An intelligent auxiliary system for eye care

Abstract

This invention relates to the field of medical and nursing technology, specifically to an intelligent auxiliary system for ophthalmic nursing. The system includes a health indicator distribution difference analysis module that acquires health indicator data of ophthalmic patients before and after nursing care, compares the differences between the health indicator data before and after nursing care, and generates health indicator transfer characteristic analysis results. In this invention, by analyzing the differences between health indicator data before and after nursing care, multidimensional health data can be effectively normalized and its change characteristics quantified, thereby accurately presenting the specific effects of nursing intervention. This quantification of change characteristics not only makes the dynamic changes of health indicators more comparable but also significantly improves the accuracy of data processing. Regarding dynamic trend prediction, by extracting the changing trends of health indicators in a time series and dividing the trend change points into multiple stages, the system can accurately capture the transition nodes and trend directions of health status.

Description

Technical Field

[0001] This invention relates to the field of medical care technology, and in particular to an intelligent auxiliary system for ophthalmic care. Background Technology

[0002] The field of medical and nursing technology involves the application of technologies such as artificial intelligence, big data, and the Internet of Things to achieve real-time monitoring of patients' health status, remote nursing, and precise health intervention, thereby improving the efficiency of medical resource utilization and the quality of patient care.

[0003] Among them, the intelligent auxiliary system for ophthalmic care is a digital nursing solution based on artificial intelligence and big data technology, used for health monitoring, nursing guidance, and recovery management of patients with ophthalmic diseases. This system can intelligently analyze patients' eye health data to provide medical staff with diagnostic and treatment assistance suggestions.

[0004] Current technologies primarily rely on basic analysis and diagnostic support of eye health data, but they have limitations in processing complex time-series data and capturing dynamic trends in health indicators. Due to a lack of in-depth application of data normalization before and after nursing care, it is often difficult to intuitively present the specific differences in the effects of nursing interventions. Furthermore, the failure to effectively integrate health data trends with dynamic time-series characteristics means that diagnostic recommendations derived solely from static analysis lack precise capture of trend change nodes, potentially leading to lags in health assessment and prediction. When identifying key change nodes in health data, current technologies struggle to extract key data features from continuous curves, resulting in incomplete extraction of nursing information and affecting the accurate assessment of nursing outcomes. Summary of the Invention

[0005] The purpose of this invention is to address the shortcomings of existing technologies by proposing an intelligent auxiliary system for ophthalmic care.

[0006] To achieve the above objectives, the present invention adopts the following technical solution: an intelligent auxiliary system for ophthalmic care includes:

[0007] The Health Indicator Distribution Difference Analysis Module acquires health indicator data of ophthalmology patients before and after nursing care, compares the differences between the health indicator data of ophthalmology patients before and after nursing care, and generates health indicator transfer characteristic analysis results.

[0008] The dynamic trend prediction and segmented modeling module determines the trend of health indicator data of ophthalmology patients from before to after nursing care based on the analysis results of the health indicator transfer characteristics. It extracts the trend change points from the data and constructs a local fitting model based on the trend change points of each stage. It analyzes the direction of change of health indicator data and predicts the future characteristic changes of health data in each stage based on the direction of change, generating segmented trend prediction results.

[0009] The health data curvature feature extraction module converts the segmented trend prediction results into a continuous change curve and fits it. It calculates the curvature value of discrete points in the fitted change curve and compares it with a preset curvature threshold. Curvature values ​​exceeding the curvature threshold are marked as key trend nodes, and key trend node identification results are generated.

[0010] The health indicator dynamic update module classifies the key trend nodes in the key trend node identification results, analyzes the distribution pattern of the key trend nodes in the fitted change curve based on the classification results, and generates reference information for ophthalmic care.

[0011] As a further aspect of the present invention, the steps for obtaining the health indicator transfer characteristic analysis results are specifically as follows:

[0012] Based on patient health index data before and after nursing care, intraocular pressure, retinal thickness and corneal morphology parameters were obtained, and the data from multiple dimensions were normalized to obtain an integrated health index data table.

[0013] Based on the integrated health indicator data table, the formula is used:

[0014] ;

[0015] Calculate the first Difference values ​​of health indicators This yields data on the differences in indicators;

[0016] in, It is the first The values ​​of several health indicators after nursing care. It is the first The values ​​of the following health indicators before nursing care;

[0017] Based on the difference data of the indicators, the maximum value, minimum value and concentrated distribution range of each type of indicator are statistically analyzed. The positive and negative change range of the health indicators are determined by combining the upper and lower limits of the difference values. The results of the health indicator transfer characteristic analysis are generated based on the statistical information.

[0018] As a further aspect of the present invention, the step of extracting trend change points specifically includes:

[0019] Based on the analysis results of the health indicator transfer characteristics, by organizing the intraocular pressure, retinal thickness and corneal morphology parameter data before and after nursing care, a data sequence is generated according to the time point order. The rate and direction of change of adjacent time points in the time sequence are analyzed one by one to generate the analysis results of the time series change trend of health indicators.

[0020] Based on the time series trend analysis results of the aforementioned health indicators, the following formula is used:

[0021] ;

[0022] Calculate the maximum absolute value in the time series of health indicators This allows us to identify the points where the trend changes.

[0023] in, The variable representing the maximum value of the objective function. It is a point-in-time variable. Indicates the time point at which nursing care begins. and the time point at which the care ends , It is the change in health indicator values. It represents the interval between two adjacent time points in a time series.

[0024] As a further aspect of the present invention, the step of obtaining the local fitting model specifically includes:

[0025] The time series is divided into multiple stages based on the trend change points. Health indicator data in each stage are organized and grouped to generate a stage health indicator data table.

[0026] Based on the aforementioned phased health indicator data table, by analyzing the changing patterns of data within each phase and the changing trends of health indicators within the corresponding phase, a local fitting model is obtained through data fitting.

[0027] As a further aspect of the present invention, the steps for obtaining the segmented trend prediction results are specifically as follows:

[0028] Based on the data in the local fitting model, by extracting the key parameters in the fitting model at each stage, and combining the fitting results to analyze the direction of change of health indicators, the trend of change is segmented and organized to generate the analysis results of the current direction of change of health indicators.

[0029] Based on the analysis results of the current health indicator changes, the following formula is used:

[0030] ;

[0031] Calculate the future Health indicator values ​​at each time point The segmented trend prediction results are obtained.

[0032] in, These are health indicator values ​​at the current point in time. It is a linear slope. It is the coefficient of second variation. This indicates the length of time between the current time point and the predicted time point.

[0033] As a further aspect of the present invention, the step of obtaining the curvature values ​​of discrete points in the fitted change curve specifically comprises:

[0034] The segmented trend prediction results are supplemented with the time points and corresponding health indicator values ​​of each data segment by linear interpolation, and then converted into a continuous change curve to generate a continuous change dataset of health indicators.

[0035] Based on the aforementioned dataset of continuous changes in health indicators, the following formula is used:

[0036] ;

[0037] Calculate the curvature of the target discrete points This yields the curvature values ​​at discrete points;

[0038] in, This represents the slope of the fitted curve at the target position. This represents the acceleration of the fitted curve at the target position. Used to normalize curvature values.

[0039] As a further aspect of the present invention, the steps for obtaining the key trend node identification results are specifically as follows:

[0040] The curvature values ​​of the discrete points are compared point by point with a preset curvature threshold. The comparison determines whether the curvature at each time point exceeds the threshold, and the data is recorded to generate a curvature threshold comparison result.

[0041] Based on the curvature threshold comparison results, time points exceeding the curvature threshold and corresponding curvature value data are extracted. By marking target time points as key trend nodes and recording the marking information, key trend node identification results are established.

[0042] As a further aspect of the present invention, the step of obtaining reference information for ophthalmic care specifically includes:

[0043] Based on the key trend node identification results, all key trend nodes in the fitted curve are collected, including the time point and curvature value corresponding to the node. By sorting the curvature values ​​and classifying them into multiple node types according to grouping rules, the classification results of key trend nodes are generated.

[0044] Based on the classification results of the key trend nodes, the time point and curvature value of each category node are extracted, the distribution density of nodes on the time axis and the time span of density regions are analyzed, the changing trend of the curve and the dense node segments are marked, and reference information for ophthalmic care is obtained.

[0045] Compared with the prior art, the advantages and positive effects of the present invention are as follows:

[0046] In this invention, by analyzing the differences in health indicator data before and after nursing care, multidimensional health data can be effectively normalized and its change characteristics quantified, thereby accurately presenting the specific effects of nursing intervention. This quantification of change characteristics not only makes the dynamic changes of health indicators more comparable but also significantly improves the accuracy of data processing. Regarding dynamic trend prediction, by extracting the changing trends of health indicators in a time series and dividing trend change points into multiple stages, the analysis of time series data becomes more detailed and targeted, accurately capturing the nodes of health status transition and trend direction. Segmented modeling of trend changes allows for precise fitting of data features at each stage, helping to more reliably predict the trajectory and development characteristics of future health data changes, providing a scientific basis for ophthalmic nursing. Through the extraction of curvature features, precise analysis of discrete points in the data change curve can effectively identify key trend nodes in health data. This labeling and analysis of the distribution patterns of trend nodes makes nursing reference information more comprehensive, facilitating the formulation of more targeted nursing decisions. Attached Figure Description

[0047] Figure 1 This is a system flowchart of the present invention;

[0048] Figure 2 This is a flowchart illustrating the process of obtaining the health indicator transfer characteristic analysis results in this invention;

[0049] Figure 3 This is a flowchart illustrating the process of extracting trend change points in this invention;

[0050] Figure 4 This is a flowchart of the process for obtaining a local fitting model in this invention;

[0051] Figure 5 This is a flowchart illustrating how the present invention obtains segmented trend prediction results;

[0052] Figure 6 This is a flowchart illustrating the calculation of the curvature values ​​of discrete points in the fitted variation curve according to the present invention.

[0053] Figure 7 This is a flowchart illustrating the process of obtaining key trend node identification results in this invention;

[0054] Figure 8 This is a flowchart illustrating the process of obtaining reference information for ophthalmic care according to the present invention. Detailed Implementation

[0055] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0056] In the description of this invention, it should be understood that the terms "length," "width," "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "top," "bottom," "inner," and "outer," etc., indicating orientation or positional relationships, are based on the orientation or positional relationships shown in the accompanying drawings and are only for the convenience of describing the invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation, and therefore should not be construed as a limitation of the invention. Furthermore, in the description of this invention, "a plurality of" means two or more, unless otherwise explicitly specified.

[0057] Please see Figure 1 An intelligent auxiliary system for eye care includes:

[0058] The Health Indicator Distribution Difference Analysis Module acquires health indicator data of ophthalmology patients before and after nursing care, including intraocular pressure, retinal thickness and corneal morphology parameters, and performs normalization processing. It compares the differences between the health indicator data of ophthalmology patients before and after nursing care, quantifies the change characteristics of health indicator data based on the difference values, and generates health indicator transfer characteristic analysis results.

[0059] The dynamic trend prediction and segmented modeling module determines the trend of health indicator data of ophthalmology patients from before to after nursing care based on the analysis results of health indicator transition characteristics. It extracts trend change points from the trend, which are the moments when the health indicators change in the trend. Based on the trend change points, the time series is divided into multiple stages, and a local fitting model of health indicator data corresponding to each stage is constructed. The direction of change of health indicator data is analyzed based on the data in the local fitting model, and the future characteristic changes of health data in each stage are predicted based on the direction of change, generating segmented trend prediction results.

[0060] The health data curvature feature extraction module converts the segmented trend prediction results into a continuous change curve. By fitting the change curve of each segment of data in the continuous change curve, the curvature value of discrete points in the fitted change curve is calculated and compared with the preset curvature threshold. Curvature values ​​exceeding the curvature threshold are marked as key trend nodes, and key trend node identification results are generated.

[0061] The health indicator dynamic update module classifies the key trend nodes in the key trend node identification results, analyzes the distribution pattern of key trend nodes in the fitted change curve based on the classification results, and generates reference information for ophthalmic care.

[0062] The results of the health indicator transfer characteristic analysis include the difference values ​​of health indicators before and after nursing care, the characteristics of the trend of difference changes, and the direction of health indicator transfer. The results of the segmented trend prediction include the time location of the trend change point, the local fitting characteristics of the segmented health indicators, and the future change trend information of the stage-specific health indicators. The results of the key trend node identification include the location distribution of key trend nodes, the curvature value corresponding to the node, and the categories of rapidly changing nodes and steadily changing nodes. The reference information for ophthalmological nursing includes the dynamic change characteristics of key trend nodes, the stage-specific change direction of health indicators, and suggestions for the priority of nursing adjustment indicators.

[0063] Please see Figure 2 The specific steps for obtaining the results of the health indicator transfer characteristic analysis are as follows:

[0064] Based on patient health index data before and after nursing care, intraocular pressure, retinal thickness and corneal morphology parameters were obtained, and the data from multiple dimensions were normalized to obtain an integrated health index data table.

[0065] Health indicator data of patients before and after nursing care were collected from the ophthalmology nursing data acquisition system. Specifically, the data included intraocular pressure, retinal thickness and corneal morphology parameters. After grouping the data by patient, the data was input into the data processing module. The maximum and minimum values ​​of each group of data were normalized. The normalization process was as follows: first, the maximum and minimum values ​​of each indicator were selected from the original dataset, and then the normalization process was applied to each data in turn to transform all data into the interval [0, 1]. A normalized health indicator table was generated by the data recording module.

[0066] Based on the integrated health indicator data table, the formula is used:

[0067] ;

[0068] Calculate the first Difference values ​​of health indicators This yields data on the differences in indicators;

[0069] in, It is the first The values ​​of several health indicators after care were collected using ophthalmic examination equipment, such as intraocular pressure values ​​obtained using a non-contact tonometer, retinal thickness values ​​obtained using a retinal imaging system, and corneal morphology parameters obtained using a corneal topography system. It is the first The values ​​of these health indicators before and after nursing care were obtained in the same way as after nursing care, through the collection of examination data before nursing care. Categories representing health indicators include intraocular pressure, retinal thickness, and corneal morphology parameters.

[0070] Suppose that the patient's pre-care health indicators, measured by the device, yielded the following results: intraocular pressure of 21 mmHg, retinal thickness of 250 μm, and corneal morphology parameter of 0.8. After care, the same device measured the following results: intraocular pressure of 18 mmHg, retinal thickness of 240 μm, and corneal morphology parameter of 0.85.

[0071] Regarding intraocular pressure: ;

[0072] Regarding retinal thickness: ;

[0073] Regarding corneal morphological parameters: ;

[0074] The results showed that intraocular pressure decreased by 3 mmHg, retinal thickness decreased by 10 μm, and corneal morphological parameters increased by 0.05.

[0075] Based on the data on differences in indicators, the maximum, minimum and concentrated distribution range of each type of indicator are statistically analyzed. The positive and negative change ranges of each health indicator are determined by combining the upper and lower limits of the difference values. Based on the statistical information, the analysis results of the health indicator transfer characteristics are generated.

[0076] First, the differences in patient data before and after nursing care were categorized and organized according to health indicators, such as intraocular pressure, retinal thickness, and corneal morphology parameters. All patient data for each health indicator were summarized, and the average change range and upper and lower limits of each category of differences were calculated. The positive and negative change ranges were determined using the maximum and minimum values ​​of the difference data. For example, a change in intraocular pressure of -0.3 to -0.1 after nursing care indicates an overall decreasing trend in intraocular pressure. Then, based on the distribution of all patient data, the concentrated distribution range of the difference values ​​was calculated. For example, the difference in retinal thickness was concentrated in the range of -0.2 to -0.1, while some individual values ​​showed larger deviations. These statistical results were transformed into a change characteristic range graph for each health indicator. The change trend of each health indicator was presented by drawing a data distribution chart, and positive and negative ranges were distinguished by color to visually show the positive or negative impact of nursing care on health indicators. Finally, the results of the health indicator transfer characteristic analysis were generated.

[0077] Please see Figure 3 The specific steps for extracting trend change points are as follows:

[0078] Based on the analysis results of the health indicator transfer characteristics, by organizing the intraocular pressure, retinal thickness and corneal morphology parameter data before and after nursing care, a data sequence was generated according to the time point order. The rate and direction of change of adjacent time points in the time sequence were analyzed one by one to generate the analysis results of the time series change trend of health indicators.

[0079] By organizing intraocular pressure (IOP), retinal thickness, and corneal morphology parameters before and after nursing care, a data sequence was generated according to time points. For example, if a patient's IOP was 21 mmHg before nursing care, and the measurements after nursing care were 19 mmHg on day 7, 18 mmHg on day 14, and 18 mmHg on day 21, then the IOP changes in the time series would be: Calculate the rate of change for the sequence at adjacent time points, such as the rate of change from day 1 to day 7. mmHg / day, the rate of change from day 7 to day 14 mmHg / day is used to comprehensively determine the downward trend of intraocular pressure. Similarly, the time series data changes of retinal thickness and corneal morphology parameters are analyzed to extract their rate and direction of change, and a complete trend analysis report of the time series changes before and after care is generated.

[0080] Based on the time series trend analysis results of health indicators, the following formula is used:

[0081] ;

[0082] Calculate the maximum absolute value in the time series of health indicators This allows us to identify the points where the trend changes.

[0083] in, This indicates finding the variable from the function that maximizes the value of the objective function. It is a point-in-time variable, representing the time point at which each data point is collected during the nursing process, and belongs to the element range of time series. The time interval is determined by pre-setting the time interval in the experimental design, for example, collecting data once a day. For each day's time, It refers to the start and end range of the time series, indicating the point in time when nursing care begins. and the time point at which the care ends , This refers to the change in health indicator values, representing the difference in health indicator values ​​between two adjacent time points before and after nursing care (such as the difference in intraocular pressure). These values ​​are obtained through health indicator data collection devices before and after nursing care. For example, intraocular pressure is measured using a non-contact tonometer, retinal thickness is measured using a retinal imaging system, and corneal morphology parameters are measured using a corneal topography system. This represents the interval between two adjacent time points in a time series. It is the rate of change of health indicators, which represents the magnitude of change of health indicators per unit time in a time series.

[0084] For example, obtaining time series data from nursing data collection. Internal health indicator data, such as a patient's intraocular pressure data. mmHg, corresponding to days 0, 7, 14, 21, and 28 respectively.

[0085] Calculate the change between adjacent time points and time interval :

[0086] Day 0 to Day 7: mmHg, sky;

[0087] Day 7 to Day 14: mmHg, sky;

[0088] Day 14 to Day 21: mmHg, sky;

[0089] Day 21 to Day 28: mmHg, sky.

[0090] Calculate the rate of change of health indicators :

[0091] Day 0 to Day 7: mmHg / day;

[0092] Day 7 to Day 14: mmHg / day;

[0093] Day 14 to Day 21: mmHg / day;

[0094] Day 21 to Day 28: mmHg / day.

[0095] Calculate the absolute value of the rate of change Find the maximum value and the corresponding time point:

[0096] mmHg / day (this rate applies from day 0 to day 21);

[0097] mmHg / day (rate from day 21 to day 28).

[0098] The results show that the maximum rate of change mmHg / day appears from day 0 to day 21. The trend change point can be day 7 or day 14. For example, the trend of intraocular pressure decrease peaks on day 14 of care and then gradually stabilizes.

[0099] Please see Figure 4 The specific steps for obtaining the local fitting model are as follows:

[0100] The time series is divided into multiple stages based on the trend change points. The health indicator data in each stage is organized and grouped to generate a stage health indicator data table.

[0101] First, trend change points are extracted from the time series of health indicators. For example, the time point with the largest change rate is determined by calculating the absolute value of the change rate of health indicators, and this is taken as the trend change point. Then, the time series is divided into multiple consecutive stages according to the location of the change points. For each stage, the health indicator data within the stage is reorganized according to the time point sequence, and the data is further grouped. For example, the intraocular pressure of a patient changes in the two stages from day 0 to day 7 and from day 7 to day 14, respectively. and A separate time series table of health indicators is generated for each stage. The segmented data facilitates the construction of subsequent local fitting models and the analysis of characteristic trends. When organizing segmented time series data, Microsoft Excel, MATLAB, or the pandas library in Python can be used to group and organize the data. For example, data can be filtered by stage in Excel, or the groupby function of pandas in Python can be used to segment and reconstruct data within a stage. After data grouping, a stage-specific health indicator data table is generated, which facilitates subsequent analysis and modeling.

[0102] Based on the phased health indicator data table, by analyzing the changing patterns of data in each phase and the changing trends of health indicators in the corresponding phase, a local fitting model is obtained through data fitting.

[0103] A local fitting model for health indicator data corresponding to each stage is constructed. Health indicator data are selected as input variables within the defined time series stages. By analyzing the changing patterns of data within each stage, a linear fitting method is used to model the data for each stage. For example, the intraocular pressure change sequence for a certain stage is... The values ​​correspond to day 0, day 7, and day 14, respectively. First, calculate the rate of change between each pair of time points. Then substitute the time points and changes into the fitting formula. The fitting parameters were calculated using the least squares method. and ,in, The target variable (dependent variable) represents the value of a health indicator, such as intraocular pressure, retinal thickness, or corneal morphology parameters, at each time point. It is the independent variable (input variable), representing point-in-time data, such as day 0, day 7, day 14, etc., corresponding to each point in time in the stage time series. It is the slope of the fitted straight line, representing the rate of change of health indicators over time within a period, reflecting the trend of health indicators in the time series (such as rising, falling, or remaining unchanged). The intercept of the fitted line represents the initial value of the health indicator when the time point is zero, reflecting the baseline level of the health indicator. Calculated using the least squares method, the intercept is automatically generated by the fitted model, reflecting the basic health status within the stage. Finally, a linear fitting model for each stage is obtained, used for subsequent prediction and analysis of stage data. Model construction can be accomplished using MATLAB's CurveFitting tool. The specific process involves first importing the stage-specific health indicator data, selecting the fitting model type (such as a linear or multinomial model), then running the fitting algorithm to output model parameters, and finally verifying the fitting accuracy and saving the model results. Through the automatic calculation functions of MATLAB or Python tools, local fitting models for multiple stages can be quickly constructed and model parameters obtained, ultimately enabling quantitative analysis of the data change trends for each stage.

[0104] Please see Figure 5 The specific steps for obtaining the segmented trend prediction results are as follows:

[0105] Based on the data in the local fitting model, the key parameters in the fitting model at each stage are extracted, and the direction of change of health indicators is analyzed in combination with the fitting results. The trend of change is then segmented and organized to generate the analysis results of the current direction of change of health indicators.

[0106] First, the results of the local fitting model for each stage are extracted. For example, for intraocular pressure data at a certain stage, a linear fitting model is used. The slope parameter was calculated. , where the slope It directly reflects the direction of change in health indicators during this stage, when At that time, health indicators showed an increasing trend. At that time, health indicators showed a downward trend, combined with the model's fitting error (such as mean squared error). To determine the reliability of the model, the slope and error parameters are first obtained through the fitting results of MATLAB tools, or the corresponding slope value is extracted through the fitting results of the scipy library in Python. For example, the fitting result for the intraocular pressure data in the first stage is... The MSE was 0.02, indicating that the intraocular pressure was decreasing during this stage. A report on the direction of change of health indicators was generated by combining the fitting results of all stages.

[0107] Based on the analysis of the current trend of health indicator changes, the following formula is used:

[0108] ;

[0109] Calculate the future Health indicator values ​​at each time point The segmented trend prediction results are obtained.

[0110] in, These are health indicator values ​​at the current point in time, obtained directly through health data collection devices. For example, intraocular pressure is measured using a non-contact tonometer, retinal thickness is obtained using a retinal imaging system, and corneal morphology parameters are obtained using a corneal topography system. This is the slope of the linear change, representing the rate and direction of change of health indicators within the current time period, reflecting the linear trend of health indicators. For example, a positive value indicates an increase, and a negative value indicates a decrease. The coefficient of the linear term is calculated through a local fitting model, for example, using the scipy library in Python to fit data within a period (e.g., ...). ), directly extract , These are quadratic coefficients, representing the acceleration or deceleration of changes in health indicators. They reflect the non-linear trends of health indicators, such as the degree to which a declining trend in health indicators slows or accelerates. The coefficients of the quadratic term are obtained by fitting time-series data with a quadratic polynomial. For example, the `numpy.polyfit` function in Python can be used to fit data within a period and extract the quadratic coefficients. It is a time interval, representing the length of time between the current time point and the predicted time point.

[0111] If it is predicted that a patient will be discharged on the 14th day after treatment (i.e., the day of treatment) The intraocular pressure (IOP) values ​​at several time points, assuming the IOP on day 7 is known to be... mmHg; linear change slope: ; Quadratic variation coefficient: Time interval: .

[0112] Substitute the formula into the calculation:

[0113] ;

[0114] Calculate the linearly changing part:

[0115] ;

[0116] Calculate the quadratic change part:

[0117] ;

[0118] Comprehensive calculation of predicted values:

[0119] ;

[0120] The results indicate that the patient's intraocular pressure was projected to be 18.195 mmHg on day 14 after treatment. The formula, by combining linear and quadratic trends, can more accurately describe the trajectory of health indicators at future points in time. It is particularly suitable for scenarios with non-linear trends, such as intraocular pressure that may decrease rapidly in the early stages of treatment and then level off later. The introduction of the formula effectively solves the problem of error accumulation that may result from purely linear prediction, providing a more scientific calculation method for the segmented trend prediction of health indicators.

[0121] Please see Figure 6 The specific steps for obtaining the curvature values ​​of discrete points in the fitted curve are as follows:

[0122] The segmented trend prediction results are supplemented with the time points and corresponding health indicator values ​​of each data segment through linear interpolation, and then converted into a continuous change curve to generate a continuous change dataset of health indicators.

[0123] First, import the segmented data into a data table in chronological order, ensuring that the time points and corresponding health indicator values ​​for each segment are completely recorded. Then, use linear interpolation to interpolate the data between time points. Extract the start and end coordinates of each time period, input these coordinates, and perform linear interpolation. The specific operation of linear interpolation is as follows: select the start and end coordinates, insert intermediate values ​​in segments according to the proportional relationship within the time interval, and generate a complete continuous change dataset. Then, input all segmented fitted data into a data visualization tool (such as MATLAB or Python's matplotlib library) to plot curves, ensuring that each segment of data is smoothly connected on the continuous change curve. Finally, generate a continuous curve that reflects the changes of health indicators over time for subsequent change characteristic analysis.

[0124] Based on a dataset of continuous changes in health indicators, the following formula is used:

[0125] ;

[0126] Calculate the curvature of the target discrete points This yields the curvature values ​​at discrete points;

[0127] in, This represents the slope of the fitted curve at the target position, reflecting the rate of change of the curve. This represents the acceleration of the fitted curve at the target position, reflecting the change in the rate of change of the curve. Used to normalize curvature values ​​to eliminate differences in the magnitude of curvature between different discrete points. and The parameters of the fitted equation can be obtained by calculating the fitted curve equation, and the expression can be obtained by taking the first and second derivatives of the fitted equation.

[0128] Assume the curve equation obtained by fitting the collected health indicator data is:

[0129] ;

[0130] Find the first derivative:

[0131] ;

[0132] Find the second derivative:

[0133] ;

[0134] exist Calculation at point:

[0135]

[0136] ;

[0137] Calculate curvature :

[0138] ;

[0139] This result indicates that, The curvature value at that point is 0.3755.

[0140] Please see Figure 7 The specific steps for obtaining the key trend node identification results are as follows:

[0141] The curvature values ​​of discrete points are compared point by point with a preset curvature threshold. The comparison determines whether the curvature at each time point exceeds the threshold, and the data is recorded to generate curvature threshold comparison results.

[0142] Based on the calculated curvature values At the point of time In specific scenarios, a preset curvature threshold is selected. The calculated value is compared with the threshold point by point. When the curvature value is greater than the threshold, it is judged as a point of drastic curvature change. , indicating a point in time The curve shows a significant change in curvature. Based on the comparison results, this point is marked as a candidate key trend point that needs further analysis. The curvature threshold comparison results show that the curvature value at this time point has exceeded the preset threshold range.

[0143] Based on the curvature threshold comparison results, the time points exceeding the curvature threshold and their corresponding curvature values ​​are extracted. By marking the target time points as key trend nodes and recording the marking information, the key trend node identification results are established.

[0144] Mark the time points corresponding to curvature values ​​exceeding the curvature threshold as key trend nodes, and combine them with the previously compared time points. curvature value A labeled dataset is created to record the curvature change at that time point, and the time... and curvature value Write the key trend nodes into a list. This marking process forms a clear key trend node identification result and generates a complete node list, recording all points marked beyond the threshold and their curvature values.

[0145] Please see Figure 8 The specific steps for obtaining reference information on eye care are as follows:

[0146] Based on the key trend node identification results, all key trend nodes in the fitted curve are collected, including the time point and curvature value corresponding to the node. By sorting the curvature values ​​and classifying them into multiple node types according to grouping rules, the classification results of key trend nodes are generated.

[0147] First, all key trend nodes in the fitted curve are collected, including the corresponding time point and curvature value. The curvature values ​​are sorted from high to low and classified according to the set grouping rules (e.g., high curvature interval, medium curvature interval, and low curvature interval). The specific operations include: reading the curvature value array of the nodes, calculating the quantiles of the array, setting the grouping thresholds to the 75th and 25th quantiles of the curvature values, classifying nodes above the 75th quantile as high curvature nodes, nodes below the 25th quantile as low curvature nodes, and the rest as medium curvature nodes. The classification results of the key trend nodes are obtained through the grouping operation. The classification results are then reorganized according to the time point for further analysis of distribution patterns.

[0148] Based on the classification results of key trend nodes, the time point and curvature value of each category node are extracted, the distribution density of nodes on the time axis and the time span of density regions are analyzed, the changing trend of the curve and the dense node segments are marked, and reference information for ophthalmic care is obtained.

[0149] Based on the classification results, the distribution patterns of nodes in each category are analyzed. Combining specific parameters such as intraocular pressure (IOP), retinal thickness, and corneal morphology, the time points of IOP changes and corresponding curvature values ​​are first extracted for nodes with high curvature. A density distribution map of IOP changes over time is generated. Kernel density estimation is used to analyze IOP changes, marking the time intervals where densely distributed high IOP nodes are located. For example, the density peak is between 6:00 AM and 10:00 AM, corresponding to the high-risk period when IOP is above 30 mmHg. Subsequently, the same density analysis is performed on retinal thickness changes, extracting and marking time points where retinal thickness changes drastically. For example, time points where retinal thickness changes by more than 50 μm are designated as key trend nodes. Combined with corneal morphology parameters, time points where corneal curvature changes drastically are extracted. For example, the rapid change segment where the central corneal thickness decreases from 520 μm to 480 μm is detected early. By combining the node distribution results of the three types of parameters, the interaction and trend correlation on the time axis are analyzed, and finally, a report on the distribution patterns of key parameters is generated for intelligent auxiliary decision-making in ophthalmic care.

[0150] The above are merely preferred embodiments of the present invention and are not intended to limit the present invention in any other way. Any person skilled in the art may make changes or modifications to the above-disclosed technical content to create equivalent embodiments that can be applied to other fields. However, any simple modifications, equivalent changes, and modifications made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention shall still fall within the protection scope of the present invention.

Claims

1. An intelligent auxiliary system for ophthalmic care, characterized in that, The system includes: The Health Indicator Distribution Difference Analysis Module acquires health indicator data of ophthalmology patients before and after nursing care, compares the differences between the health indicator data of ophthalmology patients before and after nursing care, and generates health indicator transfer characteristic analysis results. The dynamic trend prediction and segmented modeling module determines the trend of health indicator data of ophthalmology patients from before to after nursing care based on the analysis results of the health indicator transfer characteristics. It extracts the trend change points from the data and constructs a local fitting model based on the trend change points of each stage. It analyzes the direction of change of health indicator data and predicts the future characteristic changes of health data in each stage based on the direction of change, generating segmented trend prediction results. The specific steps for obtaining the local fitting model are as follows: The time series is divided into multiple stages based on the trend change points. Health indicator data in each stage are organized and grouped to generate a stage-specific health indicator data table. Based on the aforementioned phased health indicator data table, by analyzing the changing patterns of data within each phase and the changing trends of health indicators within the corresponding phase, a local fitting model is obtained through data fitting. The health data curvature feature extraction module converts the segmented trend prediction results into a continuous change curve and fits it. It calculates the curvature value of discrete points in the fitted change curve and compares it with a preset curvature threshold. Curvature values ​​exceeding the curvature threshold are marked as key trend nodes, and key trend node identification results are generated. The health indicator dynamic update module classifies the key trend nodes in the key trend node identification results, analyzes the distribution pattern of the key trend nodes in the fitted change curve based on the classification results, and generates reference information for ophthalmic care.

2. The intelligent auxiliary system for ophthalmic care according to claim 1, characterized in that, The specific steps for obtaining the health indicator transfer characteristic analysis results are as follows: Based on patient health index data before and after nursing care, intraocular pressure, retinal thickness and corneal morphology parameters were obtained, and the data from multiple dimensions were normalized to obtain an integrated health index data table. Based on the integrated health indicator data table, the formula is used: ； Calculate the first Difference values ​​of health indicators This yields data on the differences in indicators; in, It is the first The values ​​of several health indicators after nursing care. It is the first The values ​​of the following health indicators before nursing care; Based on the difference data of the indicators, the maximum value, minimum value and concentrated distribution range of each type of indicator are statistically analyzed. The positive and negative change range of the health indicators are determined by combining the upper and lower limits of the difference values. The results of the health indicator transfer characteristic analysis are generated based on the statistical information.

3. The intelligent auxiliary system for ophthalmic care according to claim 2, characterized in that, The specific steps for extracting trend change points are as follows: Based on the analysis results of the health indicator transfer characteristics, by organizing the intraocular pressure, retinal thickness and corneal morphology parameter data before and after nursing care, a data sequence is generated according to the time point order. The rate and direction of change of adjacent time points in the time sequence are analyzed one by one to generate the analysis results of the time series change trend of health indicators. Based on the time series trend analysis results of the aforementioned health indicators, the following formula is used: ； Calculate the maximum absolute value in the time series of health indicators This allows us to identify the points where the trend changes. in, The variable representing the maximum value of the objective function. It is a point-in-time variable. Indicates the time point at which nursing care begins. and the time point at which the care ends , It is the change in health indicator values. It represents the interval between two adjacent time points in a time series.

4. The intelligent auxiliary system for ophthalmic care according to claim 1, characterized in that, The specific steps for obtaining the segmented trend prediction results are as follows: Based on the data in the local fitting model, by extracting the key parameters in the fitting model at each stage, and combining the fitting results to analyze the direction of change of health indicators, the trend of change is segmented and organized to generate the analysis results of the current direction of change of health indicators. Based on the analysis results of the current health indicator changes, the following formula is used: ； Calculate the future Health indicator values ​​at each time point The segmented trend prediction results are obtained. in, These are health indicator values ​​at the current point in time. It is a linear slope. It is the coefficient of second variation. This indicates the length of time between the current time point and the predicted time point.

5. The intelligent auxiliary system for ophthalmic care according to claim 4, characterized in that, The specific steps for obtaining the curvature values ​​of discrete points in the fitted curve are as follows: The segmented trend prediction results are supplemented with the time points and corresponding health indicator values ​​of each data segment through linear interpolation, and then converted into a continuous change curve to generate a continuous change dataset of health indicators. Based on the aforementioned dataset of continuous changes in health indicators, the following formula is used: ； Calculate the curvature of the target discrete points This yields the curvature values ​​at discrete points; in, This represents the slope of the fitted curve at the target position. This represents the acceleration of the fitted curve at the target position. Used to normalize curvature values.

6. The intelligent auxiliary system for ophthalmic care according to claim 5, characterized in that, The specific steps for obtaining the key trend node identification results are as follows: The curvature values ​​of the discrete points are compared point by point with a preset curvature threshold. The comparison determines whether the curvature at each time point exceeds the threshold, and the data is recorded to generate a curvature threshold comparison result. Based on the curvature threshold comparison results, time points exceeding the curvature threshold and corresponding curvature value data are extracted. By marking target time points as key trend nodes and recording the marking information, key trend node identification results are established.

7. The intelligent auxiliary system for ophthalmic care according to claim 6, characterized in that, The specific steps for obtaining reference information on eye care are as follows: Based on the key trend node identification results, all key trend nodes in the fitted curve are collected, including the time point and curvature value corresponding to the node. By sorting the curvature values ​​and classifying them into multiple node types according to grouping rules, the classification results of key trend nodes are generated. Based on the classification results of the key trend nodes, the time point and curvature value of each category node are extracted, the distribution density of nodes on the time axis and the time span of density regions are analyzed, the changing trend of the curve and the dense node segments are marked, and reference information for ophthalmic care is obtained.

Images