Urolithiasis examination image recognition and data visualization analysis and early warning system

Abstract

The present application relates to the technical field of medical data processing, and discloses a urinary calculus examination image recognition and data visualization analysis and early warning system, which comprises: an examination data analysis module that extracts calculus disease-related sequence data from user urinary calculus examination images and blood and urine report synchronous sequence data, covering position distribution, morphological characteristics and biochemical indicators; a data visualization module that converts the above data into calculus volume dynamic change simulation graphs, biochemical indicator abnormal value correlation graphs and calculus position dynamic mapping graphs, facilitating user queries; a dynamic threshold reasoning module that performs time-space alignment and cross-local dimension abnormality fusion reasoning on the visualization results to obtain user personalized multi-dimensional hierarchical early warning thresholds and clinical decision simulation paths; and an early warning and report integration module that determines the current early warning level based on the latest indicators and early warning thresholds, and generates a comprehensive diagnosis and treatment report in combination with the visualization curve data and the decision path, facilitating patients and doctors to understand the disease and develop treatment plans.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of medical data processing, in particular to a urinary calculus examination image recognition and data visualization analysis and early warning system. BACKGROUND

[0002] In the field of medical health, urinary calculus is a common urological disease that seriously affects the quality of life of patients. Accurate diagnosis of urinary calculus and timely adoption of effective treatment measures are crucial. Urinary calculus examination images (such as ultrasound, CT images, etc.) and blood and urine test reports contain rich disease information, which plays a key role in helping doctors judge the condition of the calculus and develop treatment plans. Traditional urinary calculus diagnosis mainly relies on doctors' naked-eye observation of examination images and manual analysis of test report data. However, with the development of medical technology, the amount of data generated is increasingly large and complex, and relying solely on manual analysis is not only inefficient, but also susceptible to the influence of doctors' subjective experience and fatigue, etc., making it difficult to ensure the accuracy and consistency of the diagnosis results. In recent years, with the continuous progress of artificial intelligence and data visualization technology, applying these technologies to the diagnosis and analysis of urinary calculus has become a research hotspot. Image recognition technology can quickly and accurately extract and analyze the features of urinary calculus examination images, helping doctors more clearly understand the location, shape, and other information of the calculus; data visualization technology can present complex examination data in an intuitive and easy-to-understand graphical manner, facilitating doctors and patients to understand; and the early warning function based on data analysis can timely detect potential changes in the disease, providing strong support for clinical decision-making. By constructing a urinary calculus examination image recognition and data visualization analysis and early warning system, it is expected to achieve comprehensive and in-depth analysis of urinary calculus-related data, improve diagnosis efficiency and accuracy, and provide more personalized and precise medical services for patients. This not only helps to improve the treatment effect of patients, but also promotes the development of urological disease diagnosis technology towards intelligence and precision, and has broad application prospects and important clinical significance.

[0003] The existing technology has many defects in the diagnosis and analysis of urinary calculus. Firstly, it is difficult to effectively extract key calculus disease-related sequence data from the synchronous sequence data of urinary calculus examination images and blood and urine test reports, making it impossible to achieve comprehensive analysis of time sequence alignment and leading to incomplete understanding of calculus disease characteristics. Secondly, there is a lack of effective visualization means to convert complex calculus disease-related sequence data into intuitive and easy-to-understand graphics, such as the inability to generate dynamic change simulation graphs of calculus volume, abnormal value correlation graphs of biochemical indicators, and dynamic mapping graphs of calculus location, making it difficult for doctors and patients to clearly understand the changes in the disease. Furthermore, it is also impossible to perform deep multi-dimensional abnormal fusion reasoning on multi-dimensional visualized data, making it impossible to obtain personalized multi-dimensional grading early warning thresholds and clinical decision-making simulation paths, resulting in inaccurate early warning and lack of targeted clinical guidance.

[0004] Therefore, this invention proposes a system for image recognition, data visualization analysis, and early warning of urinary stones. Summary of the Invention

[0005] This invention provides a system for image recognition, data visualization analysis, and early warning of urinary stones. The system uses a data analysis module to accurately extract time-aligned data on stone location distribution, morphology, and pathological biochemical indicators from synchronized sequence data of urinary stone examination images and blood and urine test reports, laying a comprehensive foundation for subsequent analysis. The data visualization module then transforms this data into a simulation diagram of dynamic changes in stone volume, a correlation map of abnormal values ​​in biochemical indicators, and a dynamic mapping map of stone location, providing intuitive graphics for users to easily access information about their condition. A dynamic threshold inference module performs spatiotemporal alignment and cross-local dimension anomaly fusion inference on these visualization results to derive personalized, multi-dimensional, graded early warning thresholds and clinical decision simulation paths, considering individual differences to aid in precision treatment. Finally, the early warning and report integration module determines the current early warning level based on the latest indicators and early warning thresholds, and generates a comprehensive diagnosis and treatment report by combining visualized curve data and the clinical decision simulation path. This provides patients with comprehensive information about their condition and treatment guidance, improving the scientific rigor and effectiveness of data processing, condition analysis, and decision-making in the diagnosis and treatment of urinary stones, making it easier for patients and doctors to understand their condition and develop treatment plans.

[0006] This invention provides a system for image recognition, data visualization analysis, and early warning of urinary calculi examination, comprising:

[0007] The examination data parsing module is used to extract stone-related sequence data from the synchronized sequence data of the user's urinary stone examination images and blood and urine test reports, including time-aligned urinary stone location distribution sequence features, morphological sequence features, and stone-related biochemical sequence indicators;

[0008] The data visualization module is used to convert sequence data related to the occurrence of kidney stones into a simulation diagram of dynamic changes in kidney stone volume, a correlation map of abnormal values ​​of biochemical indicators, and a dynamic mapping map of kidney stone location, which serve as a comprehensive examination visualization result for users to query.

[0009] The dynamic threshold reasoning module is used to perform spatiotemporal alignment and cross-local dimension anomaly fusion reasoning on the simulation map of dynamic changes in stone volume, the correlation map of abnormal values ​​of biochemical indicators, and the dynamic mapping map of stone location, so as to obtain the user's personalized multi-dimensional graded early warning threshold and clinical decision simulation path.

[0010] An early warning and report integration module is configured to determine a current early warning level of a user based on the latest related indicators of the user's stone disease and the user's personalized multi-dimensional early warning threshold, and to generate a comprehensive diagnosis and treatment report of urinary stones in combination with visual curve data generated based on the comprehensive examination visual results and a clinical decision simulation path.

[0011] Optionally, the examination data analysis module comprises:

[0012] An image recognition submodule is configured to extract urinary stone location data and morphological data of each urinary stone examination image in the user's synchronous sequence data based on all the local images of urinary stones in each urinary stone examination image in the user's synchronous sequence data.

[0013] A data serialization submodule is configured to generate position distribution sequence features and morphological sequence features of urinary stones based on the urinary stone location data and morphological data of all the urinary stone examination images in the user's synchronous sequence data.

[0014] A report serialization analysis submodule is configured to generate stone disease-related biochemical sequence indicators based on the stone disease-related biochemical indicators in all the blood and urine examination reports in the user's synchronous sequence data.

[0015] A time sequence alignment submodule is configured to perform time sequence alignment on the position distribution sequence features, morphological sequence features, and stone disease-related biochemical sequence indicators of urinary stones to obtain stone disease-related sequence data.

[0016] Optionally, the data visualization module comprises:

[0017] A urinary stone dynamic simulation submodule is configured to generate a stone volume dynamic change simulation map based on the morphological sequence features of urinary stones in the stone disease-related sequence data, and to generate a stone position dynamic mapping map based on the position distribution sequence features of urinary stones in the stone disease-related sequence data.

[0018] A first abnormal indicator extraction submodule is configured to extract first abnormal biochemical indicators and first sequence abnormal indicators from the position distribution sequence features and morphological sequence features of urinary stones.

[0019] A second abnormal indicator extraction submodule is configured to extract second abnormal biochemical indicators and second sequence abnormal indicators from the stone disease-related biochemical sequence indicators.

[0020] An abnormal indicator mapping submodule is configured to generate a biochemical indicator abnormal value correlation map based on the first abnormal biochemical indicators, the first sequence abnormal indicators, the second abnormal biochemical indicators, and the second sequence abnormal indicators, and to provide the map as a comprehensive examination visual result for user query.

[0021] Optionally, the dynamic threshold reasoning module comprises:

[0022] The cross-local-dimension anomaly fusion reasoning submodule is configured to perform spatio-temporal alignment and cross-local-dimension anomaly fusion reasoning on the stone volume dynamic change simulation graph, the biochemical index abnormal value correlation graph and the stone position dynamic mapping graph to obtain a plurality of cross-local-dimension anomaly fusion reasoning information and a cross-local-dimension anomaly fusion reasoning context of the patient.

[0023] The dynamic threshold reasoning submodule is configured to determine the personalized multi-dimensional grading early warning threshold and the clinical decision simulation path of the user based on the standard cross-local-dimension anomaly fusion reasoning context and the plurality of cross-local-dimension anomaly fusion reasoning information and the cross-local-dimension anomaly fusion reasoning context of the patient.

[0024] Optionally, the cross-local-dimension anomaly fusion reasoning submodule comprises:

[0025] The abnormal data extraction unit is configured to extract first urinary stone diagnosis related abnormal data from the stone volume dynamic change simulation graph, and extract second urinary stone diagnosis related abnormal data from the stone position dynamic mapping graph.

[0026] The time span division determination unit is configured to determine a local time span division sequence based on the same abnormal value interval sequence in the first urinary stone diagnosis related abnormal data and the same abnormal value interval sequence in the second urinary stone diagnosis related abnormal data.

[0027] The synchronous division unit is configured to perform synchronous division on the first urinary stone diagnosis related abnormal data, the second urinary stone diagnosis related abnormal data and the biochemical index abnormal value correlation graph based on the local time span division sequence to obtain multi-dimensional local abnormal data.

[0028] The cross-dimension fusion reasoning unit is configured to perform cross-dimension fusion reasoning on the multi-dimensional local abnormal data to obtain a plurality of cross-local-dimension anomaly fusion reasoning information and a cross-local-dimension anomaly fusion reasoning context of the patient.

[0029] Optionally, the time span division determination unit comprises:

[0030] The cause correlation coefficient acquisition subunit is configured to acquire a urinary stone cause correlation coefficient between each index data in the first urinary stone diagnosis related abnormal data and each index data in the second urinary stone diagnosis related abnormal data.

[0031] The time span sequence generation unit is used to determine the locally segmented time span sequence based on the correlation coefficient of urinary stone etiology between each indicator data in the first urinary stone diagnosis-related abnormal data, the urinary stone etiology correlation coefficient between each indicator data in the second urinary stone diagnosis-related abnormal data, the interval sequence of the same outlier in the first urinary stone diagnosis-related abnormal data, and the interval sequence of the same outlier in the second urinary stone diagnosis-related abnormal data.

[0032] ;

[0033] In the formula, To locally divide the time span sequence, This represents the total number of all indicator data types in the abnormal data related to the diagnosis of urinary calculi. This represents the total number of all types of indicators in the abnormal data related to the diagnosis of second urinary tract stones. The first abnormal data related to the diagnosis of urinary calculi The first indicator data and the second abnormal data related to the diagnosis of urinary stones The correlation coefficient between the data of these indicators regarding the causes of urinary stones To find the average function, The first abnormal data related to the diagnosis of urinary calculi The first value in the sequence of outlier intervals for the same type of indicator data. The first abnormal data related to the diagnosis of urinary calculi The first value in the sequence of outlier intervals for the same type of indicator data. The first abnormal data related to the diagnosis of urinary calculi The second value in the sequence of outlier intervals for the same type of indicator data. The first abnormal data related to the diagnosis of urinary calculi The second value in the sequence of outlier intervals for the same type of indicator data. The first abnormal data related to the diagnosis of urinary calculi The first outlier in the interval sequence of the same type of index data A number, The first abnormal data related to the diagnosis of urinary calculi The first outlier in the interval sequence of the same type of index data A number of values, and The value is equal to the first abnormal data related to the diagnosis of urinary stones. The total number of values ​​in the same outlier interval sequence of the first type of indicator data and the first outlier in the second type of urinary calculi diagnosis-related outlier data. The minimum value among the total number of values ​​in the interval sequence of the same outlier of a certain indicator data.

[0034] Optionally, the cross-dimension fusion reasoning unit comprises:

[0035] The cause association rule obtaining subunit is configured to obtain all stone cause association rules.

[0036] The compliance degree determining subunit is configured to perform cross-dimension combination on the multi-dimensional local abnormal data based on all stone cause association rules, to obtain a plurality of cross-dimension combination abnormal data, and to determine a compliance degree of each cross-dimension combination abnormal data to a corresponding stone cause association rule.

[0037] The fusion reasoning subunit is configured to obtain a plurality of cross-local-dimension abnormal fusion reasoning information of the patient based on all stone cause association rules and all cross-dimension combination abnormal data.

[0038] The reasoning context generating subunit is configured to generate a cross-local-dimension abnormal fusion reasoning context based on all cross-dimension combination abnormal data, corresponding all stone cause association rules and corresponding compliance degrees, and in combination with corresponding all cross-local-dimension abnormal fusion reasoning information.

[0039] Optionally, the dynamic threshold reasoning sub-module comprises:

[0040] The macro context deviation analysis unit is configured to determine an abnormal fusion reasoning macro deviation degree based on the standard cross-local-dimension abnormal fusion reasoning context and the cross-local-dimension abnormal fusion reasoning context.

[0041] The hierarchical early warning threshold determining unit is configured to determine a personalized multi-dimensional hierarchical early warning threshold of the user based on all cross-local-dimension abnormal fusion reasoning information of the patient and the abnormal fusion reasoning macro deviation degree.

[0042] The clinical decision path simulation unit is configured to determine a clinical decision simulation path of the user based on the cross-local-dimension abnormal fusion reasoning context.

[0043] Optionally, the hierarchical early warning threshold determining unit comprises:

[0044] The basic threshold determining subunit is configured to determine a multi-dimensional hierarchical early warning basic threshold based on all cross-local-dimension abnormal fusion reasoning information of the patient.

[0045] The hierarchical threshold determining subunit is configured to correct the multi-dimensional hierarchical early warning basic threshold based on the abnormal fusion reasoning macro deviation degree, to obtain the personalized multi-dimensional hierarchical early warning threshold of the user.

[0046] Optionally, the early warning and report integration module comprises:

[0047] The early warning level diagnosis submodule is used for determining the current early warning level of the user based on the latest related indicators of the stone disease of the user and the personalized multi-dimensional early warning threshold of the user, and issuing an early warning signal based on the current early warning level;

[0048] The intervention suggestion generation submodule is used for determining the subsequent intervention priority suggestion of the user based on the clinical decision simulation path;

[0049] The report integration submodule is used for generating a plurality of visual curve data based on the comprehensive examination visual result, and integrating the current early warning level of the user, the plurality of visual curve data and the subsequent intervention priority suggestion of the user to generate a comprehensive diagnosis and treatment report of urinary calculi.

[0050] The present application has the following advantages over the prior art: the examination data analysis module accurately extracts the related data such as the stone position distribution, morphology and biochemical sequence indicators from the urinary calculi examination image and blood and urine examination report synchronous sequence data, laying a comprehensive foundation for subsequent analysis; then the data visualization module converts them into a stone volume dynamic change simulation diagram, a biochemical indicator abnormal value correlation diagram and a stone position dynamic mapping diagram to facilitate user query of the disease condition; then the dynamic threshold reasoning module performs spatio-temporal alignment and cross-local dimension abnormality fusion reasoning on these visual results to obtain a personalized multi-dimensional early warning threshold and a clinical decision simulation path, considering individual differences to assist precise treatment; finally, the early warning and report integration module determines the current early warning level according to the latest indicators and the early warning threshold, and generates a comprehensive diagnosis and treatment report in combination with the visual curve data and the clinical decision simulation path to provide comprehensive disease condition information and treatment guidance for the patient, and overall improves the scientificity and effectiveness of data processing, disease condition analysis and decision-making in the urinary calculi diagnosis and treatment process, facilitating the patient and the doctor to understand the disease condition and develop a treatment plan.

[0051] Other features and advantages of the present application will be set forth in the following description, and in part will become apparent to those skilled in the art from the description, or can be learned by practice of the present application. The objects and other advantages of the present application can be realized and attained by the structure particularly pointed out in the written description and claims hereof.

[0052] The technical solutions of the present application will be further described in detail below with the help of the drawings and examples. BRIEF DESCRIPTION OF DRAWINGS

[0053] The accompanying drawings are included to provide a further understanding of the present application, and constitute a part of the specification, illustrate embodiments of the present application and explain the present application together with the embodiments, and do not constitute a limitation on the present application. In the drawings:

[0054] Figure 1 It is a urinary calculi examination image recognition and data visualization analysis and early warning reminding system architecture schematic diagram in the embodiment of the present application;

[0055] Figure 2 An example diagram of cross-local-dimension anomaly fusion reasoning context in an embodiment of the present application. DETAILED DESCRIPTION

[0056] The preferred embodiments of the present application will be described herein below with reference to the accompanying drawings, in which it is understood that the preferred embodiments described herein are merely intended to illustrate and explain the present application, and are not intended to limit the present application.

[0057] Reference Figure 1 The present application provides a urinary stone examination image recognition and data visualization analysis and early warning system, comprising:

[0058] An examination data analysis module is configured to extract stone disease-related sequence data from the user's urinary stone examination image and the blood and urine examination report synchronous sequence data, including time-series aligned urinary stone position distribution sequence features, morphological sequence features, and stone disease-related biochemical sequence indexes. The module is responsible for extracting sequence data closely related to stone disease from the user's urinary stone examination image and the blood and urine examination report synchronous sequence data corresponding thereto. For example, urinary stone examination images taken at different time points and blood and urine examination report data at the same period. By processing these data, time-series aligned urinary stone position distribution sequence features (such as the change of stones in different positions of the urinary system over time), morphological sequence features (such as the change of stone size and shape over time), and stone disease-related biochemical sequence indexes (such as the fluctuation of components related to stone formation in blood or urine over time) are obtained, providing a comprehensive and orderly data basis for in-depth analysis of subsequent system modules.

[0059] A data visualization module is configured to convert the stone disease-related sequence data into a stone volume dynamic change simulation diagram, a biochemical index abnormal value correlation map, and a stone position dynamic mapping diagram as a comprehensive examination visualization result for user query. This module converts the stone disease-related sequence data extracted by the examination data analysis module into intuitive and easy-to-understand graphics. The stone volume dynamic change simulation diagram is generated based on the morphological sequence features of urinary stones, allowing users to clearly see the change of stone volume over time. The stone position dynamic mapping diagram is generated based on the position distribution sequence features, visually displaying the change of stone position in the urinary system. At the same time, abnormal indexes are extracted from various sequence features and indexes to generate a biochemical index abnormal value correlation map, revealing the relationship between various biochemical index abnormalities. These visualization results facilitate user query and quick understanding of disease trend.

[0060] A dynamic threshold reasoning module is used to perform spatio-temporal alignment and cross-local dimension anomaly fusion reasoning on the stone volume dynamic change simulation graph, biochemical index abnormal value correlation graph, and stone position dynamic mapping graph, to obtain personalized multi-dimensional graded early warning thresholds and clinical decision simulation paths for the user. For the stone volume dynamic change simulation graph, biochemical index abnormal value correlation graph, and stone position dynamic mapping graph generated by the data visualization module, the module performs spatio-temporal alignment and cross-local dimension anomaly fusion reasoning. For example, the abnormal information displayed in each graph at different time points is considered comprehensively, and the correlation between abnormalities in different dimensions (such as position, shape, and biochemical index) is analyzed. Through this reasoning, personalized multi-dimensional graded early warning thresholds (such as limit values for different risk levels according to the development of the stone) and clinical decision simulation paths (simulating reasonable diagnosis and treatment steps for the user's condition) are obtained, providing an important reference for precision medicine.

[0061] An early warning and report integration module is used to determine the current early warning level of the user based on the latest related indicators of the user's stone disease and the user's personalized multi-dimensional graded early warning thresholds, and to generate a comprehensive urinary stone diagnosis and treatment report in combination with the visualization curve data generated based on the comprehensive examination visualization results and the clinical decision simulation path. This module compares the latest related indicators of the user's stone disease with the personalized multi-dimensional graded early warning thresholds obtained by the dynamic threshold reasoning module to determine the current early warning level, such as low risk, medium risk, or high risk. At the same time, in combination with the visualization curve data (such as stone volume change curve, biochemical index fluctuation curve, etc.) generated based on the comprehensive examination visualization results and the clinical decision simulation path, a comprehensive urinary stone diagnosis and treatment report is generated. This report can help patients and doctors understand the current condition and development trend of the disease, so as to develop a scientific and reasonable treatment plan.

[0062] In an alternative embodiment, the examination data analysis module includes:

[0063] An image recognition sub-module is used to extract the urinary stone position data and morphological data of each urinary stone examination image in the user's synchronous sequence data based on all urinary stone local images in each urinary stone examination image in the user's synchronous sequence data. It works based on all urinary stone local images in each urinary stone examination image in the user's synchronous sequence data. For example, in ultrasound or CT examination images, the local image area corresponding to each stone is identified. Then, the urinary stone position data (such as the specific coordinate position of the stone in the urinary system) and morphological data (such as the size, shape, and contour of the stone) are extracted from these local images. These data can provide detailed information about the spatial position and morphological characteristics of each urinary stone in the examination image, providing original information for subsequent analysis of the changes in the stone at different times.

[0064] data serialization submodule, configured to generate position distribution sequence features and morphological sequence features of urinary calculi based on urinary calculi position data and morphological data of all urinary calculi examination images in the user's synchronous sequence data; the module processes the urinary calculi position data and morphological data of all urinary calculi examination images extracted by the image recognition submodule. The position data of examination images at different time points are integrated to generate position distribution sequence features of urinary calculi, reflecting the regularity of the position change of the calculi in a period of time; similarly, the morphological data are sorted into morphological sequence features, embodying the evolution of the shape of the calculi over time. For example, through these sequence features, it can be seen whether the calculi gradually move in a certain direction, or the shape gradually becomes larger or smaller, etc., providing a basis for comprehensively understanding the development process of the calculi.

[0065] report serialization analysis submodule, configured to generate calculi-related biochemical sequence indicators based on calculi-related biochemical indicators in all portions of blood and urine examination reports in the user's synchronous sequence data; the module focuses on the calculi-related biochemical indicators in all portions of blood and urine examination reports in the user's synchronous sequence data. The indicators are analyzed and sorted to generate calculi-related biochemical sequence indicators. For example, the contents of calcium, oxalic acid, uric acid, etc. in blood, and the indicators such as pH and crystal components in urine are processed to form sequences that change over time. These sequences can reflect the dynamic changes of the biochemical environment related to the occurrence of calculi, which is helpful for analyzing the formation and development mechanism of calculi from the biochemical point of view.

[0066] time sequence alignment submodule, configured to perform time sequence alignment on the position distribution sequence features, morphological sequence features and calculi-related biochemical sequence indicators of urinary calculi to obtain calculi-related sequence data. The position distribution sequence features, morphological sequence features and calculi-related biochemical sequence indicators of urinary calculi generated in the foregoing are aligned in time sequence. Because different types of data all reflect calculi-related information, but the acquisition time may be different, through time sequence alignment, these data are matched with each other in the time dimension, so that complete calculi-related sequence data are obtained. The integrated data can more accurately reflect the correlation in time between the changes of the position and shape of the calculi and the changes of the biochemical indicators, providing an ordered and comprehensive data basis for subsequent comprehensive analysis and disease judgment.

[0067] In an alternative embodiment, the data visualization module comprises:

[0068] The urinary calculus dynamic simulation submodule is used to generate a calculus volume dynamic change simulation diagram based on the morphological sequence characteristics of the urinary calculus in the calculus occurrence related sequence data, and simultaneously generate a calculus position dynamic mapping diagram based on the position distribution sequence characteristics of the urinary calculus in the calculus occurrence related sequence data. The morphological sequence characteristics of the urinary calculus in the calculus occurrence related sequence data are used to generate the calculus volume dynamic change simulation diagram. The morphological sequence characteristics include information about the size and shape of the calculus changing over time. Through analysis and calculation of the information, the change of the calculus volume at different time points is simulated in a graphical manner, so that the user can directly see how the calculus volume grows or shrinks. Meanwhile, the position distribution sequence characteristics of the urinary calculus are used to generate the calculus position dynamic mapping diagram. The characteristics record the change of the position of the calculus in the urinary system over time. Through the mapping diagram, the moving track of the calculus in the body can be clearly presented, helping the user to understand the dynamic change of the position of the calculus. For example, if the morphological sequence characteristics show that the major diameter and the minor diameter of the calculus gradually increase over time, the dynamic simulation submodule can generate a simulation diagram showing that the volume of the calculus gradually increases. If the position distribution sequence characteristics show that the calculus moves from a certain part of the kidney to the ureter, the calculus position dynamic mapping diagram can directly show the moving process.

[0069] The first abnormal index extraction submodule is used to extract a first abnormal biochemical index and a first sequence abnormal index from the position distribution sequence characteristics and the morphological sequence characteristics of the urinary calculus. The submodule focuses on mining abnormal information from the position distribution sequence characteristics and the morphological sequence characteristics of the urinary calculus. The first abnormal biochemical index refers to finding clues that may suggest abnormal biochemical indexes from the sequence related to the physical characteristics of the calculus. For example, if the morphology of the calculus suddenly changes abnormally, it may be related to the change of certain biochemical indexes. The submodule can extract the biochemical indexes that may be related. The first sequence abnormal index is for the position distribution and the morphological sequence itself, and finds abnormal sequence conditions that do not conform to the normal change rule. For example, the position of the calculus should move slowly, but suddenly moves in a large amplitude and jumps. The abnormal position sequence change will be extracted. These abnormal indexes provide key clues for further analysis of the calculus occurrence.

[0070] The second abnormal index extraction submodule is configured to extract a second abnormal biochemical index and a second sequence abnormal index from the stone disease-related biochemical sequence index. The second abnormal biochemical index refers to an index value that is found to be outside the normal range directly from the biochemical sequence index data. For example, the content of a certain component in the blood that is closely related to stone formation is too high or too low. The second sequence abnormal index focuses on the regularity of the biochemical sequence index over time. If there is a situation that does not conform to the conventional change pattern, such as a certain index that has been stable but suddenly fluctuates continuously, the abnormal sequence change will be extracted by this module, providing a basis for analyzing the biochemical mechanism change of stone disease.

[0071] The abnormal index mapping submodule is configured to generate a biochemical index abnormal value correlation graph based on the first abnormal biochemical index, the first sequence abnormal index, the second abnormal biochemical index, and the second sequence abnormal index, as a comprehensive inspection visualization result for user query. This module integrates the first abnormal biochemical index, the first sequence abnormal index, the second abnormal biochemical index, and the second sequence abnormal index obtained by the two preceding abnormal index extraction submodules to generate a biochemical index abnormal value correlation graph. The graph visually displays the mutual relationship between various abnormal indexes, such as which stone position or shape abnormality is closely related to a specific biochemical index abnormality, and the possible internal relationship between different biochemical index abnormalities. As a comprehensive inspection visualization result for user query, doctors or patients can understand the correlation of various abnormal conditions in the stone disease process more comprehensively and deeply by viewing the graph, providing strong support for accurate judgment of the disease and development of a treatment plan.

[0072] In an alternative embodiment, the dynamic threshold reasoning module includes:

[0073] The cross-local dimension abnormality fusion reasoning submodule is configured to perform spatio-temporal alignment and cross-local dimension abnormality fusion reasoning on the stone volume dynamic change simulation graph, the biochemical index abnormal value correlation graph, and the stone position dynamic mapping graph to obtain a plurality of cross-local dimension abnormality fusion reasoning information and a cross-local dimension abnormality fusion reasoning context of the patient (for reference Figure 2); this submodule is responsible for processing three kinds of visualization results: stone volume dynamic change simulation graph, biochemical index abnormal value correlation graph, and stone position dynamic mapping graph. First, temporal and spatial alignment is performed, which means matching data from different dimensions (volume, biochemical index, position) in terms of time and space concepts to ensure consistency in analysis. For example, the change in stone volume at a certain time is matched with the abnormality of biochemical index at the same time and the change in stone position. Then, cross-local-dimension abnormality fusion reasoning is carried out, which comprehensively considers abnormal information from different local dimensions (volume, biochemical index, position in a short period of time) to explore potential relationships between them. For example, sudden increase in stone volume, combined with changes in certain component indicators in the biochemical index abnormal value correlation graph, and changes in position in the stone position dynamic mapping graph, analysis of the possible meaning of these simultaneous abnormalities can represent the patient's condition. Through this reasoning method, multiple cross-local-dimension abnormality fusion reasoning information of the patient is obtained, which is the conclusion obtained by comprehensive analysis of abnormalities in different dimensions. At the same time, the cross-local-dimension abnormality fusion reasoning context is generated, which is like a logical chain showing how various abnormalities are related and influenced each other, providing clear clues for further analysis of the patient's condition.

[0074] The dynamic threshold reasoning submodule is used to determine the user's individualized multi-dimensional grading early warning threshold and clinical decision simulation path based on the standard cross-local-dimension abnormality fusion reasoning context and the patient's multiple cross-local-dimension abnormality fusion reasoning information and cross-local-dimension abnormality fusion reasoning context. This submodule works based on the standard cross-local-dimension abnormality fusion reasoning context and the patient's own cross-local-dimension abnormality fusion reasoning information and context. The standard context is a general pattern established based on a large amount of clinical data or medical knowledge, representing the reasonable relationship and development path between different dimension abnormalities under normal circumstances. Comparing the patient's specific situation with it can reveal the particularity of the patient's condition. For example, if the standard context shows that the increase in stone volume is accompanied by specific biochemical index changes and position movement patterns, but the patient's situation is different, the unique aspects of the patient's condition can be analyzed. Through this comparative analysis, the user's individualized multi-dimensional grading early warning threshold is determined. These thresholds are developed according to the patient's individual situation and are used to divide the severity of the condition or risk level, such as determining whether the patient is at low, medium, or high risk level based on various abnormalities related to the stone. At the same time, based on the patient's cross-local-dimension abnormality fusion reasoning context, the clinical decision simulation path is determined. This path simulates the reasonable treatment steps and decision-making order for the patient's condition, providing an important reference for doctors to develop actual treatment plans and helping doctors provide more accurate medical services to patients.

[0075] In an alternative embodiment, the cross-local-dimension abnormality fusion reasoning submodule includes:

[0076] The abnormal data extraction unit is configured to extract first urinary calculus diagnosis related abnormal data from the calculus volume dynamic change simulation diagram and second urinary calculus diagnosis related abnormal data from the calculus position dynamic mapping diagram. The main task of the unit is to extract the urinary calculus diagnosis related abnormal data from the calculus volume dynamic change simulation diagram and the calculus position dynamic mapping diagram, respectively. In the calculus volume dynamic change simulation diagram, attention can be paid to the case that the calculus volume suddenly and rapidly increases or decreases, and these abnormal change data are marked as the first urinary calculus diagnosis related abnormal data. For example, under normal circumstances, the calculus volume increases slowly, but the volume growth rate is significantly accelerated in a certain period of time, and this abnormal growth data is extracted. Meanwhile, in the calculus position dynamic mapping diagram, if the abnormal moving track of the calculus is found, such as the calculus originally stably located in a certain region of the kidney suddenly and rapidly moves to the ureter, the abnormal position change data is extracted as the second urinary calculus diagnosis related abnormal data. These abnormal data are important basis for subsequent analysis of the condition change.

[0077] The division time span determination unit is configured to determine a local division time span sequence based on the same abnormal value interval sequence in the first urinary calculus diagnosis related abnormal data and the same abnormal value interval sequence in the second urinary calculus diagnosis related abnormal data. The same abnormal value interval sequence, for example, in the first urinary calculus diagnosis related abnormal data, the calculus volume suddenly increases multiple times, and the time interval between these increase cases constitutes a same abnormal value interval sequence. In the second urinary calculus diagnosis related abnormal data, the time interval between similar abnormal moving cases of the calculus also forms a corresponding sequence. By analyzing the two groups of same abnormal value interval sequences, a local division time span sequence is determined. This time span sequence can reflect the distribution rule of the abnormal cases in the time dimension, and provide a time division basis for subsequent synchronous analysis of abnormal data in different dimensions. For example, if it is found that the case of sudden increase of the calculus volume appears every certain period of time, and there is a certain correlation with the time interval of the abnormal movement of the calculus, by calculating and analyzing these intervals, a suitable local division time span is obtained, so as to better correspondingly analyze the abnormal data in different dimensions in time.

[0078] The synchronization division unit is configured to synchronize division of the first urinary stone diagnosis related abnormal data, the second urinary stone diagnosis related abnormal data, and the biochemical index abnormal value correlation graph based on the local division time span sequence, to obtain multi-dimensional local abnormal data. The purpose of this step is to make the abnormal data from different dimensions (stone volume, position, and biochemical index) consistent and comparable in time. Through synchronization division, the abnormal data of different dimensions is sorted according to the same time segment, thereby obtaining multi-dimensional local abnormal data. For example, based on the determined time span, the time period of abnormal increase in stone volume, the time period of abnormal movement of stone position, and the time period of abnormal change of biochemical index are correspondingly divided, so that the relationship between different dimensional abnormalities can be observed and analyzed under the same time scale, and a more orderly data basis is provided for subsequent cross-dimensional fusion reasoning.

[0079] The cross-dimensional fusion reasoning unit is configured to perform cross-dimensional fusion reasoning on the multi-dimensional local abnormal data to obtain a plurality of cross-local dimensional abnormal fusion reasoning information and a cross-local dimensional abnormal fusion reasoning context of the patient. It comprehensively considers the abnormal conditions of stone volume, position, and biochemical index and other dimensions, and digs deep connections between them. Through such fusion reasoning, a plurality of cross-local dimensional abnormal fusion reasoning information of the patient is obtained. These information may reveal the internal relationship between abnormal changes in stone volume and specific position changes and related biochemical index changes, such as rapid increase in stone volume, movement of its position to a certain specific area, and abnormal fluctuation of certain biochemical indexes. The causal relationship or mutual influence mechanism between the three may constitute a cross-local dimensional abnormal fusion reasoning information. At the same time, the cross-local dimensional abnormal fusion reasoning context is generated to show the mutual correlation path between different dimensional abnormalities in a logically clear way, helping doctors to comprehensively understand the development and change mechanism of the disease, and providing strong support for accurate diagnosis and treatment.

[0080] In an alternative embodiment, the division time span determination unit comprises:

[0081] The causal correlation coefficient acquisition subunit is used to obtain the causal correlation coefficient between each indicator data in the first set of abnormal data related to the diagnosis of urinary stones and each indicator data in the second set of abnormal data related to the diagnosis of urinary stones. This subunit is responsible for finding the causal correlation coefficient between each indicator data in the first set of abnormal data related to the diagnosis of urinary stones (from the dynamic change simulation map of stone volume) and the second set of abnormal data related to the diagnosis of urinary stones (from the dynamic mapping map of stone location). This means that it needs to analyze the abnormal indicators related to stone volume and abnormal indicators related to stone location to see the degree of correlation between them in the process of leading to the formation or development of urinary stones. Assuming that the indicator in the first set of abnormal data related to the diagnosis of urinary stones is the stone volume growth rate, and the indicator in the second set of abnormal data related to the diagnosis of urinary stones is the time the stone stays in a specific area, the causal correlation coefficient acquisition subunit will determine the degree of correlation between these two indicators in terms of the cause of urinary stones through medical knowledge, past case data, or specific algorithms, and derive a coefficient representing the degree of correlation between the two. Such analysis is performed between each type of indicator data to obtain the corresponding causal correlation coefficient.

[0082] The time span sequence generation unit is used to determine the locally segmented time span sequence based on the correlation coefficient of urinary stone etiology between each indicator data in the first urinary stone diagnosis-related abnormal data, the urinary stone etiology correlation coefficient between each indicator data in the second urinary stone diagnosis-related abnormal data, the interval sequence of the same outlier in the first urinary stone diagnosis-related abnormal data, and the interval sequence of the same outlier in the second urinary stone diagnosis-related abnormal data.

[0083] ;

[0084] In the formula, To locally divide the time span sequence, This represents the total number of all indicator data types in the abnormal data related to the diagnosis of urinary calculi. This represents the total number of all types of indicators in the abnormal data related to the diagnosis of second urinary tract stones. The first abnormal data related to the diagnosis of urinary calculi The first indicator data and the second abnormal data related to the diagnosis of urinary stones The correlation coefficient between the data of these indicators regarding the causes of urinary stones To find the average value function, The first abnormal data related to the diagnosis of urinary calculi The first value in the interval sequence of outliers of the same type of index data (i.e., the first value in the first abnormal data related to the diagnosis of urinary stones). (the time interval between the first and second outliers of the corresponding index data). The first abnormal data related to the diagnosis of urinary calculi The first value in the interval sequence of outliers of the same type of index data (i.e., the first value in the second outlier-related abnormal data of urinary calculi diagnosis) (the time interval between the first and second outliers of the corresponding index data). The first abnormal data related to the diagnosis of urinary calculi The second value in the sequence of outlier intervals for the same type of indicator data. The first abnormal data related to the diagnosis of urinary calculi The second value in the sequence of outlier intervals for the same type of indicator data. The first abnormal data related to the diagnosis of urinary calculi The first outlier in the interval sequence of the same type of index data A number, The first abnormal data related to the diagnosis of urinary calculi The first outlier in the interval sequence of the same type of index data There are several values, and The value is equal to the first abnormal data related to the diagnosis of urinary stones. The total number of values ​​in the same outlier interval sequence of the first type of indicator data and the first outlier in the second type of urinary calculi diagnosis-related outlier data. The minimum value among the total number of values ​​in the interval sequence of the same outlier of a certain indicator data.

[0085] This time span sequence reflects the comprehensive temporal correlation characteristics of different abnormal indicators, providing an important basis for time division in subsequent synchronous analysis of abnormal data in different dimensions. For example, if the abnormality of rapid increase in stone volume and the abnormality of sudden change in stone location are closely related in time (i.e., the causal correlation coefficient is large), then when determining the local division time span sequence, the interval sequence between these two sets of abnormal values ​​will be given a higher weight, making the time span more reflective of the temporal relationship between these two abnormalities.

[0086] In an alternative implementation, the cross-dimensional fusion inference unit includes:

[0087] The causal association rule acquisition subunit is used to obtain all causal association rules for urinary stones. These rules are derived from medical research, clinical experience, and extensive case analysis, describing the relationship between different factors and the formation of urinary stones. For example, abnormal changes in certain biochemical indicators may have specific causal relationships with changes in the location and size of stones; these relationships constitute the causal association rules for urinary stone formation. These rules form the basis for subsequent analysis, providing a theoretical framework for understanding patients' abnormal data.

[0088] The degree of compliance determination sub-unit is configured to combine the multi-dimensional local abnormal data across dimensions based on all stone-cause association rules to obtain a plurality of cross-dimension combination abnormal data, and determine a degree of compliance of each cross-dimension combination abnormal data to a corresponding stone-cause association rule;

[0089] Suppose that the three abnormal data of sudden increase in stone volume, stone location moving to a specific area, and increase in a certain biochemical indicator come from different dimensions, and that they are combined together to form a cross-dimension combination abnormal data. Then, for each cross-dimension combination abnormal data, the degree of compliance of the combination to the rule is determined according to the stone-cause association rule, that is, the degree of compliance of the combination to the corresponding stone-cause association rule is determined. This step helps to evaluate the rationality and possibility of each cross-dimension combination abnormal data under the known medical knowledge system, and provides a quantitative basis for subsequent fusion reasoning.

[0090] The fusion reasoning sub-unit is configured to obtain a plurality of cross-local-dimension abnormal fusion reasoning information of the patient based on all stone-cause association rules and all cross-dimension combination abnormal data. This step comprehensively considers the degree of compliance of each cross-dimension combination abnormal data to the corresponding rule, and the potential relationship between different combinations. For example, if a cross-dimension combination abnormal data has a high degree of compliance to a stone-cause association rule, it means that this abnormal combination has a high possibility to follow the rule in the stone formation process. Through comprehensive analysis of all combinations and rules, a plurality of cross-local-dimension abnormal fusion reasoning information of the patient is obtained. These information further reveals the internal relationship between different dimensional abnormalities, and their influence on the formation and development of urinary stones.

[0091] The reasoning context generation sub-unit is configured to generate a cross-local-dimension abnormal fusion reasoning context based on all cross-dimension combination abnormal data and corresponding all stone-cause association rules and corresponding degrees of compliance, and combining corresponding all cross-local-dimension abnormal fusion reasoning information. This context shows in a visual or logical way how different dimensional abnormalities are related to each other, and how they follow or deviate from the known stone-cause association rules. It is like a detailed map that provides a clear idea for doctors and researchers to understand the complexity of the patient's condition and the role of different abnormal conditions in the stone formation process, thereby providing strong support for formulating more accurate diagnosis and treatment plans. For example, through the reasoning context, it can be intuitively seen how the abnormal increase in stone volume is influenced by the change in specific biochemical indicators and the movement of stone location, and ultimately leads to the development of stone condition.

[0092] In an alternative embodiment, the dynamic threshold reasoning sub-module comprises:

[0093] The macroscopic context deviation analysis unit is used to determine the abnormal fusion reasoning macroscopic deviation degree based on the standard cross-local dimension abnormal fusion reasoning context and the cross-local dimension abnormal fusion reasoning context. The core task of this unit is to evaluate the difference between the patient's condition and the standard situation. It achieves this goal by comparing the standard cross-local dimension abnormal fusion reasoning context and the patient's own cross-local dimension abnormal fusion reasoning context. The standard cross-local dimension abnormal fusion reasoning context is based on a large number of normal or common cases, representing the correlation and development pattern between different dimension abnormal situations under general cognition. The patient's cross-local dimension abnormal fusion reasoning context is inferred from the patient's specific stone volume, location, and biochemical index, and other multi-dimensional abnormal information. By comparing the two contexts in detail, the deviation between the patient's abnormal fusion reasoning mode and the standard mode is found out, and the abnormal fusion reasoning macroscopic deviation degree is determined. For example, if the standard context shows that the increase in stone volume is usually accompanied by a specific location movement and biochemical index change sequence, while the patient's condition shows a different development path in some aspects, the macroscopic context deviation analysis unit will quantify this difference and give a numerical value representing the deviation degree. This deviation degree reflects the speciality of the patient's condition relative to the standard situation, providing an important reference for subsequent personalized warning threshold determination and clinical decision-making. The abnormal fusion reasoning macroscopic deviation degree reflects the deviation degree of the patient's condition from the standard situation.

[0094] The deviation degree between the standard cross-local dimension abnormal fusion reasoning context and the patient's cross-local dimension abnormal fusion reasoning context can be achieved by the following steps:

[0095] Extract the key features in the two contexts, which may include the sequence, frequency, severity, etc. of the dimensional abnormalities such as stone volume, location, biochemical index, etc. For example, in the standard context, a specific biochemical index changes after a period of stone volume increase, while in the patient's context, the time interval or change sequence may be different. Match the extracted features according to the dimension and time sequence for subsequent comparative analysis.

[0096] For the matched features, the quantitative difference calculation of each dimension's abnormal situation is performed. For example, for the stone volume dimension, the difference in volume growth rate between the standard context and the patient's context is calculated; for the biochemical index dimension, the difference in the deviation degree of the specific index value from the standard range is calculated, etc.

[0097] The sequence and frequency difference of abnormal occurrence also needs to be quantified. For example, in the standard context, a certain abnormality occurs once every 10 days, while in the patient's context, it occurs once every 5 days. The frequency multiple is calculated to quantify this difference.

[0098] Different dimensions of feature differences are given weights, which are determined based on medical knowledge and clinical experience to determine the importance of each dimension to the disease. For example, the change in stone volume may have a greater impact on the disease and be given a higher weight; while the change in some secondary biochemical indicators has a relatively low weight.

[0099] According to the quantified values of the dimensional feature differences and the weights, a comprehensive value is calculated by mathematical methods such as weighted summation, which is the abnormal fusion reasoning macroscopic deviation degree, reflecting the overall deviation degree of the two contexts.

[0100] The hierarchical warning threshold determination unit is used to determine the user's personalized multi-dimensional hierarchical warning threshold based on the patient's all cross-local dimensional abnormal fusion reasoning information and abnormal fusion reasoning macroscopic deviation degree. Considering these two factors, this unit will set personalized multi-dimensional hierarchical warning thresholds for the patient. For example, if the patient's cross-local dimensional abnormal fusion reasoning information shows that multiple abnormal conditions related to the stone are more serious, and the abnormal fusion reasoning macroscopic deviation degree indicates that the patient's condition deviates from the standard condition, it may mean that the patient's condition is special and serious, so the hierarchical warning threshold determination unit will set a lower warning threshold accordingly, so as to more timely warn the patient's condition. These hierarchical warning thresholds can more accurately classify the severity of the disease and the risk level according to the actual situation of the patient individual, and provide more targeted warning prompts for the patient.

[0101] The clinical decision path simulation unit is used to determine the user's clinical decision simulation path based on the cross-local dimensional abnormal fusion reasoning context. This unit is based on the cross-local dimensional abnormal fusion reasoning context to determine the user's clinical decision simulation path. The cross-local dimensional abnormal fusion reasoning context clearly shows the logical relationship and development path between different dimensional abnormalities in the patient's stone disease, reflecting the internal mechanism of the disease. The clinical decision path simulation unit simulates the reasonable clinical decision steps and order for the patient's condition according to this context. For example, if the reasoning context shows that the increase in stone volume is closely related to a specific biochemical indicator abnormality, and this abnormal combination may lead to a high risk of further development of the stone, the clinical decision path simulation unit may suggest first intervening in the treatment of the biochemical indicator, and then adjusting the subsequent treatment plan according to the change in stone volume. This simulation path provides an important reference framework for doctors to make actual clinical treatment decisions, helping doctors to develop more scientific and personalized treatment plans according to the patient's specific condition.

[0102] In an alternative embodiment, the hierarchical warning threshold determination unit comprises:

[0103] The basic threshold determination subunit is configured to determine a multi-dimensional hierarchical early warning basic threshold based on all the cross-local dimension abnormality fusion reasoning information of the patient. The cross-local dimension abnormality fusion reasoning information covers the comprehensive analysis results of the correlation between the abnormal conditions of the stone in multiple dimensions such as volume, position, and biochemical indicators. Assuming that these information indicates that the patient's stone volume grows rapidly, accompanied by movement in a specific position and significant changes in certain biochemical indicators, the basic threshold determination subunit will preliminarily set a multi-dimensional hierarchical early warning basic threshold based on the severity, frequency of occurrence, and mutual influence relationship of these abnormal conditions. For example, if the stone volume increases rapidly and the biochemical indicators fluctuate frequently, the basic threshold may be relatively low, meaning that a lower degree of abnormal change may trigger a higher level of early warning, thereby reflecting a relatively serious condition. This basic threshold provides an initial framework for subsequent precise adjustment and is a preliminary early warning limit based on the patient's own abnormal conditions.

[0104] The hierarchical threshold determination subunit is configured to correct the multi-dimensional hierarchical early warning basic threshold based on the abnormality fusion reasoning macroscopic deviation degree, thereby obtaining the personalized multi-dimensional hierarchical early warning threshold of the user. This subunit corrects the multi-dimensional hierarchical early warning basic threshold based on the abnormality fusion reasoning macroscopic deviation degree, thereby obtaining the personalized multi-dimensional hierarchical early warning threshold of the user. The abnormality fusion reasoning macroscopic deviation degree reflects the deviation degree between the patient's abnormality fusion mode and the standard mode. If the abnormality fusion reasoning macroscopic deviation degree is large, it means that the patient's disease development mode is significantly different from the common standard mode, and at this time, the hierarchical threshold determination subunit will adjust the basic threshold accordingly. For example, the early warning threshold may be further reduced to make the early warning more sensitive, because the patient's special disease development may require more timely attention and intervention. Conversely, if the deviation degree is small, the basic threshold may be appropriately maintained to make the early warning more in line with the regular situation. Through this correction based on the deviation degree, the finally determined personalized multi-dimensional hierarchical early warning threshold can more accurately reflect the characteristics and risk degree of the patient's individual disease, and provide more targeted early warning standards for the clinic.

[0105] In an alternative embodiment, the early warning and reporting integration module comprises:

[0106] The early warning level diagnosis submodule is configured to determine the current early warning level of the user based on the latest relevant indicators of the user's stone disease and the personalized multi-dimensional grading early warning threshold of the user, and to issue an early warning signal based on the current early warning level. The submodule compares the latest relevant indicators of the user's stone disease with the personalized multi-dimensional grading early warning threshold obtained by the dynamic threshold reasoning module. The latest relevant indicators include the position, size, and morphological changes of the stone, as well as real-time data such as biochemical indicators in the blood and urine related to the occurrence of stone disease. By comparing these indicators with the corresponding early warning threshold, the current early warning level of the user is determined. For example, if the volume of the user's stone increases beyond the threshold of the corresponding level, and some key biochemical indicators are also in the abnormal range, it may be determined that the early warning level is high. After determining the early warning level, a corresponding early warning signal is issued according to the level. The early warning signal can be a direct visual prompt (such as color marking in the electronic medical record system), a sound reminder (mobile notification for medical staff), etc., the purpose of which is to timely inform the patient and medical staff of the severity of the current condition, so as to take appropriate measures.

[0107] The intervention suggestion generation submodule is configured to determine the subsequent intervention priority suggestion of the user based on the clinical decision simulation path. The clinical decision simulation path is formulated according to the patient's cross-local dimension anomaly fusion reasoning context, and includes a series of reasonable diagnosis and treatment steps for the patient's condition. The submodule evaluates the urgency and importance of each diagnosis and treatment step. For example, if the clinical decision simulation path shows that the stone is about to block the ureter, the related intervention (such as surgery or drug stone expulsion) to remove the blockage will be given a high priority; while some suggestions for improving life habits to prevent stone recurrence may have a relatively low priority. In this way, a clear sequence of subsequent intervention guidance is provided for medical staff and patients.

[0108] The report integration submodule is configured to generate a variety of visual curve data based on the comprehensive examination visualization results, and to integrate the current early warning level of the user, the variety of visual curve data, and the subsequent intervention priority suggestion of the user to generate a comprehensive diagnosis and treatment report of urinary stones. Based on the comprehensive examination visualization results provided by the data visualization module, such as the stone volume dynamic change simulation graph, the biochemical indicator abnormal value correlation graph, and the stone position dynamic mapping graph, a variety of visual curve data is generated. These curves can more directly show the trend of the condition over time, such as the growth curve of the stone volume over time, the fluctuation curve of a specific biochemical indicator, etc. The comprehensive diagnosis and treatment report provides comprehensive and detailed information on the patient's condition to the patient and medical staff, including the severity and trend of the current condition, as well as targeted guidance suggestions for subsequent intervention, which helps to develop a scientific and reasonable treatment plan.

[0109] It will be apparent to those skilled in the art that various modifications and variations can be made to the present application without departing from the spirit or scope of the application. Thus, it is intended that the present application cover modifications and variations of this application provided they come within the scope of the application and their equivalent technology.

Claims

1. A urinary calculus examination image recognition and data visualization analysis and early warning system, characterized in that, The method comprises the following steps: checking data analysis module, for extracting the user's urinary calculus examination image and the synchronous sequence data of the blood and urine examination report related to the sequence data of the calculus disease, including the time sequence alignment of the position distribution sequence characteristics, the morphological sequence characteristics and the calculus disease related biochemical sequence index of the urinary calculus; data visualization module, for converting the calculus disease related sequence data into a calculus volume dynamic change simulation diagram, a biochemical index abnormal value correlation graph and a calculus position dynamic mapping diagram as a comprehensive examination visualization result for user query; dynamic threshold reasoning module, for spatiotemporal alignment and cross-local dimension anomaly fusion reasoning of the calculus volume dynamic change simulation diagram, the biochemical index abnormal value correlation graph and the calculus position dynamic mapping diagram, to obtain the user's personalized multi-dimensional graded warning threshold and clinical decision simulation path; warning and report integration module, for determining the current warning level of the user based on the latest related index of the user's calculus disease and the user's personalized multi-dimensional graded warning threshold, and combining the visual curve data and the clinical decision simulation path generated based on the comprehensive examination visualization result to generate a urinary calculus comprehensive diagnosis and treatment report; The dynamic threshold reasoning module comprises: a cross-local dimension anomaly fusion reasoning submodule for spatiotemporal alignment and cross-local dimension anomaly fusion reasoning of the calculus volume dynamic change simulation diagram, the biochemical index abnormal value correlation graph and the calculus position dynamic mapping diagram to obtain a plurality of cross-local dimension anomaly fusion reasoning information and a cross-local dimension anomaly fusion reasoning context of the patient; The cross-local dimension anomaly fusion reasoning submodule comprises: an abnormal data extraction unit for extracting first urinary calculus diagnosis related abnormal data in the calculus volume dynamic change simulation diagram, and simultaneously extracting second urinary calculus diagnosis related abnormal data in the calculus position dynamic mapping diagram; a time span division determination unit for determining a local division time span sequence based on the same abnormal value interval sequence in the first urinary calculus diagnosis related abnormal data and the same abnormal value interval sequence in the second urinary calculus diagnosis related abnormal data; a synchronous division unit for synchronously dividing the first urinary calculus diagnosis related abnormal data, the second urinary calculus diagnosis related abnormal data and the biochemical index abnormal value correlation graph based on the local division time span sequence to obtain multi-dimensional local abnormal data; a cross-dimension fusion reasoning unit for cross-dimension fusion reasoning of the multi-dimensional local abnormal data to obtain a plurality of cross-local dimension anomaly fusion reasoning information and a cross-local dimension anomaly fusion reasoning context of the patient; The time span division determination unit comprises: a cause correlation coefficient acquisition submodule for acquiring a urinary calculus cause correlation coefficient between each index data in the first urinary calculus diagnosis related abnormal data and each index data in the second urinary calculus diagnosis related abnormal data; The time span sequence generation unit is configured to determine a local division time span sequence based on each index data in the first urinary calculus diagnosis related abnormal data, a urinary calculus cause correlation coefficient between each index data in the second urinary calculus diagnosis related abnormal data, a same abnormal value interval sequence in the first urinary calculus diagnosis related abnormal data, and a same abnormal value interval sequence in the second urinary calculus diagnosis related abnormal data. ； In the formula, To locally divide the time span sequence, This represents the total number of all indicator data types in the abnormal data related to the diagnosis of urinary calculi. This represents the total number of all types of indicators in the abnormal data related to the diagnosis of second urinary tract stones. The first abnormal data related to the diagnosis of urinary calculi The first indicator data and the second abnormal data related to the diagnosis of urinary stones The correlation coefficient between the data of these indicators regarding the causes of urinary stones To find the average value function, The first abnormal data related to the diagnosis of urinary calculi The first value in the sequence of outlier intervals for the same type of indicator data. The first abnormal data related to the diagnosis of urinary calculi The first value in the sequence of outlier intervals for the same type of indicator data. The first abnormal data related to the diagnosis of urinary calculi The second value in the sequence of outlier intervals for the same type of indicator data. The first abnormal data related to the diagnosis of urinary calculi The second value in the sequence of outlier intervals for the same type of indicator data. The first abnormal data related to the diagnosis of urinary calculi The first outlier in the interval sequence of the same type of index data A number, The first abnormal data related to the diagnosis of urinary calculi The first outlier in the interval sequence of the same type of index data There are several values, and The value is equal to the first abnormal data related to the diagnosis of urinary stones. The total number of values ​​in the same outlier interval sequence of the first type of indicator data and the first outlier in the second type of urinary calculi diagnosis-related outlier data. The minimum value among the total number of values ​​in the interval sequence of the same outlier of a certain indicator data.

2. The urinary stone detection image recognition and data visualization analysis and early warning system according to claim 1, characterized in that, The examination data analysis module comprises: The image recognition submodule is configured to extract urinary calculus position data and morphological data of each urinary calculus examination image in the user's synchronous sequence data based on all urinary calculus local images in each urinary calculus examination image in the user's synchronous sequence data. The data serialization submodule is configured to generate position distribution sequence features and morphological sequence features of urinary calculi based on the urinary calculus position data and morphological data of all urinary calculus examination images in the user's synchronous sequence data. The report serialization analysis submodule is configured to generate calculus occurrence related biochemical sequence indexes based on calculus occurrence related biochemical indexes in all portions of blood and urine examination reports in the user's synchronous sequence data. The time sequence alignment submodule is configured to perform time sequence alignment on the position distribution sequence features, morphological sequence features, and calculus occurrence related biochemical sequence indexes of urinary calculi to obtain calculus occurrence related sequence data.

3. The urinary stone detection image recognition and data visualization analysis and early warning system of claim 1, wherein, The data visualization module comprises: The urinary calculus dynamic simulation submodule is configured to generate a calculus volume dynamic change simulation diagram based on the morphological sequence features of urinary calculi in the calculus occurrence related sequence data, and simultaneously generate a calculus position dynamic mapping diagram based on the position distribution sequence features of urinary calculi in the calculus occurrence related sequence data. The first abnormal index extraction submodule is configured to extract first abnormal biochemical indexes and first sequence abnormal indexes from the position distribution sequence features and morphological sequence features of urinary calculi. The second abnormal index extraction submodule is configured to extract second abnormal biochemical indexes and second sequence abnormal indexes from the calculus occurrence related biochemical sequence indexes. The abnormal index mapping submodule is configured to generate a biochemical index abnormal value correlation map based on the first abnormal biochemical indexes, the first sequence abnormal indexes, the second abnormal biochemical indexes, and the second sequence abnormal indexes, as a comprehensive examination visualization result for user query.

4. The urinary stone detection image recognition and data visualization analysis and early warning system of claim 1, wherein, The dynamic threshold reasoning module further comprises: The dynamic threshold reasoning submodule is configured to determine a user's individualized multi-dimensional hierarchical early warning threshold and a clinical decision simulation path based on a standard cross-local dimension abnormal fusion reasoning context, a plurality of cross-local dimension abnormal fusion reasoning information of the patient, and the cross-local dimension abnormal fusion reasoning context.

5. The urinary stone detection image recognition and data visualization analysis and early warning system according to claim 1, characterized in that, The cross-dimension fusion reasoning unit comprises: The cause correlation rule acquisition submodule is configured to acquire all calculus cause correlation rules. The compliance degree determination submodule is configured to perform cross-dimension combination on multi-dimensional local abnormal data based on all calculus cause correlation rules, to obtain a plurality of cross-dimension combination abnormal data, and to determine a compliance degree of each cross-dimension combination abnormal data to a corresponding calculus cause correlation rule. The fusion reasoning subunit is configured to obtain multiple cross-local-dimension abnormal fusion reasoning information of the patient based on all stone cause association rules and all cross-dimension combined abnormal data. The reasoning context generation subunit is configured to generate cross-local-dimension abnormal fusion reasoning context based on all cross-dimension combined abnormal data, corresponding all stone cause association rules and corresponding degrees of compliance, and corresponding all cross-local-dimension abnormal fusion reasoning information.

6. The urinary stone detection image recognition and data visualization analysis and early warning system according to claim 1, characterized in that, The dynamic threshold reasoning sub-module comprises: The macroscopic context deviation analysis unit is configured to determine abnormal fusion reasoning macroscopic deviation degree based on the standard cross-local-dimension abnormal fusion reasoning context and the cross-local-dimension abnormal fusion reasoning context. The hierarchical early warning threshold determination unit is configured to determine the personalized multi-dimensional hierarchical early warning threshold of the user based on all cross-local-dimension abnormal fusion reasoning information of the patient and the abnormal fusion reasoning macroscopic deviation degree. The clinical decision path simulation unit is configured to determine the clinical decision simulation path of the user based on the cross-local-dimension abnormal fusion reasoning context.

7. The urinary stone detection image recognition and data visualization analysis and early warning system according to claim 6, characterized in that, The hierarchical early warning threshold determination unit comprises: The basic threshold determination subunit is configured to determine a multi-dimensional hierarchical early warning basic threshold based on all cross-local-dimension abnormal fusion reasoning information of the patient. The hierarchical threshold determination subunit is configured to correct the multi-dimensional hierarchical early warning basic threshold based on the abnormal fusion reasoning macroscopic deviation degree to obtain the personalized multi-dimensional hierarchical early warning threshold of the user.

8. The urinary stone detection image recognition and data visualization analysis and early warning system according to claim 1, characterized in that, The early warning and report integration module comprises: The early warning level diagnosis sub-module is configured to determine the current early warning level of the user based on the latest related indicators of the stone disease of the user and the personalized multi-dimensional hierarchical early warning threshold of the user, and to send an early warning signal based on the current early warning level. The intervention suggestion generation sub-module is configured to determine the subsequent intervention priority suggestion of the user based on the clinical decision simulation path. The report integration sub-module is configured to generate multiple visual curve data based on the comprehensive examination visualization result, and to integrate the current early warning level of the user, the multiple visual curve data and the subsequent intervention priority suggestion of the user to generate a urinary stone comprehensive diagnosis and treatment report.

Images