly detect the progression of fundus lesions in diabetic patients but also accurately provide intervention suggestions to medical staff, such as medication adjustments or laser treatment, significantly improving the targeting and effectiveness of treatment. Compared with existing technologies, the main innovation of the deep assessment and early warning module lies in the joint assessment of dynamic monitoring of blood glucose control and fundus disease risk. Traditional assessment methods may rely solely on a single indicator such as blood glucose or fundus lesions. This module, however, significantly improves the accuracy and timeliness of assessments by integrating multi-dimensional data and performing in-depth analysis. Through meticulous tracking and real-time alerts of patient conditions, the system effectively reduces the rate of missed or misdiagnosed fundus lesions, promotes the development of personalized treatment plans, thereby enhancing the prevention and treatment of diabetic retinopathy and improving patient prognosis.

[0135] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention.

Claims

1. An intelligent assessment and early warning system for emergency internal medicine patients, characterized in that: It includes a disease course extraction module, a feature extraction module, a blood glucose fluctuation and fundus disease analysis module, a comprehensive assessment and early warning module, and a depth assessment and early warning module; The disease course extraction module extracts disease course data from the electronic health records (EHRs) of diabetic patients by constructing a disease assessment and early warning platform. At the same time, it constructs a time-series database and connects the time-series database with the disease assessment and early warning platform to store the collected disease course data in the time-series database. The feature extraction module extracts disease course data from the time-series database for preprocessing to obtain a pathological dataset, and then extracts features from the pathological dataset to obtain pathological feature vectors. The blood glucose fluctuation and fundus disease analysis module constructs a disease course algorithm model and a fundus disease risk algorithm model, inputs the obtained pathological feature vectors into the disease course algorithm model and the fundus disease risk algorithm model, and calculates and outputs the disease course influencing factor Rdt, the risk contribution value of retinal microvascular damage Rwx and the risk contribution value of blood glucose fluctuation Rbv respectively. The blood glucose fluctuation and fundus disease analysis module includes a disease course and lesion analysis unit, a microvascular damage analysis unit, and a blood glucose analysis unit. The disease course and lesion analysis unit extracts the duration of diabetes Dt from the pathological feature vector, inputs it into the disease course algorithm model, calculates and outputs the disease course impact factor Rdt, and analyzes the impact of the disease course on fundus lesions. The disease course impact factor Rdt is calculated and output using the following disease course algorithm model; In the formula, tanh represents the hyperbolic tangent function, α1 represents the parameter that modulates the effect of disease course on fundus lesions, and e represents the exponential function. α1 represents the exponential decay term, α2 represents the sensitivity of the adjustment of the disease course to changes in risk, α3 represents the adjustment coefficient of the exponential decay term, and α4 represents the rate control coefficient of the exponential decay term. The microvascular injury analysis unit constructs a fundus disease risk algorithm model, which includes a microvascular injury risk algorithm model and a blood glucose fluctuation risk algorithm model. Simultaneously, the retinal microvascular injury index Rmi is extracted from the pathological feature vector and input into the microvascular injury risk algorithm model to calculate and output the risk contribution value Rwx of retinal microvascular injury, thereby quantifying the degree of retinal microvascular injury. The risk contribution value Rwx of retinal microvascular injury is calculated and output by the following microvascular injury risk algorithm model; Rwx=β1·ln(1+Rmi)+β2·Rmi 3 ; In the formula, ln represents the natural logarithm, ln(1+Rmi) represents the logarithmic transformation term, β1 represents the adjustment coefficient of the logarithmic transformation term, and Rmi 3 β2 represents the cubic term of the retinal microvascular damage index, and β2 represents the adjustment coefficient of the cubic term. The blood glucose analysis unit extracts the blood glucose fluctuation index Bv from the pathological feature vector, inputs it into the blood glucose fluctuation risk algorithm model, calculates and outputs the blood glucose fluctuation risk contribution value Rbv, and quantifies the impact of blood glucose fluctuation on fundus lesions. The blood glucose fluctuation risk contribution value Rbv is calculated and output using the following blood glucose fluctuation risk algorithm model; Rbv=γ1·e γ2·Bv ·sin(γ3·Bv); In the formula, e represents an exponential function, e γ2·Bv γ1 represents the growth rate control coefficient of the exponential term of the blood glucose fluctuation index, sin represents the sine function, γ2 represents the sensitivity control coefficient of the blood glucose fluctuation index, and γ3 represents the periodicity control coefficient of the blood glucose fluctuation index. The comprehensive assessment and early warning module calculates and outputs the comprehensive fundus disease risk factor Rrisk by combining the output results of the disease course algorithm model and the fundus disease risk algorithm model. Based on the output result of the comprehensive fundus disease risk factor Rrisk, it makes a preliminary judgment on the risk of diabetic fundus disease and classifies the fundus disease of diabetic patients into different risk levels based on the preliminary judgment result. The risk levels include Level 1 risk, Level 2 risk, and Level 3 risk. The depth assessment and early warning module triggers depth analysis when a diabetic patient is identified as being at level 2 or 3 risk. The depth analysis calculates depth based on the comprehensive fundus lesion risk factor Rrisk and outputs a depth risk factor Rdeep. Based on the output of the depth risk factor Rdeep, the module determines the degree of fundus lesions and generates relevant early warning measures based on the depth determination results.

2. The intelligent assessment and early warning system for emergency internal medicine patients according to claim 1, characterized in that: The disease course extraction module includes a disease course extraction unit and a data storage unit; The disease course extraction unit extracts the disease course data of diabetic patients by setting up an API application interface to connect the disease assessment and early warning platform to the hospital's electronic health record (EHR). The disease progress data includes patient information, duration of diabetes (Dt), blood glucose monitoring data (B), and fundus examination images; The data storage unit constructs a time-series database and sets up write and output ports. The write port stores the disease progress data into the time-series database according to the extraction timestamp. At the same time, it generates a unique character ID based on the patient information using encoding technology. Each unique character ID corresponds to the disease progress data of the relevant diabetic patient.

3. The intelligent assessment and early warning system for emergency internal medicine patients according to claim 2, characterized in that: The feature extraction module includes a preprocessing unit and a feature extraction unit; The preprocessing unit extracts the disease course data through the write port and preprocesses the disease course data. The preprocessing methods include image preprocessing and data preprocessing. The image preprocessing involves using AI image processing technology on fundus examination images in the disease course data to intelligently identify the number of retinal microbleeds (Nmic), the area of ​​retinal lesions (Aretina), the area of ​​retinal edema (Aedema), and the total area of ​​the retina (Azqymj) in the fundus examination images. The data preprocessing involves cleaning and normalizing the diabetes duration Dt and blood glucose monitoring data B in the disease course data, combined with the obtained retinal microbleeds number Nmic, retinal lesion area Aretina, retinal edema area Aedema, and total retinal area Azqymj. All data in the disease course data are normalized to the range [0,1], outliers, missing data, and the influence of dimensions are removed. The pathological dataset is generated by summarizing the preprocessed data, including the number of retinal microbleeds (Nmic), the area of ​​retinal lesions (Aretina), the area of ​​retinal edema (Aedema), the total area of ​​the retina (Azqymj), the duration of diabetes (Dt), and the blood glucose monitoring data (B). The feature extraction unit extracts pathological feature vectors based on the acquired pathological dataset. The pathological feature vector includes the duration of diabetes Dt, blood glucose monitoring B, blood glucose fluctuation index Bv, retinal microvascular damage index Rmi, and retinal edema index Rsw. The blood glucose fluctuation index Bv is extracted by calculating the standard deviation based on the blood glucose monitoring value B detected each time. The specific extraction formula is as follows: Where n represents the total number of measurements, and Bi represents the i-th blood glucose measurement value. This represents the average blood glucose measurement. The retinal microvascular damage index Rmi is calculated and extracted based on the number of retinal microbleeds Nmic and the area of ​​retinal lesions Aretina. The specific extraction formula is as follows: The retinal edema index Rsw is calculated and extracted based on the area of ​​the retinal edema region Aedema and the total retinal area Azqymj. The specific extraction formula is as follows:

4. The intelligent assessment and early warning system for emergency internal medicine patients according to claim 3, characterized in that: The comprehensive assessment and early warning module includes a comprehensive summary unit and a graded assessment unit; The comprehensive aggregation unit calculates and outputs a comprehensive fundus disease risk factor Rrisk by combining blood glucose fluctuation with the disease course influencing factor Rdt, the risk contribution value of retinal microvascular damage Rwx, and the risk contribution value of blood glucose fluctuation Rbv output by the fundus disease analysis module, and the retinal edema index Rsw, thereby comprehensively quantifying the risk of fundus diseases. The comprehensive fundus disease risk factor Rrisk is calculated and output using the following algorithm formula; Rrisk=(Rdt+Rwx+Rbv)·(1+δ·Rsw); In the formula, δ represents the retinal edema index adjustment coefficient.

5. The intelligent assessment and early warning system for emergency internal medicine patients according to claim 4, characterized in that: The grading assessment unit uses the user to set the risk range of fundus lesions for diabetic patients [0.3, 0.6] based on the risk of fundus lesions, and then makes a preliminary judgment on the current risk of fundus lesions of diabetic patients based on the output of the comprehensive fundus lesion risk factor Rrisk, and classifies the lesion level of diabetic patients. The specific judgment content is as follows: When the comprehensive risk factor for fundus diseases, Rrisk, is <0.3, it is marked as a level 1 risk. When 0.3 ≤ comprehensive fundus disease risk factor Rrisk < 0.6, it is marked as level 2 risk; When the comprehensive risk factor for fundus diseases, Rrisk, is ≥0.6, it is marked as level three risk; Among them, the risk level of the lesion is from low to high, with level 1 risk < level 2 risk < level 3 risk.

6. The intelligent assessment and early warning system for emergency internal medicine patients according to claim 5, characterized in that: The in-depth assessment and early warning module includes an in-depth analysis unit and an in-depth assessment unit; The deep analysis unit triggers deep analysis after initial assessment of level 2 and level 3 risk. This deep analysis extracts the average glycated hemoglobin (HbA1c) value of diabetic patients over 3 months from their electronic health records (EHR) and calculates the glycemic control index (Dci). The specific algorithm formula is as follows: HbA1c represents the average glycated hemoglobin value, 6.5 represents the diagnostic criteria for diabetes, and 8.5 represents abnormal blood glucose control. Then, based on the blood glucose control index Dci and the preliminary comprehensive fundus disease risk factor Rrisk, the deep risk factor Rdeep is output for in-depth analysis of the fundus disease risk in diabetic patients. The deep risk factor Rdeep is calculated and output using the following algorithm formula; Rdeep=(Rrisk+eta·Dci)·(1+δ2·Rsw); In the formula, η represents the blood glucose control index adjustment coefficient, and δ2 represents the second adjustment coefficient of the retinal edema index.

7. The intelligent assessment and early warning system for emergency internal medicine patients according to claim 6, characterized in that: The depth assessment unit extracts historical diabetes data from the hospital's electronic health record (EHR) via an API application interface, combines it with clinical examination and treatment results, allows the user to set the lesion severity range [0.4, 0.7], and then outputs the obtained depth risk factor Rdeep to determine the risk of fundus lesions in diabetic patients. The specific determination content is as follows: When the deep risk factor Rdeep < 0.4, it is recommended to check regularly. When 0.4 ≤ deep risk factor Rdeep < 0.7, a first warning message is generated to prompt medical staff to adjust the medication for diabetic patients. When the deep risk factor Rdeep ≥ 0.7, a second early warning message is generated to prompt medical staff to intervene and treat the patient.

Images