A magnetic resonance detection system and method for bile and pancreatic duct imaging

Abstract

The present invention discloses a magnetic resonance detection system and method for bile and pancreatic duct imaging, which belongs to the technical field of bile and pancreatic duct imaging, and comprises an image analysis module and an image detection module; the image analysis module is used to analyze a target image, determine various doubtful problems corresponding to the target image, and calculate problem evaluation values ​​corresponding to the doubtful problems; determine detection interference problems according to the problem evaluation values; the image detection module is used to perform bile and pancreatic duct disease detection on the target image, and determine various target analysis images; perform bile and pancreatic duct disease recognition on the target image and various target analysis images respectively, obtain various disease recognition results, mark the disease recognition result corresponding to the target image as a baseline disease result, and mark the disease recognition result corresponding to the target analysis image as an auxiliary disease result; perform consistency evaluation on the baseline disease result of the target image according to the auxiliary disease results of each target analysis image, and obtain corresponding consistent evaluation values; and mark the consistent evaluation values ​​on the target image.

Description

Technical Field

[0001] The present invention belongs to the technical field of bile and pancreatic duct imaging, and in particular relates to a magnetic resonance detection system and method for bile and pancreatic duct imaging. Background Art

[0002] With the continuous development of medical technology, magnetic resonance imaging (MRI) technology has been widely used in the diagnosis of bile and pancreatic duct diseases. In particular, magnetic resonance cholangiopancreatography (MRCP) technology, as a non-invasive pancreaticobiliary imaging technology, can obtain high-quality pancreaticobiliary images without the use of contrast agents, providing important diagnostic basis for doctors.

[0003] However, during MRI imaging, imaging problems may occur in the image due to the influence of various factors, such as patient movement, breathing, magnetic field inhomogeneity, equipment failure, etc. These imaging problems not only affect the clarity of the image, but may also cause doctors to misjudge the lesion and reduce the accuracy of the diagnosis. Especially in the diagnosis of bile and pancreatic duct diseases, due to the complex lesion site and higher requirements for image quality, the impact of imaging problems is more significant.

[0004] In order to improve the accuracy of MRI in the diagnosis of bile and pancreatic duct diseases, artificial intelligence technology has been widely used in recent years. Through technical means such as deep learning, data analysis and algorithm optimization, automatic analysis and diagnosis of MRI images can be achieved. However, existing intelligent diagnosis systems often ignore the impact of imaging problems on diagnostic results, resulting in large errors in diagnostic results in some cases.

[0005] Based on this, in order to solve the above problems, the present invention provides a magnetic resonance detection system and method for bile and pancreatic duct imaging. Summary of the invention

[0006] In order to solve the problems existing in the above scheme, the present invention provides a magnetic resonance detection system and method for bile and pancreatic duct imaging.

[0007] The purpose of the present invention can be achieved through the following technical solutions:

[0008] A magnetic resonance detection system for bile and pancreatic duct imaging, comprising an image analysis module and an image detection module;

[0009] The image analysis module is used to analyze the target image, determine the suspicious issues corresponding to the target image, and determine the image area corresponding to the suspicious issue, marking it as the suspicious area; obtain the problem attribute of the suspicious issue, calculate the problem value corresponding to the suspicious issue according to the problem attribute; match the corresponding problem interference value according to the suspicious issue;

[0010] According to the formula Calculate the corresponding question evaluation value;

[0011] Where: WQ is the problem evaluation value; WTP is the problem value; YP is the problem interference value; e is a natural constant;

[0012] The detection interference problem is determined according to the problem evaluation value corresponding to each suspicious problem.

[0013] Furthermore, the method for determining doubtful issues includes:

[0014] Establishing a question library, the question library is used to store various imaging problems and the problem attributes, problem interference values ​​and image recognition features corresponding to each imaging problem; establishing a feature recognition model based on the question library, the feature recognition model is used to perform feature recognition on the target image;

[0015] A target image is acquired, and the target image is analyzed by the feature recognition model to obtain image feature values ​​of various imaging problems, and imaging problems with image feature values ​​equal to 0 are marked as doubtful problems.

[0016] Furthermore, the expression of the feature recognition model is: ;

[0017] In the formula: si is the image recognition feature of the corresponding imaging problem, i=1, 2, ..., n, n is a positive integer; the output data is the image feature value 1 or 0.

[0018] Furthermore, the method for calculating the problem value includes:

[0019] Determine each attribute item corresponding to the problem attribute, obtain the patient's status data, process the status data and the problem attribute according to each attribute item, and obtain the status comparison data and attribute item data corresponding to each attribute item;

[0020] Set the attribute judgment model. The expression of the attribute judgment model is: ;

[0021] Where: dj is the state comparison data and attribute item data corresponding to the corresponding attribute item, j=1, 2, ..., m, m is a positive integer; the output data is the attribute judgment value 1 or 0;

[0022] According to the formula Calculate the corresponding attribute value;

[0023] Where: SW is the attribute value; βj is the weight coefficient of the corresponding attribute item;

[0024] Calculate the corresponding question value according to the attribute value and image feature value corresponding to the question;

[0025] The formula for calculating the problem value is WTP = SW × TP´;

[0026] Where: WTP is the problem value; SW is the attribute value; TP´ is the image feature value of the corresponding questionable problem.

[0027] Furthermore, the calculation method of the problem interference value includes:

[0028] Obtain historical detection data corresponding to each imaging problem, and count the influence probability corresponding to the imaging problem according to the historical detection data; set a corresponding influence value for each imaging problem according to a preset influence value matching table, where the value range of the influence value is [1, 10];

[0029] Calculate the corresponding problem interference value according to the formula YP=δ×YZ;

[0030] Where: YP is the problem interference value; δ is the impact probability; YZ is the impact value.

[0031] The image detection module is used to perform bile and pancreatic duct disease detection on the target image, obtain the target image, identify the interference area corresponding to the interference problem in the target image; perform image simulation on the interference area to obtain a plurality of target simulation images; and screen each of the target simulation images to obtain each target analysis image;

[0032] Perform bile and pancreatic duct disease recognition on the target image and each target analysis image respectively, obtain disease recognition results corresponding to the target image and each target analysis image respectively, the disease recognition results include no disease and bile and pancreatic duct disease types, mark the disease recognition result corresponding to the target image as a baseline disease result, and mark the baseline disease result on the target image; mark the disease recognition result corresponding to the target analysis image as an auxiliary disease result; and mark each auxiliary disease result on the corresponding target analysis image;

[0033] The consistency evaluation of the benchmark disease results of the target image is performed according to the attached disease results of each target analysis image to obtain a corresponding consistent evaluation value; and the consistent evaluation value is marked on the target image.

[0034] Furthermore, the method for screening the target simulation image includes:

[0035] Set the associated verification items corresponding to each target simulation image; collect data according to each associated verification item to obtain the data of each associated verification item; establish an associated judgment model;

[0036] Analyze the data of each associated verification item through the associated judgment model to obtain the associated verification value corresponding to each associated verification item;

[0037] The obtained correlation check values ​​are marked as GLz, where z represents the corresponding correlation check item, z=1, 2, ..., v, and v is a positive integer; according to the formula Calculate the corresponding screening value;

[0038] Where: PS is the screening value; bz is the proportional coefficient of the corresponding associated verification item, 0 <bz≤1;

[0039] The target simulation images are screened according to the screening values ​​to obtain analysis images of each target.

[0040] Furthermore, the expression of the association judgment model is ;

[0041] Where: k is the corresponding associated verification item data; the output data is the associated verification value 1 or 0.

[0042] Furthermore, the method of evaluating the consistency of the baseline disease results of the target image according to the attached disease results of each target analysis image includes:

[0043] Compare each subsidiary disease result with the benchmark disease result, and mark the target analysis image with a difference between the subsidiary disease result and the benchmark disease result as a difference image; set a corresponding difference value according to the subsidiary disease result of the difference image, and the value range of the difference value is [1, 2];

[0044] The obtained difference value is marked as CAr, r represents the corresponding difference image, r=1, 2, ..., q, q is a positive integer;

[0045] According to the formula Calculate the corresponding consistent evaluation value;

[0046] Where: YZW is the consistent evaluation value; FL is the number of target analysis images.

[0047] A magnetic resonance detection method for bile and pancreatic duct imaging, the method comprising:

[0048] Determine each suspicious point problem of the target image, and determine the suspicious point area corresponding to the suspicious point problem; obtain the problem attribute of the suspicious point problem, and calculate the problem value corresponding to the suspicious point problem according to the problem attribute; match the corresponding problem interference value according to the suspicious point problem;

[0049] According to the formula Calculate the corresponding question evaluation value;

[0050] Where: WQ is the problem evaluation value; WTP is the problem value; YP is the problem interference value; e is a natural constant;

[0051] Determine the detection interference problem according to the problem evaluation value corresponding to each doubtful problem;

[0052] Acquire a target image, and identify an interference area corresponding to the interference problem in the target image; perform image simulation on the interference area to obtain a number of target simulation images; and screen each target simulation image to obtain each target analysis image;

[0053] Perform bile and pancreatic duct disease recognition on the target image and each target analysis image respectively, obtain disease recognition results corresponding to the target image and each target analysis image respectively, the disease recognition results include no disease and bile and pancreatic duct disease types, mark the disease recognition result corresponding to the target image as a baseline disease result, and mark the baseline disease result on the target image; mark the disease recognition result corresponding to the target analysis image as an auxiliary disease result; and mark each auxiliary disease result on the corresponding target analysis image;

[0054] The consistency of the benchmark disease results of the target image is evaluated according to the subsidiary disease results of each target analysis image to obtain a corresponding consistent evaluation value; and the consistent evaluation value is marked on the target image.

[0055] Compared with the prior art, the present invention has the following beneficial effects:

[0056] Through the cooperation between the image analysis module and the image detection module, intelligent detection of bile and pancreatic duct imaging is achieved, imaging problems in the target image are automatically identified, and the intelligently identified disease types are optimized and analyzed according to the type and degree of the imaging problem, thereby improving the accuracy of intelligent identification; by providing sufficient auxiliary data for doctors to refer to, it helps doctors to judge the patient's condition more accurately; using deep learning algorithms for image analysis and disease diagnosis can achieve rapid processing and analysis of a large number of target images, thereby improving the efficiency and accuracy of diagnosis. BRIEF DESCRIPTION OF THE DRAWINGS

[0057] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0058] Figure 1 This is a principle block diagram of the present invention. DETAILED DESCRIPTION

[0059] The technical solution of the present invention will be clearly and completely described below in conjunction with the embodiments. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0060] like Figure 1 As shown, a magnetic resonance detection system for bile and pancreatic duct imaging includes an image analysis module and an image detection module;

[0061] The image analysis module is used to analyze the bile and pancreatic duct imaging image and mark the bile and pancreatic duct imaging image as the target image; establish a problem library, which is used to store various imaging problems that may occur during the bile and pancreatic duct imaging, mainly problems that have an impact on the identification of bile and pancreatic duct diseases; and set image recognition features and problem attributes for each imaging problem, where the problem attribute is the reason for the image problem, such as fast-flowing blood signals, patient movement, breathing, uneven magnetic field, equipment failure, etc.;

[0062] Identify each imaging problem in the problem library, analyze the impact of each imaging problem on the detection of bile and pancreatic duct diseases, and count the probability of corresponding bile and pancreatic duct disease analysis errors caused by each imaging problem based on a large amount of historical detection data, and mark it as the impact probability; set a corresponding impact value for the corresponding affected bile and pancreatic duct disease, and the impact value is set according to the impact of the bile and pancreatic duct disease analysis error on the patient. The impact value range is [1, 10]. Sort the bile and pancreatic duct diseases from high to low according to the severity, set the impact value of the first one to 10, and the lowest one to 1, and set the corresponding impact values ​​according to the severity difference; it can be set by professionals, and the corresponding impact value matching table can be established by summarizing;

[0063] Calculate the corresponding problem interference value according to the formula YP=δ×YZ; where: YP is the problem interference value; δ is the impact probability; YZ is the impact value;

[0064] The interference value of each question is stored in the question library.

[0065] Based on the existing image recognition technology, a corresponding feature recognition model is established. The feature recognition model is used to identify the target image and determine whether it has the imaging problem in the problem library. The image recognition technology and the image recognition features of each imaging problem are combined to perform recognition and judgment. The expression is: ; Wherein: si is the image recognition feature of the corresponding imaging problem, i=1, 2, ..., n, n is a positive integer; si does not meet the recognition requirements, indicating that there is no image recognition feature corresponding to si in the target image; the output data is the image feature value 1 or 0; when the image feature value is 1, it indicates that the target image has the image recognition feature and the corresponding position can be determined.

[0066] Obtain a target image, analyze the target image through a feature recognition model, obtain image feature values ​​of each imaging problem, and mark the imaging problem with an image feature value equal to 1 as a suspicious problem; determine the image area corresponding to the suspicious problem and mark it as the suspicious area; match the corresponding problem attributes and problem interference values ​​from the problem library according to the image recognition features corresponding to the suspicious area;

[0067] According to the problem attributes corresponding to the doubtful questions, the corresponding attribute values ​​are obtained, that is, the patient's state data during the test is compared and evaluated with the problem attributes. The state data includes the patient's gender, age, heart rate and other data that need to be collected for comparison of the problem data; each data item in the problem attribute is marked as an attribute item, and the corresponding state comparison data is extracted from the state data according to each attribute item, that is, the state data corresponding to the attribute item in the problem attribute is used to analyze whether the state comparison data conforms to the attribute item data corresponding to the attribute item, and the existing data recognition and judgment technology is used to make a judgment, and the attribute judgment model is set. The expression of the attribute judgment model is: ; Where: dj is the input data, j=1, 2, ..., m, m is a positive integer; that is, the state comparison data and attribute item data corresponding to the corresponding attribute item; dj does not meet the judgment requirements means that the state comparison data does not meet the attribute item data requirements; the output data is the attribute judgment value 1 or 0;

[0068] According to the formula Calculate the corresponding attribute value; where: SW is the attribute value; βj is the weight coefficient of the corresponding attribute item. The determination weight of the imaging problem is set according to the attribute item, that is, when the attribute item is clearly met, the probability of the imaging problem can be determined. The weight coefficient is set according to the ratio of each probability. It is generally set by professionals. If it is not set, the default is 1.

[0069] Calculate the corresponding question value according to the attribute value and image feature value corresponding to the question;

[0070] The formula for calculating the problem value is WTP=SW×TP´, where WTP is the problem value, SW is the attribute value, and TP´ is the image feature value of the corresponding questionable problem.

[0071] Match the corresponding problem interference value from the problem database according to the doubtful problem;

[0072] According to the formula Calculate the corresponding problem evaluation value; where: WQ is the problem evaluation value; WTP is the problem value; YP is the problem interference value; e is a natural constant;

[0073] When the problem evaluation value is not greater than the threshold X1, no corresponding operation is performed;

[0074] When the problem evaluation value is greater than the threshold X1, the corresponding suspicious problem is marked as a detection interference problem.

[0075] The image detection module is used to detect bile and pancreatic duct diseases on the target image, obtain the target image, identify the suspicious areas corresponding to each interference problem in the target image, and mark them as interference areas; obtain the possible impact of the interference problem on the target image, such as bile duct obstruction may cause bile stasis, which may cover part of the lesion area; blood signals with fast flow rate may show low signal or no signal, which may make the vascular structure unclear in the image; artifacts may cover the lesion area; analyze the interference area according to the possible impact of the interference problem and the current patient's physical condition and other areas in the target image, and estimate what kind of image the interference area may be; use the existing technology to perform simulation analysis, for example, based on a neural network such as a DNN network to establish a corresponding image simulation model, establish a corresponding training set by artificial means for training, use a large amount of historical detection data for labeling and integration, form possible images corresponding to different interference problems, interference areas, and patient conditions, and set a training set; perform simulation analysis through the image simulation model after successful training, and obtain each target simulation image, which is the target image after image simulation of the interference area. Perform image simulation on the interference area to obtain several target simulation images. And set the associated verification items corresponding to each simulation situation. The associated verification items refer to the aspects from which the simulation situation can be verified, and are mainly set according to the accompanying phenomena of the simulation situation. For example, if the simulation situation corresponding to a target simulation image has a characteristic representation in other positions of the target image, the associated verification items are set according to the characteristics. Specifically, the corresponding associated verification items are set by professionals according to the actual situation of each imaging problem.

[0076] According to each associated check item, the corresponding target simulation image is evaluated and screened to obtain each target analysis image. The target analysis image does not include the target image and is set for simulation based on the interference area;

[0077] The target image and each target analysis image are respectively subjected to bile and pancreatic duct disease recognition, and the disease recognition results corresponding to the target image and each target analysis image are obtained. The disease recognition results include no disease and bile and pancreatic duct disease type. The bile and pancreatic duct disease type is to clarify which bile and pancreatic duct disease it is. Specifically, the target image is recognized by using the existing bile and pancreatic duct disease recognition technology; the disease recognition result corresponding to the target image is marked as the baseline disease result, and the baseline disease result is marked on the target image; the disease recognition result corresponding to the target analysis image is marked as the subsidiary disease result; and each subsidiary disease result is marked on the corresponding target analysis image;

[0078] The consistency of the baseline disease results of the target image is evaluated according to the subsidiary disease results of each target analysis image to obtain a corresponding consistent evaluation value; the consistent evaluation value is marked on the target image, and when the consistent evaluation value is lower than the preset value, a corresponding early warning mark can be made to prompt the doctor.

[0079] Methods for evaluating and screening corresponding target simulation images according to various associated verification items include:

[0080] Data is collected according to each associated verification item to obtain the data of each associated verification item; an associated judgment model is established, and the associated judgment model is used to judge whether the associated verification item data has the characteristics of the associated verification item. If so, it is considered to be an abnormal verification, otherwise it is considered to be normal verification; the expression of the associated judgment model is ; In the formula: k is the input data, is the corresponding associated verification item data; the output data is the associated verification value 1 or 0;

[0081] Analyze the data of each associated verification item through the associated judgment model to obtain the associated verification value corresponding to each associated verification item;

[0082] The obtained correlation check values ​​are marked as GLz, where z represents the corresponding correlation check item, z=1, 2, ..., v, and v is a positive integer; according to the formula Calculate the corresponding screening value;

[0083] Where: PS is the screening value; bz is the proportional coefficient of the corresponding associated verification item, 0 <bz≤1;

[0084] The target simulation images whose screening values ​​are greater than the threshold value X2 are eliminated, and the remaining target simulation images are marked as target analysis images.

[0085] The method for evaluating the consistency of the baseline disease results of the target image based on the attached disease results of each target analysis image includes:

[0086] Compare each subsidiary disease result with the baseline disease result, and mark the target analysis image with a difference between the subsidiary disease result and the baseline disease result as a difference image; set the corresponding difference value according to the subsidiary disease result of the difference image, and the difference value range is [1, 2], which is set according to the degree of difference between the subsidiary disease result and the baseline disease result. If the disease results are directly opposite, the difference value is 2. If there is a disease, when the disease location, size, severity, etc. are different, it is also considered to be different, and the corresponding difference size is determined according to the impact of the disease diagnosis on the patient compared with the target image; in actual applications, various possible disease identification differences will be summarized and sorted from high to low according to the severity, and then the corresponding difference value will be set in combination with the interpolation method; a difference value statistical table will be formed, and then analyzed in combination with the difference value statistical table.

[0087] The obtained difference value is marked as CAr, r represents the corresponding difference image, r=1, 2, ..., q, q is a positive integer;

[0088] According to the formula Calculate the corresponding consistent evaluation value;

[0089] Where: YZW is the consistent evaluation value; FL is the number of target analysis images.

[0090] A magnetic resonance detection method for bile and pancreatic duct imaging, the method comprising:

[0091] Determine each suspicious point problem of the target image, and determine the suspicious point area corresponding to the suspicious point problem; obtain the problem attribute of the suspicious point problem, and calculate the problem value corresponding to the suspicious point problem according to the problem attribute; match the corresponding problem interference value according to the suspicious point problem;

[0092] According to the formula Calculate the corresponding question evaluation value;

[0093] Where: WQ is the problem evaluation value; WTP is the problem value; YP is the problem interference value; e is a natural constant;

[0094] Determine the detection interference problem according to the problem evaluation value corresponding to each doubtful problem;

[0095] Acquire a target image, and identify an interference area corresponding to the interference problem in the target image; perform image simulation on the interference area to obtain a number of target simulation images; and screen each target simulation image to obtain each target analysis image;

[0096] Perform bile and pancreatic duct disease recognition on the target image and each target analysis image respectively, obtain disease recognition results corresponding to the target image and each target analysis image respectively, the disease recognition results include no disease and bile and pancreatic duct disease types, mark the disease recognition result corresponding to the target image as a baseline disease result, and mark the baseline disease result on the target image; mark the disease recognition result corresponding to the target analysis image as an auxiliary disease result; and mark each auxiliary disease result on the corresponding target analysis image;

[0097] The consistency of the benchmark disease results of the target image is evaluated according to the subsidiary disease results of each target analysis image to obtain a corresponding consistent evaluation value; and the consistent evaluation value is marked on the target image.

[0098] The above formulas are all calculated by removing dimensions and taking numerical values. The formula is a formula that is closest to the actual situation obtained by collecting a large amount of data and performing software simulation. The preset parameters and preset thresholds in the formula are set by technical personnel in this field according to actual conditions or obtained by simulating a large amount of data.

[0099] The above embodiments are only used to illustrate the technical method of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical method of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical method of the present invention.

Claims

1. A magnetic resonance detection system for bile and pancreatic duct imaging, characterized in that: It includes an image analysis module and an image detection module; The image analysis module is used to analyze the target image, determine the doubtful issues corresponding to the target image, and determine the image area corresponding to the doubtful issues, and mark it as the doubtful area; Obtaining the question attribute of the doubtful question, and calculating the question value corresponding to the doubtful question according to the question attribute; Match the corresponding problem interference value according to the doubtful question; According to the formula Calculate the corresponding question evaluation value; Where: WQ is the problem evaluation value; WTP is the problem value; YP is the problem interference value; e is a natural constant; Determine the detection interference problem according to the problem evaluation value corresponding to each doubtful problem; The image detection module is used to perform bile and pancreatic duct disease detection on the target image, obtain the target image, identify the interference area corresponding to the interference problem in the target image; perform image simulation on the interference area to obtain a plurality of target simulation images; and screen each of the target simulation images to obtain each target analysis image; Respectively perform bile and pancreatic duct disease recognition on the target image and each target analysis image to obtain disease recognition results corresponding to the target image and each target analysis image, the disease recognition results include no disease and bile and pancreatic duct disease types, mark the disease recognition result corresponding to the target image as a baseline disease result, and mark the baseline disease result on the target image; Marking the disease recognition result corresponding to the target analysis image as an auxiliary disease result; Mark each subsidiary disease result on the corresponding target analysis image; Perform consistency evaluation on the benchmark disease results of the target image according to the subsidiary disease results of each target analysis image to obtain a corresponding consistent evaluation value; marking the consistent evaluation value on the target image; Methods for determining suspicious issues include: Establishing a question library, the question library is used to store various imaging problems and the problem attributes, problem interference values ​​and image recognition features corresponding to each imaging problem; establishing a feature recognition model based on the question library, the feature recognition model is used to perform feature recognition on the target image; Acquire a target image, analyze the target image using the feature recognition model, obtain image feature values ​​of each imaging problem, and mark the imaging problem with an image feature value equal to 1 as a suspicious problem; The expression of the feature recognition model is: ; Where: si is the image recognition feature of the corresponding imaging problem, i=1, 2, ..., n, n is a positive integer; the output data is the image feature value 1 or 0; The calculation method of the problem interference value includes: Obtain historical detection data corresponding to each imaging problem, and count the influence probability corresponding to the imaging problem according to the historical detection data; set a corresponding influence value for each imaging problem according to a preset influence value matching table, where the value range of the influence value is [1, 10]; Calculate the corresponding problem interference value according to the formula YP=δ×YZ; Where: YP is the problem interference value; δ is the impact probability; YZ is the impact value.

2. A magnetic resonance detection system for bile and pancreatic duct imaging according to claim 1, characterized in that: The calculation methods of the problem value include: Determine each attribute item corresponding to the problem attribute, obtain the patient's status data, process the status data and the problem attribute according to each attribute item, and obtain the status comparison data and attribute item data corresponding to each attribute item; Set the attribute judgment model. The expression of the attribute judgment model is: ; Where: dj is the state comparison data and attribute item data corresponding to the corresponding attribute item, j=1, 2, ..., m, m is a positive integer; the output data is the attribute judgment value 1 or 0; According to the formula Calculate the corresponding attribute value; Where: SW is the attribute value; βj is the weight coefficient of the corresponding attribute item; Calculate the corresponding question value according to the attribute value and image feature value corresponding to the question; The formula for calculating the problem value is WTP = SW × TP´; Where: WTP is the problem value; SW is the attribute value; TP´ is the image feature value of the corresponding questionable problem.

3. The magnetic resonance detection system for bile and pancreatic duct imaging according to claim 1, characterized in that: Methods for screening target simulation images include: Set the associated verification items corresponding to each target simulation image; collect data according to each associated verification item to obtain the data of each associated verification item; establish an associated judgment model; Analyze the data of each associated verification item through the associated judgment model to obtain the associated verification value corresponding to each associated verification item; The obtained correlation check values ​​are marked as GLz, where z represents the corresponding correlation check item, z=1, 2, ..., v, and v is a positive integer; according to the formula Calculate the corresponding screening value; Where: PS is the screening value; bz is the proportional coefficient of the corresponding associated verification item, 0 <bz≤1; The target simulation images are screened according to the screening values ​​to obtain analysis images of each target.

4. A magnetic resonance detection system for bile and pancreatic duct imaging according to claim 3, characterized in that: The expression of the association judgment model is: ; Where: k is the corresponding associated verification item data; the output data is the associated verification value 1 or 0.

5. The magnetic resonance detection system for bile and pancreatic duct imaging according to claim 1, characterized in that: The method for evaluating the consistency of the baseline disease results of the target image based on the attached disease results of each target analysis image includes: Compare each subsidiary disease result with the benchmark disease result, and mark the target analysis image with a difference between the subsidiary disease result and the benchmark disease result as a difference image; set a corresponding difference value according to the subsidiary disease result of the difference image, and the value range of the difference value is [1, 2]; The obtained difference value is marked as CAr, r represents the corresponding difference image, r=1, 2, ..., q, q is a positive integer; According to the formula Calculate the corresponding consistent evaluation value; Where: YZW is the consistent evaluation value; FL is the number of target analysis images.

6. A magnetic resonance detection method for bile and pancreatic duct imaging, characterized in that: A magnetic resonance detection system for bile and pancreatic duct imaging as described in any one of claims 1 to 5, the method comprising: Determine each suspicious point problem of the target image, and determine the suspicious point area corresponding to the suspicious point problem; obtain the problem attribute of the suspicious point problem, and calculate the problem value corresponding to the suspicious point problem according to the problem attribute; match the corresponding problem interference value according to the suspicious point problem; According to the formula Calculate the corresponding question evaluation value; Where: WQ is the problem evaluation value; WTP is the problem value; YP is the problem interference value; e is a natural constant; Determine the detection interference problem according to the problem evaluation value corresponding to each doubtful problem; Acquire a target image, and identify an interference area corresponding to the interference problem in the target image; perform image simulation on the interference area to obtain a number of target simulation images; and screen each target simulation image to obtain each target analysis image; Perform bile and pancreatic duct disease recognition on the target image and each target analysis image respectively, obtain disease recognition results corresponding to the target image and each target analysis image respectively, the disease recognition results include no disease and bile and pancreatic duct disease types, mark the disease recognition result corresponding to the target image as a baseline disease result, and mark the baseline disease result on the target image; mark the disease recognition result corresponding to the target analysis image as an auxiliary disease result; and mark each auxiliary disease result on the corresponding target analysis image; The consistency of the benchmark disease results of the target image is evaluated according to the subsidiary disease results of each target analysis image to obtain a corresponding consistent evaluation value; and the consistent evaluation value is marked on the target image.

Images