Arterial lesion treatment scheme auxiliary generation system based on ultrasound and radiography data

Abstract

The invention discloses an arterial lesion treatment scheme auxiliary generation system based on ultrasonic and radiography data, and belongs to the technical field of data processing. Comprising a data model storage unit used for storing relation models, an acquisition unit used for acquiring medical record data, and a reference medical record database generation unit used for screening medical records similar to a current patient and storing the medical record data of the medical records as a reference medical record database. The system comprises a data acquisition unit used for acquiring detection data, a data abnormity screening unit used for screening and outputting abnormal data and missing data, a reference scheme generation unit used for confirming and outputting a reference treatment scheme, and an interaction unit used for realizing data and information interaction between the system and a user. Meanwhile, a reference treatment scheme is automatically generated based on a case big database, risk points in an operation are locked, powerful assistance and support are provided for accurate treatment of doctors, and the treatment effect of arterial lesion is improved.

Description

Technical Field

[0001] This application belongs to the field of data processing technology and relates to an auxiliary system for generating treatment plans for arterial lesions based on ultrasound and angiography data. Background Technology

[0002] The lower extremities are a common site for arteriosclerosis obliterans. The root cause of lower extremity arteriosclerosis obliterans is vascular plaque (atherosclerosis): plaques that resemble porridge are formed when lipids, cholesterol, calcium and other substances are deposited under the arterial intima. This leads to thickening and hardening of the arterial wall, loss of elasticity, narrowing or even blockage of the lumen, ultimately affecting the normal flow of blood.

[0003] The formation of vascular plaques is associated with a variety of medical factors, including high cholesterol, high blood pressure, diabetes, smoking, obesity, lack of exercise, genetic factors, aging, and chronic inflammation. These factors ultimately lead to plaque formation by damaging the vascular endothelium, promoting lipid deposition, or triggering inflammation.

[0004] Typically, when a patient presents with symptoms suggestive of atherosclerosis, such as cold extremities, intermittent claudication, rest pain, or gangrene, the doctor will first perform an arterial ultrasound examination to obtain data on the vessel wall structure, plaque characteristics, blood flow velocity and direction. Based on the analysis of this data, a preliminary assessment of the degree of arterial stenosis is made. If the ultrasound examination reveals severe stenosis or occlusion, interventional surgery needs to be considered. To improve the effectiveness of surgical treatment and reduce surgical risks, angiography is usually performed before surgery to further determine the morphology of the vessel lumen, the location and degree of stenosis, collateral circulation, and other data. In complex cases, intravascular ultrasound (IVUS) is also required to provide more accurate images from inside the vessel. (IVUS measurements of atherosclerotic plaques include: plaque CSA (cross sectional area), maximum plaque thickness, minimum plaque thickness, plaque eccentricity, plaque burden, and luminal area stenosis rate; calcified lesions are analyzed semi-quantitatively based on the quadrant occupied by calcification). In existing applications, once doctors obtain the above-mentioned test data, they need to formulate corresponding surgical plans based on their own experience and judgment. For example, they need to determine the puncture point, stent length and type (stent implantation) based on the location, length and degree of stenosis, and determine the thrombolysis method based on the stability of the plaque.

[0005] In practical applications, the treatment of atherosclerosis involves a large amount of testing data, and the treatment of a lesion site in the blood vessel during surgery can potentially trigger a chain reaction in related sites (for example, unstable plaque fragments may detach during surgery and flow along the blood vessels, leading to distal embolism). Therefore, it is essential to ensure the accuracy of all testing data and a comprehensive surgical plan. Clearly, relying solely on doctors to perform precise analysis of the aforementioned testing data and formulate a comprehensive treatment plan would not only significantly increase the workload of doctors but also increase the risks of the surgery. Summary of the Invention

[0006] To reduce the workload of doctors and improve the treatment effect of arteriosclerosis, this application aims to propose a system for assisting in the generation of treatment plans for arterial lesions based on ultrasound and angiography data. This system ensures the accuracy and completeness of the test data through correlation analysis between various test data, and automatically generates reference treatment plans based on a large case database, identifying risk points during surgery. This provides strong assistance and support for doctors to provide accurate treatment and improves the treatment effect of arterial lesions. The specific solution is as follows:

[0007] A system for assisting in the generation of treatment plans for arterial lesions based on ultrasound and angiography data, comprising:

[0008] The data model storage unit is configured to store a first relational model reflecting the relationship between various data in arterial ultrasound data and arterial angiography data, a second relational model reflecting the relationship between various arterial ultrasound data and arterial angiography data and various disease influencing factors, and a third relational model reflecting the relationship between reference treatment plan and arterial ultrasound data and / or arterial angiography data.

[0009] The data acquisition unit is configured to acquire the current patient's medical record data.

[0010] The reference case database generation unit is configured to connect to the case database, acquire and filter cases and their medical records that are similar to the current patient from the case database based on the current patient's medical record data, and store them as the reference case database.

[0011] The data anomaly screening unit is configured to receive the current patient's medical record data and, based on the first relational model, the second relational model, and the reference case database, screen and output abnormal and missing data.

[0012] The reference treatment plan generation unit is configured to confirm and output a reference treatment plan based on the current patient's medical record data and the third relationship model.

[0013] The interaction unit includes data input and output components, configured to enable data and information interaction between the system and the user;

[0014] The data acquisition unit is connected to the data anomaly screening unit and automatically acquires or reviews relevant medical record data in response to abnormal or missing data, or outputs a data acquisition request through the interaction unit and monitors and acquires medical record data input by user feedback.

[0015] The medical record data includes one or more of the following: various influencing factors of the disease, arterial ultrasound data, and arterial angiography data.

[0016] The data anomaly screening unit includes:

[0017] The first screening subunit is configured to filter and output first abnormal data based on arterial ultrasound data and / or arterial angiography data in combination with the first relationship model, and mark it as a type of abnormality.

[0018] The second screening subunit is configured to filter and output second abnormal data based on the influencing factors of each disease and the second relationship model, and mark them as Class II abnormalities.

[0019] The third screening subunit is configured to compare and determine the third abnormal data and missing data in the current patient's medical record data based on the reference case database, and mark them as three types of abnormalities;

[0020] The exception notification sub-unit is configured to alert the user of data exceptions through an interactive unit in a set display manner, based on the type of data exception.

[0021] Through the above technical solution, when a user inputs the current patient's medical record data, the system can search and match similar cases and their medical record data from the large medical record database to form a reference case database. This provides a reliable basis for subsequent data accuracy, screening, and treatment plan customization. By reviewing the current patient's own ultrasound / contrast imaging data and the correlation between them, the accuracy of the ultrasound and contrast imaging data is initially determined. Then, by combining various influencing factors of the current patient's condition with the correlation between the ultrasound / contrast imaging data and these influencing factors, the accuracy of the data is further determined. Finally, the reference case database is used for final data accuracy screening (including screening for missing data). Ensuring the accuracy, completeness, and reliability of relevant medical record data provides a data guarantee for the generation of reference treatment plans. In the above technical solution, all medical record data acquisitions undergo relevance verification, ensuring the completeness of required data while avoiding unnecessary data collection. For example, if an arterial ultrasound examination shows that the patient's blood vessels have only mild stenosis, only drug treatment is needed, without further arterial angiography, thus avoiding over-medicalization and examination. By using three types of data anomaly markers and displaying them in different ways, users can be effectively alerted to abnormal data. Users can focus more on the formulation of treatment plans rather than data verification, and combined with the provided reference treatment plans, it is easier for users to customize more accurate treatment plans in the future.

[0022] Optionally, the reference case database generation unit includes:

[0023] The medical record data parsing subunit is configured to parse one or more of the following from the current patient's medical record data: various disease influencing factors, arterial ultrasound data, and arterial angiography data.

[0024] The similarity determination algorithm subunit is configured to match or generate a similarity determination algorithm based on at least one specific symptom influencing factor in the current patient's medical record data to determine the similarity between the current patient's arterial ultrasound data and / or arterial angiography data and related medical record data in the case database;

[0025] The first matching subunit is configured to match the same or similar cases and their medical records from the case database based on the parsed arterial ultrasound data and / or arterial angiography data, and form the first intermediate database.

[0026] The second matching subunit is configured to match cases and their medical records that are the same as or similar to the above data from the medical record database based on the disease influencing factors obtained from the analysis, and form a second intermediate database.

[0027] The database fusion subunit is configured to acquire data from the first intermediate database and the second intermediate database, and take the intersection of cases and their medical records to form the reference case database.

[0028] The similarity determination algorithm subunit includes:

[0029] The weight configuration table is used to store the correspondence between one or more specific disease influencing factors and the weight coefficients for similarity judgment of various data in arterial ultrasound data and arterial angiography data;

[0030] The similarity score relationship table is used to store the similarity scores corresponding to the range of differences between various data in arterial ultrasound data and arterial angiography data;

[0031] The similarity determination module is configured to acquire arterial ultrasound data and / or arterial angiography data and the values ​​of each data item contained in the current patient's medical record data, compare them with the relevant data contained in each case in the case database, obtain the similarity score of each data item, and then calculate the similarity based on the similarity determination weight coefficient corresponding to each data item.

[0032] The cases and their medical records in the first intermediate database, the second intermediate database, and the reference case database are sorted and stored according to their similarity.

[0033] If the number of cases in the reference case database is lower than the set value, then the same number or a set proportion of cases and their medical records are extracted from the first intermediate database and the second intermediate database respectively to expand the data until the number of cases in the reference case database reaches the set value.

[0034] The above technical solution can determine the conditions for generating a reference case database based on one or more influencing factors of a patient's current condition, thereby making the data in the generated reference case database more valuable. At the same time, maintaining the number of reference cases in the reference case database at a set value based on similarity helps to generate personalized data models and improve the accuracy of reference treatment plans.

[0035] Optionally, the reference case database generation unit further includes:

[0036] The database correction subunit is configured to obtain the relevant medical record data after review or the medical record data input by the user feedback and temporarily store it as correction data, and look up the similarity judgment weight coefficient corresponding to the above correction data from the weight configuration relationship table.

[0037] The difference between the corrected data and its corresponding original data is calculated by combining the similarity judgment weight coefficient;

[0038] If the difference in the similarity scores is lower than the set value, the reference case database will not be updated; otherwise, similar cases and their medical records will be searched again from the large case database based on the corrected data to form a new reference case database.

[0039] The above technical solutions can ensure that the cases in the reference case database have reference value, which is conducive to improving the accuracy of subsequent reference treatment plans.

[0040] Optionally, the third screening subunit includes:

[0041] The data missing detection module is configured to compare and search for missing medical record data in the current patient's medical record data based on the cases and their medical record data contained in the reference case database, and output the missing data.

[0042] The first personalized relationship model generation module is configured to generate, based on the cases and medical records contained in the reference case database, a first personalized relationship model that reflects the relationship between various data in arterial ultrasound data and arterial angiography data, and the relationship between various arterial ultrasound data and arterial angiography data and the influencing factors of various diseases, based on multi-factor analysis or neural network training.

[0043] The data anomaly output module is configured to match and output the third anomaly data based on the current patient's medical record data and the first personalized relationship model.

[0044] Through the above technical solutions, the first personalized relational model generated based on the reference case database is more suitable for judging the anomalies in the current patient medical record data. Compared with the anomalies generated by using the first relational model and the second relational model, the generated third anomaly data is more accurate, making it easier for users to make a reasonable and comprehensive judgment on the accuracy of the data. At the same time, the data missing detection module can be used to check the data categories that have low similarity weight coefficients but are related to the generation of the reference treatment plan, ensuring the integrity of the current patient medical record data.

[0045] Optionally, the data acquisition unit includes:

[0046] The data retrieval subunit is configured to retrieve and summarize the current patient's medical record data in a set format from the data storage area containing the current patient's various medical record data.

[0047] The data verification subunit is configured to receive and respond to the abnormal data output by the data anomaly screening unit, retrieve the corresponding original data from the current patient medical record data and verify it. If the data is consistent, a user confirmation pop-up is output; if the data is inconsistent, the current abnormal data is corrected and a user confirmation pop-up is output.

[0048] The data request subunit is configured to receive and respond to missing data output by the data anomaly screening unit, and to obtain data acquisition requests and monitor and acquire medical record data input by user feedback through the interaction unit.

[0049] The data storage subunit is configured to store various types of acquired data.

[0050] Through the above technical solution, the data acquisition unit can not only retrieve various examination data from relevant examination equipment, such as arterial ultrasound data, but also actively review relevant medical record data based on the output results of the data anomaly screening unit, which helps to improve the accuracy and completeness of the data.

[0051] Optionally, the data acquisition unit further includes:

[0052] The anomaly determination subunit is configured to compare the difference between the three types of abnormal data and the first and second types of abnormal data. If the difference exceeds the set value, a data supplementation sampling request is output to the interaction unit.

[0053] The supplementary sampling data included intravascular ultrasound data.

[0054] When there is a significant difference between the abnormal data identified based on the first personalized relationship model and the abnormal data identified based on the existing relationship model, it can be understood that an important piece of data contained in the current patient's medical record data is weakened in the first or second relationship model, resulting in the omission of data anomalies when determining data correlation. By supplementing the data categories through the above technical solution, the accuracy and completeness of the data can be further improved.

[0055] Optionally, the reference treatment plan includes the treatment plan category and detailed parameters of the plan;

[0056] The treatment options include: mechanical thrombolysis, thrombectomy, balloon angioplasty, stent implantation, combined thrombectomy, and hybrid surgery.

[0057] The detailed parameters of the protocol include: drug type and dosage, stent length and deployment method, stent placement location, instrument and operation parameters, physiological indicator monitoring items and monitoring protocol, and treatment risk points.

[0058] Optionally, the reference scheme generation unit includes:

[0059] The second personalized relationship model generation module is configured to analyze and generate a second personalized relationship model based on the cases and their medical records contained in the reference case database, which reflects the correlation between the reference treatment plan and arterial ultrasound data and / or arterial angiography data.

[0060] The first reference scheme generation subunit is configured to receive the current patient's medical record data and the third relationship model, and confirm and output the first reference scheme.

[0061] The second reference scheme generation subunit is configured to receive the current patient's medical record data and the second personalized relationship model, and confirm and output the second reference scheme.

[0062] The reference scheme confirmation subunit is configured to receive and compare the first reference scheme with the second reference scheme:

[0063] If the two reference treatment plans are of the same category and the deviation of the detailed parameters of the plans is within the set range, then the second reference plan shall be selected as the reference treatment plan.

[0064] If the two treatment plans are of different categories or the deviation of the detailed parameters of the plans exceeds the set range, the first reference plan, the second reference plan, and the prompt information will be output to the interactive unit.

[0065] The above technical solution can generate different reference treatment plans based on the general relationship model and the personalized relationship model. When the generated plans differ too much, the interactive unit will output prompts to remind the user to make a judgment and choice, so as to ensure the accuracy of the final plan.

[0066] Optionally, the data model storage unit may also store a fourth relation model that reflects the correspondence between each treatment plan and its corresponding risk point;

[0067] The arterial lesion treatment plan generation system also includes:

[0068] The risk locking unit is configured to connect to the reference treatment plan generation unit, obtain the reference treatment plan, and generate risk points corresponding to the current reference treatment plan based on the fourth relationship model, and output them to the interaction unit.

[0069] Through the above technical solution, when the system generates a reference treatment plan, it can simultaneously output the corresponding treatment risk points, which facilitates users to comprehensively evaluate the reference treatment plan and improve the treatment effect.

[0070] Optionally, the step of generating a reference treatment plan based on the arterial lesion treatment plan generation system includes:

[0071] Retrieve the current patient's medical record data;

[0072] Based on the above medical record data, cases and their medical records similar to the current patient are selected from the case database and stored as a reference case database;

[0073] Based on the analysis of the aforementioned reference case database, a first personalized relationship model was generated to reflect the correlation between various data in arterial ultrasound data and arterial angiography data, as well as the correlation between various arterial ultrasound data and arterial angiography data and the influencing factors of various diseases.

[0074] Perform anomaly screening on current patient medical record data:

[0075] Determine whether the correlation between various data in arterial ultrasound data and / or arterial angiography data conforms to the first relationship model; if not, output the first abnormal data.

[0076] Determine whether the correlation between various arterial ultrasound data and / or arterial angiography data and the influencing factors of each disease conforms to the second relationship model. If it does not conform, output the second abnormal data.

[0077] Determine whether the current patient's medical record data is missing compared to the medical record data of similar cases in the reference case database; if so, output the missing data.

[0078] Determine whether the correlation between various data in arterial ultrasound data and / or arterial angiography data, and the correlation between various arterial ultrasound data and / or arterial angiography data and various disease influencing factors, are consistent with the first personalized relationship model. If they are not consistent, output the third abnormal data.

[0079] Based on the above abnormal data, review the original data and / or obtain new medical record data;

[0080] Determine whether the above reference case database needs to be revised and a new reference case database needs to be generated based on the reviewed and / or supplemented medical record data;

[0081] If the reference case database needs to be corrected, then the abnormal data in the current patient's medical record data will be screened again;

[0082] Based on the case and medical record data of each case in the revised reference case database, a second personalized relationship model is generated to reflect the correlation between the reference treatment plan and arterial ultrasound data and / or arterial angiography data.

[0083] Based on the screened medical record data, the first reference plan and the second reference plan are generated by combining the third relationship model and the second personalized relationship model, respectively.

[0084] The system compares the first reference scheme with the second reference scheme and determines a reference treatment scheme based on the degree of difference, or outputs a prompt message to the interaction unit for the user to determine the final treatment scheme.

[0085] This application solution includes at least one of the following beneficial effects:

[0086] (1) After the patient’s medical record data is entered into the system, the system can search and match similar cases and their medical record data from the medical record database to form a reference case database, which can provide a reliable basis for subsequent data accuracy, screening and program customization;

[0087] (2) Screening medical record data multiple times to improve the accuracy of the generated reference treatment plan: By reviewing the current patient's own ultrasound data / contrast imaging data and the correlation between the two, the accuracy of the ultrasound and contrast imaging data is initially determined. Then, the accuracy of the data is determined again by combining the various disease influencing factors in the current patient's medical record data and the correlation between the ultrasound / contrast imaging data and the disease influencing factors. Finally, a personalized relationship model is generated by introducing the reference case database to perform the final data accuracy screening (including the screening of missing data). This ensures the accuracy, completeness and reliability of the relevant medical record data and provides data assurance for the generation of the reference treatment plan.

[0088] (3) By marking three types of data anomalies and displaying them in different ways, users can be effectively reminded to pay attention to abnormal data. Users can focus more on the formulation of treatment plans rather than data verification and other tasks. Combined with the provided reference treatment plans, it is easier for users to customize more accurate treatment plans in the future. Attached Figure Description

[0089] Figure 1 This is a schematic diagram of the overall system of this application;

[0090] Figure 2 This is a structural diagram of the system functional modules of this application.

[0091] Figure reference numerals: 100, Data model storage unit; 200, Data acquisition unit; 210, Data retrieval subunit; 220, Data verification subunit; 230, Data request subunit; 240, Abnormality determination subunit; 250, Data storage subunit; 300, Reference case database generation unit; 310, Medical record data parsing subunit; 320, Similarity determination algorithm subunit; 3201, Weight configuration relationship table; 3202, Similarity score relationship table; 3203, Similarity determination module; 330, First matching subunit; 340, Second matching subunit; 350, Database fusion subunit; 360 400. Database Correction Subunit; 410. Data Anomaly Screening Subunit; 420. Second Screening Subunit; 430. Third Screening Subunit; 4301. Data Missing Check Module; 4302. First Personalized Relationship Model Generation Module; 4303. Data Anomaly Output Module; 440. Anomaly Prompt Subunit; 500. Reference Solution Generation Unit; 510. Second Personalized Relationship Model Generation Module; 520. First Reference Solution Generation Subunit; 530. Second Reference Solution Generation Subunit; 540. Reference Solution Confirmation Subunit; 600. Interaction Unit; 700. Risk Locking Unit. Detailed Implementation

[0092] The embodiments of this application are described in detail below, and examples of the embodiments are shown in the accompanying drawings.

[0093] In the description of this specification, the references to "certain embodiments," "one embodiment," "some embodiments," "illustrative embodiment," "example," "specific example," or "some examples" refer to specific features, structures, materials, or characteristics described in connection with the described embodiment or example, which are included in at least one embodiment or example of this application. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples.

[0094] A system for assisting in the generation of treatment plans for arterial lesions based on ultrasound and angiography data, such as Figure 1 As shown, it includes a data model storage unit 100, a data acquisition unit 200, a reference case database generation unit 300, a data anomaly screening unit 400, a reference protocol generation unit 500, and an interaction unit 600.

[0095] The data model storage unit 100 includes local storage and cloud storage. The local storage includes a hard disk or solid-state drive of a computer or personal computer, and the cloud storage is configured as a database connected to a cloud server. The data model storage unit 100 is configured to store a first relational model reflecting the correlation between various data in arterial ultrasound data and arterial angiography data, a second relational model reflecting the correlation between various arterial ultrasound data and arterial angiography data and various influencing factors of diseases, and a third relational model reflecting the correlation between reference treatment plans and arterial ultrasound data and / or arterial angiography data.

[0096] The aforementioned first relational model encompasses both the relationships between various data points within arterial ultrasound data and the relationships between arterial ultrasound data and arterial angiography data. For example, the correlation between blood flow velocity and stenosis degree obtained from arterial ultrasound data distal to the lesion location and the blood flow velocity and stenosis degree at the lesion location is such that, since blood flow is constant, the product of blood flow velocity and stenosis degree at these two locations should be a fixed value. Similarly, there is a clear correspondence between angiography data at the lesion location (from which the degree and location of stenosis can be obtained) and ultrasound data. Through this first relational model, abnormal data can be screened out using the correlations between various data points in arterial ultrasound or angiography data. This allows for a clear understanding of data anomalies caused during data acquisition; for example, during vascular ultrasound examination, if the ultrasound probe does not make close contact with the skin surface, resulting in blurred ultrasound images and thus inaccurate stenosis degree assessments.

[0097] The disease-influencing factors mentioned in the second relationship model above include, but are not limited to: high blood lipids, high blood pressure, diabetes, smoking, obesity, lack of exercise, genetic factors, aging, and chronic inflammatory response. These disease-influencing factors typically determine the location and degree of atherosclerosis to a large extent. When conducting disease risk prediction, the second relationship model can be used to estimate the risk of lesions in areas that are difficult to examine.

[0098] In the aforementioned third relationship model, the reference treatment plan includes plan classification and detailed parameters. Plan classification includes, but is not limited to: mechanical thrombolysis, thrombectomy, balloon angioplasty, stent implantation, combined thrombectomy (delivering a compressed mesh stent to the distal end of the thrombus and releasing it to embed it into the thrombus; after a short wait, the stent and the entangled thrombus are pulled out of the body together), and composite surgery (primarily for multi-segmental, diffuse, and complex lesions of chronic lower limb ischemia). Detailed parameters include not only the type and dosage of medication, stent length and deployment method, stent placement location, instrument and operational parameters, aspiration negative pressure, balloon dilation size, and treatment risk points, but also physiological monitoring items and monitoring protocols during surgery, such as monitoring blood pressure at specific points on the body.

[0099] The data acquisition unit 200 is configured to acquire the current patient's medical record data, which includes outpatient medical records, laboratory reports, medical imaging examination data, pathology data, and nursing records. According to data categories, the medical record data includes one or more of the following: influencing factors of various diseases, arterial ultrasound data, and arterial angiography data.

[0100] In detail, in the embodiments of this application, the data acquisition unit 200 includes: a data retrieval subunit 210, a data verification subunit 220, a data request subunit 230, and a data storage subunit 250.

[0101] The data storage subunit 250 is configured to store various types of acquired data.

[0102] The data retrieval subunit 210 is configured to retrieve and summarize the current patient's medical record data according to a set format from the data storage area containing the current patient's various medical record data. For example, it can retrieve various medical record information entered by doctors from the structured data entry module. The aforementioned structured data entry refers to the standardized and unified input process of data according to predefined standardized formats, models, or terminology to ensure data consistency, analyzability, and efficient retrieval. The aforementioned data storage area also includes data storage devices for various examination equipment, such as high-precision ultrasound examination instruments and contrast imaging devices, as well as the patient's personal medical record data stored on a cloud server. After retrieving the above data information, the data retrieval subunit 210 performs preliminary format filtering and classification, and stores it according to the set format.

[0103] The data verification subunit 220 is configured to receive and respond to abnormal data output by the data anomaly screening unit 400, retrieve the corresponding original data from the current patient medical record data and verify it. If the data matches, a user confirmation pop-up is output for confirmation by the user (usually a doctor). If the data does not match, the current abnormal data is corrected and a user confirmation pop-up is output. The purpose of the above-mentioned data verification subunit 220 is to reduce data deviation during data extraction and ensure data consistency and accuracy. For example, if an error occurs in updating the data in the device's data storage, resulting in incorrect data obtained by the data retrieval subunit 210, the data verification subunit 220 can investigate and resolve such problems.

[0104] The data request subunit 230 is configured to receive and respond to missing data output by the data anomaly screening unit 400, output a data acquisition request through the interaction unit 600, and monitor and acquire medical record data input by the user. In this embodiment, the interaction unit 600 includes data input and output components, including but not limited to a mouse, keyboard, monitor, joystick, touch panel, voice player, and microphone, configured to enable data and information interaction between the system and the user. When a data anomaly is detected, the data request subunit 230 prompts the user about the relevant data anomaly through the monitor or voice player and outputs a dialog box requesting the user to confirm the abnormal data. If the data needs revision, a corresponding dialog box is output, and the system monitors and acquires new medical record data input by the user to replace the original data.

[0105] As can be seen from the above-disclosed embodiments, the data acquisition unit 200 in this application can not only retrieve various examination data from relevant examination equipment, such as arterial ultrasound data, blood pressure detection data, blood glucose detection data, etc., but also actively review relevant medical record data based on the output results of the data anomaly screening unit 400, thereby achieving preliminary screening of the data and helping to improve the accuracy and completeness of the data.

[0106] The reference case database generation unit 300 is configured to connect to the case database, obtain and filter similar cases and their case data from the case database based on the current patient's medical record data, and store them as the reference case database.

[0107] Detailed, combined Figure 2 As shown, the reference case database generation unit 300 in this application embodiment includes: a medical record data parsing subunit 310, a similarity determination algorithm subunit 320, a first matching subunit 330, a second matching subunit 340, and a database fusion subunit 350.

[0108] The medical record data parsing subunit 310 is configured to parse one or more of the following from the current patient's medical record data: various influencing factors of the disease, arterial ultrasound data, and arterial angiography data. In practical applications, the above data is acquired by the data acquisition unit 200 and stored in a predefined manner, facilitating data parsing by the medical record data parsing subunit. For example, it can determine whether a particular case data belongs to an influencing factor of the disease and perform more specific classification.

[0109] Unlike existing technologies that directly use one or more medical record data as a reference for similarity determination, in this embodiment, the similarity determination algorithm subunit 320 is configured to match or generate a similarity determination algorithm based on at least one specific symptom influencing factor in the current patient's medical record data. This algorithm is used to determine the similarity between the current patient's arterial ultrasound data and / or arterial angiography data and related medical record data in the case database. That is, different patients correspond to different similarity determination rules, aiming to find more suitable cases and their medical record data for reference. Specifically, in this embodiment, the similarity determination algorithm subunit 320 includes: a weight configuration relationship table 3201, a similarity score relationship table 3202, and a similarity determination module 3203.

[0110] Table 3201, a weighting configuration table, stores the correspondence between one or more specific disease-related influencing factors and the weighting coefficients for similarity assessment in arterial ultrasound and arterial angiography data. In practice, the accuracy of arterial ultrasound data is often affected by relevant disease-related influencing factors. For example, a patient's obesity status can affect the vessel wall thickness data in arterial ultrasound data, directly impacting the reliability of the aforementioned data. If the original data is still used for comparison, ignoring these influencing factors during the data similarity comparison process will obviously lead to inaccurate similarity assessment. The ultimate effect of configuring similarity assessment weighting coefficients here is to provide correction for similarity assessment results based on data reliability, making the reference cases and their medical records more reliable in subsequent studies.

[0111] The similarity score table 3202 is used to store the similarity scores corresponding to the difference range of various data in arterial ultrasound data and arterial angiography data. For example, for the same moderate vascular stenosis (the reduction in the vascular diameter shown by imaging examination is 50%-70%), the similarity score can be further assigned according to the specific degree of difference. For example, a difference of less than 5% is scored as 10 points, and a difference of 5%-10% is scored as 6 points, etc.

[0112] The similarity determination module 3203 is configured to acquire arterial ultrasound data and / or arterial angiography data and the values ​​of various data contained in the current patient's medical record data, compare them with the relevant data contained in each case in the case database, obtain the similarity score of each data, and then calculate the similarity by weighted summation according to the similarity determination weight coefficients corresponding to each data. The above similarity is actually a specific numerical value.

[0113] Based on the aforementioned similarity determination algorithm, the first matching subunit 330 is configured to match cases and their medical records that are identical or similar to the data obtained from the parsed arterial ultrasound data and / or arterial angiography data from the large case database, forming a first intermediate database. The second matching subunit 340 is configured to match cases and their medical records that are identical or similar to the data obtained from the parsed disease influencing factors from the large case database, forming a second intermediate database.

[0114] The cases in the first and second intermediate databases are mostly the same. The database fusion subunit 350 is configured to acquire data from the first and second intermediate databases, and take the intersection of the cases and their medical records to form the reference case database. In practice, the cases and their medical records in the first, second, and reference case databases are sorted and stored according to their similarity to facilitate data querying and retrieval. In practical applications, the case information numbers can be used for sorting.

[0115] If the number of cases in the reference case database is lower than the set value, then the same number or a set proportion of cases and their medical records are extracted from the first intermediate database and the second intermediate database respectively to expand the data until the number of cases in the reference case database reaches the set value, which facilitates the generation of personalized relationship models later.

[0116] The data anomaly screening unit 400 is configured to receive the current patient's medical record data and, based on the first relational model, the second relational model, and the reference case database, screen and output abnormal and missing data.

[0117] Detailed, such as Figure 2 As shown, the data anomaly screening unit 400 includes: a first screening subunit 410, a second screening subunit 420, a third screening subunit 430, and an anomaly alert subunit 440.

[0118] The first screening subunit 410 is configured to filter and output first abnormal data based on arterial ultrasound data and / or arterial angiography data in conjunction with a first relational model, marking them as Class I abnormalities. Class I abnormalities reflect inherent abnormalities within a specific type of data, such as an incorrect correlation between the degree of stenosis at different locations of blood vessels and blood flow velocity in arterial ultrasound data. The second screening subunit 420 is configured to filter and output second abnormal data based on various disease influencing factors in conjunction with a second relational model, marking them as Class II abnormalities. For example, if the second relational model shows multiple disease influencing factors simultaneously falling within a certain interval, but a certain data point in the arterial ultrasound data, which should be within a set interval, shows an abnormality. The third screening subunit 430 is configured to compare and determine third abnormal data and missing data in the current patient's medical record data based on a reference case database, marking them as Class III abnormalities. For example, if similar cases in the reference case database all contain arterial angiography data, but the current patient's medical record data does not contain such data, it is marked as missing data. This also includes missing data from examinations such as blood glucose and blood lipids. The exception notification subunit 440 is configured to remind the user of data exceptions through the interaction unit 600 in a set display mode according to the type of data exception. For example, the user can distinguish the types of exceptions by changing the color of the data item display, making it easier for the user to differentiate them.

[0119] It should be noted that the first relational model, second relational model, etc., stored in the aforementioned data model storage unit 100 are all general relational models trained using data from a large case database. Data validation using these general relational models is merely to exclude obviously erroneous data in the current patient's medical records, not to correct the data based on the association relationships of these general relational models. However, in practice, different patients have specific characteristics, and obviously, the aforementioned general relational models cannot accurately screen for related data anomalies. For example, if multiple segments of a blood vessel have various types of sclerotic lesions, the correspondence in the general relational model will become unstable.

[0120] Therefore, in this embodiment of the application, the third screening subunit 430 specifically includes: a data missing detection module 4301, a first personalized relationship model generation module 4302, and a data anomaly output module 4303.

[0121] The data missing data detection module 4301 is configured to compare and search for missing medical record data in the current patient's medical record data based on the cases and their medical record data contained in the reference case database, and output the missing data. The first personalized relationship model generation module 4302 is configured to generate, based on the cases and their medical record data contained in the reference case database, a first personalized relationship model reflecting the correlation between various data in arterial ultrasound data and arterial angiography data, and reflecting the correlation between various arterial ultrasound data and arterial angiography data and the influencing factors of various diseases, using multifactor analysis or neural network training. In practice, multifactor analysis is usually used to generate the above-mentioned first personalized relationship model. The data anomaly output module 4303 is configured to match and output third abnormal data based on the current patient's medical record data and the first personalized relationship model. Since the source data for the training analysis of the first personalized relationship model comes from the reference case database, it is more effective in discovering anomalies in the current patient's medical record data.

[0122] Based on the above scheme, the data acquisition unit 200 also includes an anomaly determination subunit 240, configured to compare the differences between the three types of abnormal data and the first and second types of abnormal data. If the difference exceeds a set value, a data supplementation sampling request is output to the interaction unit 600. The supplemented sampling data includes, but is not limited to, intravascular ultrasound data. The significance of the above scheme is that when the difference between the abnormal data determined based on the first personalized relationship model and the abnormal data determined by the existing relationship model is too large, it can be understood that an important piece of data contained in the current patient's medical record data is weakened in the first or second relationship model, resulting in the omission of data anomalies when performing data correlation determination. Based on the above scheme, by supplementing the data categories, the accuracy and completeness of the data can be further improved.

[0123] The data acquisition unit 200 is connected to the data anomaly screening unit 400. In response to abnormal or missing data, it automatically acquires or reviews relevant medical record data, or outputs a data acquisition request through the interaction unit 600 and monitors the medical record data input by the user feedback.

[0124] In an optimized embodiment of this application, the reference case database generation unit 300 further includes a database correction subunit 360. Specifically, it is configured to acquire reviewed relevant medical record data or user-inputted medical record data and temporarily store it as corrected data. It then searches for the similarity judgment weight coefficient corresponding to the corrected data in the weight configuration relationship table 3201. Based on the difference between the corrected data and its corresponding original data, and combined with the similarity judgment weight coefficient, it calculates the difference between the two similarity scores. If the difference in the similarity scores is lower than a set value, the reference case database is not updated; otherwise, it re-screens and searches for similar cases and their medical record data from the large case database based on the corrected data to form a new reference case database. This ensures that the cases available for reference in the reference case database have reference value, which is beneficial for improving the accuracy of subsequent reference treatment plans.

[0125] The reference treatment plan generation unit 500 is configured to confirm and output a reference treatment plan based on the current patient's medical record data and the third relation model. Specifically, the reference treatment plan generation unit 500 includes: a first reference treatment plan generation subunit 520, configured to receive the current patient's medical record data and the third relation model, and confirm and output a first reference treatment plan.

[0126] To further improve the accuracy of the reference treatment plan, in this embodiment, the reference plan generation unit 500 further includes a second personalized relationship model generation module 510, a second reference plan generation subunit 530, and a reference plan confirmation subunit 540.

[0127] The second personalized relationship model generation module 510 is configured to generate a second personalized relationship model reflecting the correlation between the reference treatment plan and arterial ultrasound data and / or arterial angiography data, based on the cases and their medical records contained in the reference case database, through multi-factor analysis or neural network training analysis. This model represents the treatment plan category and detailed parameters corresponding to various arterial ultrasound and / or arterial angiography data. The second reference plan generation subunit 530 is configured to receive the current patient's medical record data and the second personalized relationship model, and confirm and output the second reference plan. The reference plan confirmation subunit 540 is configured to receive and compare the first reference plan and the second reference plan: if the treatment plan categories are the same and the deviation of the detailed parameters is within a set range, the second reference plan is selected as the reference treatment plan. If the treatment plan categories are different or the deviation of the detailed parameters exceeds a set range, the first reference plan, the second reference plan, and a prompt message are output to the interaction unit 600, allowing the user to confirm which treatment plan to use.

[0128] Furthermore, the data model storage unit 100 also stores a fourth relation model reflecting the correspondence between each treatment plan and its corresponding risk points. For example, blood vessels at specific locations are prone to damage during stent surgery or are prone to acute thrombosis after stent placement. Correspondingly, the arterial lesion treatment plan auxiliary generation system also includes a risk locking unit 700, configured to connect with the reference plan generation unit 500. It obtains reference treatment plans and, based on the fourth relation model, generates risk points corresponding to the current reference treatment plan and outputs them to the interaction unit 600. When the system generates a reference treatment plan, it can simultaneously output the corresponding treatment risk points, facilitating a comprehensive evaluation of the reference treatment plan by the user and improving treatment effectiveness.

[0129] Based on the arterial lesion treatment plan generation system described above, the steps for generating a reference treatment plan include:

[0130] S1, retrieve the current patient's medical record data;

[0131] S2. Based on the above medical record data, select cases and their medical record data that are similar to the current patient from the case database and store them as a reference case database.

[0132] S3. Based on the above reference case database, a first personalized relationship model is generated to reflect the correlation between various data in arterial ultrasound data and arterial angiography data, and to reflect the correlation between various arterial ultrasound data and arterial angiography data and the influencing factors of various diseases.

[0133] S4, perform anomaly screening on the current patient's medical record data:

[0134] S41, determine whether the correlation between various data in the arterial ultrasound data and / or arterial angiography data conforms to the first relationship model; if not, output the first abnormal data.

[0135] S42, determine whether the correlation between various arterial ultrasound data and / or arterial angiography data and the influencing factors of each disease is consistent with the second relationship model. If not, output the second abnormal data.

[0136] S43, determine whether there are missing data in the current patient's medical record data compared with the medical record data of similar cases in the reference case database; if so, output the missing data.

[0137] S44, determine whether the correlation between various data in the arterial ultrasound data and / or arterial angiography data, and the correlation between various arterial ultrasound data and / or arterial angiography data and various disease influencing factors, are consistent with the first personalized relationship model. If they are not consistent, output the third abnormal data.

[0138] S5. Based on the above abnormal data, review the original data and / or obtain new medical record data;

[0139] S6. Determine whether the above-mentioned reference case database needs to be revised and a new reference case database is generated based on the reviewed and / or supplemented medical record data.

[0140] S7. If the reference case database needs to be corrected, repeat steps S4-S6 to screen for abnormal data in the current patient medical record data again until the reference case database is determined.

[0141] S8. Based on the case and medical record data of each case in the revised reference case database, analyze and generate a second personalized relationship model to reflect the correlation between the reference treatment plan and arterial ultrasound data and / or arterial angiography data.

[0142] S9. Based on the screened medical record data, the first reference plan and the second reference plan are generated by combining the third relationship model and the second personalized relationship model, respectively.

[0143] S10, compare the first reference scheme with the second reference scheme and determine the generation of a reference treatment scheme based on the degree of difference, or output prompt information to the interaction unit 600 so that the user can determine the final treatment scheme.

[0144] Although embodiments of this application have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting this application. Those skilled in the art can make changes, modifications, substitutions and variations to the above embodiments within the scope of this application.

Claims

1. An ultrasound and contrast data based treatment plan assistance generation system for arterial lesions, the system comprising: Comprise: A data model storage unit (100) configured to store a first relationship model reflecting the correlation between each item of arterial ultrasound data and each item of arterial angiography data, a second relationship model reflecting the correlation between each item of arterial ultrasound data and each item of arterial angiography data and each disease influencing factor, and a third relationship model reflecting the correlation between a reference treatment plan and each item of arterial ultrasound data and / or each item of arterial angiography data; A data acquisition unit (200) configured to acquire the medical record data of a current patient; A reference case database generation unit (300) configured to connect with a case database, acquire and filter similar cases and their medical record data from the case database according to the medical record data of the current patient, and store them as a reference case database; A data anomaly screening unit (400) configured to receive the medical record data of the current patient and screen and output abnormal data and missing data according to the first relationship model, the second relationship model, and the reference case database; A reference plan generation unit (500) configured to confirm a reference treatment plan according to the medical record data of the current patient and the third relationship model; An interaction unit (600) including data information input and output components, configured to realize data and information interaction between the system and the user; Wherein, the data acquisition unit (200) and the data anomaly screening unit (400) are connected, and in response to the abnormal data or missing data, the relevant medical record data is automatically acquired or reviewed, or a data acquisition request is output through the interaction unit (600) and the medical record data input by the user is monitored; The medical record data includes one or more combinations of each disease influencing factor, arterial ultrasound data, and arterial angiography data; The data anomaly screening unit (400) comprises: A first screening subunit (410) configured to screen and output first abnormal data according to the arterial ultrasound data and / or the arterial angiography data in combination with the first relationship model, and mark it as a first type of abnormality; A second screening subunit (420) configured to screen and output second abnormal data according to each disease influencing factor in combination with the second relationship model, and mark it as a second type of abnormality; A third screening subunit (430) configured to compare and determine third abnormal data and missing data in the medical record data of the current patient according to the reference case database, and mark it as a third type of abnormality; An abnormality prompting subunit (440) configured to prompt the user about the data abnormality through the interaction unit (600) in a set display mode according to the type of the data abnormality.

2. The ultrasound and contrast data-based treatment plan assistance generation system for arterial lesions of claim 1, wherein, The reference case database generation unit (300) comprises: A medical record data analysis subunit (310) configured to analyze one or more of each disease influencing factor, arterial ultrasound data, and arterial angiography data from the medical record data of the current patient; The similarity determination algorithm subunit (320) is configured to match or generate a similarity determination algorithm for determining the similarity between the current patient arterial ultrasound data and / or arterial angiography data and the relevant medical record data in the case database according to at least one specific disease influencing factor in the current patient medical record data; The first matching subunit (330) is configured to match the same or similar cases and their medical record data from the case database according to the analyzed arterial ultrasound data and / or arterial angiography data to form a first intermediate database; The second matching subunit (340) is configured to match the same or similar cases and their medical record data from the case database according to the analyzed disease influencing factors to form a second intermediate database; The database fusion subunit (350) is configured to obtain the data of the first intermediate database and the second intermediate database, and take the intersection of the cases and their medical record data to form the reference case database. The similarity determination algorithm subunit (320) includes: The weight configuration relationship table (3201) is used to store the corresponding relationship between one or more specific disease influencing factors and the similarity determination weight coefficients of each item of data in the arterial ultrasound data and the arterial angiography data; The similarity score relationship table (3202) is used to store the similarity scores corresponding to the data difference ranges in the arterial ultrasound data and the arterial angiography data; The similarity determination module (3203) is configured to obtain the arterial ultrasound data and / or arterial angiography data and the data values contained therein in the current patient medical record data, compare them with the relevant data contained in each case in the case database, obtain the similarity score values of each item of data, and then calculate the similarity according to the similarity determination weight coefficients corresponding to each item of data. The cases and their medical record data in the first intermediate database, the second intermediate database, and the reference case database are stored in order according to the similarity; If the number of cases in the reference case database is less than a set value, the same number or a set proportion of cases and their medical record data are extracted from the first intermediate database and the second intermediate database for data expansion according to the case order until the number of cases in the reference case database reaches the set value.

3. The ultrasound and contrast data-based treatment plan assistance generation system for arterial lesions of claim 2, wherein, The reference case database generation unit (300) further includes: The database correction subunit (360) is configured to obtain the relevant medical record data after review or the medical record data input by the user and temporarily store it as correction data, and find the similarity determination weight coefficients corresponding to the correction data in the weight configuration relationship table (3201); The difference between the correction data and the original data is combined with the similarity determination weight coefficients to calculate the difference between the two similarity scores before and after the correction data. If the difference between the two similarity scores is less than a set value, the reference case database is not updated, otherwise, the similar cases and their medical record data are reselected and found from the case database according to the correction data to form a new reference case database.

4. The ultrasound and contrast data-based treatment plan assistance generation system for arterial lesions of claim 1, wherein, The third screening subunit (430) includes: The data missing checking module (4301) is configured to find missing data in the medical record data of the current patient according to the cases and medical record data in the reference case database, and output the missing data. The first personalized relationship model generating module (4302) is configured to generate a first personalized relationship model reflecting the correlation between the arterial ultrasound data and the arterial angiography data and the correlation between the arterial ultrasound data and the arterial angiography data and the influencing factors of the diseases based on multi-factor analysis or neural network training according to the cases and medical record data in the reference case database. The data anomaly output module (4303) is configured to match and output the third abnormal data according to the medical record data of the current patient and the first personalized relationship model.

5. The ultrasound and contrast data-based treatment plan assistance generation system for arterial lesions of claim 1, wherein, The data acquisition unit (200) comprises: The data retrieval subunit (210) is configured to retrieve and aggregate the medical record data of the current patient in a set format from a data storage area storing the medical record data of the current patient; The data review subunit (220) is configured to receive and respond to the abnormal data output by the data anomaly screening unit (400), retrieve the corresponding original data from the medical record data of the current patient, and verify the data. If the data is consistent, a user confirmation pop-up window is output. If the data is inconsistent, the current abnormal data is corrected and a user confirmation pop-up window is output. The data request subunit (230) is configured to receive and respond to the missing data output by the data anomaly screening unit (400), output a data acquisition request through the interaction unit (600), and monitor the medical record data input by the user feedback; The data storage subunit (250) is configured to store the acquired data.

6. The ultrasound and contrast data-based treatment plan assistance generation system for arterial lesions of claim 5, wherein, The data acquisition unit (200) further comprises: The abnormality degree determination subunit (240) is configured to compare the difference between the three types of abnormal data and the first and second types of abnormal data. If the difference exceeds a set value, a data supplementary sampling request is output to the interaction unit (600); The supplementary sampling data includes intravascular ultrasound data.

7. The ultrasound and contrast data-based treatment plan assistance generation system for arterial lesions of claim 1, wherein, The reference treatment plan includes a treatment plan category and a plan detail parameter; The treatment plan category includes mechanical thrombolysis, thrombus aspiration, balloon angioplasty, stent implantation, combined thrombectomy, and combined surgery. The plan detail parameter includes drug type and dosage, stent length and deployment method, stent placement position, instrument and operation parameter, physiological index monitoring item and monitoring plan, and treatment risk point.

8. The ultrasound and contrast data-based treatment plan assistance generation system for arterial lesions of claim 7, wherein, The reference plan generating unit (500) comprises: The second personalized relationship model generating module (510) is configured to analyze and generate a second personalized relationship model reflecting the correlation between the reference treatment plan and the arterial ultrasound data and / or the arterial angiography data according to the cases and medical record data in the reference case database; The first reference plan generating subunit (520) is configured to receive the medical record data of the current patient and the third relationship model, and output the first reference plan. The second reference scheme generation subunit (530) is configured to receive the medical record data of the current patient and the second personalized relationship model, and determine the output second reference scheme; The reference scheme determination subunit (540) is configured to receive and compare the first reference scheme and the second reference scheme: If the treatment scheme categories contained in the two are the same and the scheme detail parameter deviation is within the set range, the second reference scheme is selected as the reference treatment scheme; If the treatment scheme categories contained in the two are different or the scheme detail parameter deviation exceeds the set range, the first reference scheme, the second reference scheme, and the prompt information are output to the interaction unit (600).

9. The ultrasound and contrast data-based treatment plan assistance generation system for arterial lesions of claim 8, wherein, The data model storage unit (100) also stores a fourth relationship model for reflecting the corresponding relationship between each treatment scheme and its corresponding risk point; The arterial lesion treatment scheme assisted generation system further comprises: The risk locking unit (700) is configured to be in data connection with the reference scheme generation unit (500), obtain the reference treatment scheme, and generate a risk point corresponding to the current reference treatment scheme based on the fourth relationship model and output to the interaction unit (600).

10. The ultrasound and contrast data-based treatment plan assistance generation system for arterial lesions of claim 8, wherein, Based on the arterial lesion treatment scheme assisted generation system, the step of generating a reference treatment scheme comprises: Obtaining the medical record data of the current patient; According to the above medical record data, similar cases and their medical record data are screened from the case database and stored as a reference case database; According to the above reference case database, a first personalized relationship model is analyzed and generated to reflect the correlation between each item of arterial ultrasound data and arterial angiography data, and to reflect the correlation between each item of arterial ultrasound data and arterial angiography data and each disease influencing factor; Abnormal data screening is performed on the medical record data of the current patient: Determine whether the correlation between each item of arterial ultrasound data and / or arterial angiography data conforms to the first relationship model, and if not, output the first abnormal data; Determine whether the correlation between each item of arterial ultrasound data and / or arterial angiography data and each disease influencing factor conforms to the second relationship model, and if not, output the second abnormal data; Determine whether the medical record data of the current patient is missing compared to the medical record data contained in the similar cases in the reference case database, and if so, output the missing data; Determine whether the correlation between each item of arterial ultrasound data and / or arterial angiography data, and the correlation between each item of arterial ultrasound data and / or arterial angiography data and each disease influencing factor conforms to the first personalized relationship model, and if not, output the third abnormal data; According to the above abnormal data, the original data is reviewed and / or new medical record data is obtained; According to the reviewed and / or supplemented medical record data, it is determined whether the above reference case database needs to be corrected and a new reference case database is generated; If the reference case database needs to be corrected, the abnormal data in the medical record data of the current patient is screened again; Based on the cases and the medical record data of the cases in the corrected reference case database, a second personalized relationship model reflecting the correlation between the reference treatment scheme and the arterial ultrasound data and / or the arterial angiography data is analyzed and generated; According to the screened medical record data, the third relationship model and the second personalized relationship model, a first reference scheme and a second reference scheme are respectively generated; The first reference scheme and the second reference scheme are compared, and a reference treatment scheme is determined or prompt information is output to the interaction unit (600) for the user to determine the final treatment scheme according to the difference degree.

Images