Liver and gall CT diagnosis report generation method and device

Abstract

The invention is suitable for the technical field of liver and gall CT data processing, and particularly relates to a liver and gall CT diagnosis report generation method and equipment, and the method comprises the steps: obtaining historical diagnosis information and a first liver and gall CT of a patient; according to the first liver and gallbladder CT and the historical liver and gallbladder CT, determining a first abnormal area of liver and gallbladder plain scanning CT and / or a second abnormal area of liver and gallbladder enhanced CT; obtaining abnormal region information according to the first abnormal region and / or the second abnormal region; determining liver and gall text features of the patient according to the abnormal region information and a liver and gall feature library; according to the first abnormal region and / or the second abnormal region, superposing the liver and gall plain-scan CT and the liver and gall enhanced CT to obtain a second liver and gall CT; and obtaining a diagnosis report according to the liver and gall text features and the second liver and gall CT. Therefore, the liver and gall CT diagnosis report generation method provided by the embodiment of the invention can improve the efficiency of generating the liver and gall CT diagnosis report.

Description

Technical Field

[0001] The present application belongs to the technical field of hepatobiliary CT data processing, and in particular, relates to a method and device for generating a hepatobiliary CT diagnostic report. Background Art

[0002] Hepatobiliary CT is a medical imaging method that uses X-rays to scan the liver and gallbladder and forms images through computer processing to evaluate the structure and function of the liver and gallbladder. This examination method can clearly show the internal structure and abnormalities of the liver and gallbladder, such as stones and tumors, and has important reference value for the diagnosis and treatment of liver and gallbladder diseases.

[0003] In the prior art, doctors need to manually analyze a large amount of CT images and data. This means that doctors need to manually analyze CT images, identify lesion areas, and write diagnostic reports based on this information. This process is not only time-consuming, but also significantly increases the workload of doctors when faced with a large number of patients. In the context of tight medical resources, doctors need to deal with a large number of patients and cases, and it is difficult to have enough time and energy to carefully analyze each hepatobiliary CT image and data. This resource shortage further exacerbates the low diagnostic efficiency. Therefore, there is currently a problem of inefficiency in the generation of hepatobiliary CT diagnostic reports. Summary of the invention

[0004] The embodiment of the present application provides a method and device for generating a hepatobiliary CT diagnosis report, which can solve the problem of low efficiency in generating a hepatobiliary CT diagnosis report.

[0005] In a first aspect, an embodiment of the present application provides a method for generating a hepatobiliary CT diagnosis report, comprising: Acquire historical diagnosis information and the patient's first hepatobiliary CT scan; wherein the historical diagnosis information includes a historical diagnosis report and a corresponding historical hepatobiliary CT scan, and the first hepatobiliary CT scan includes a hepatobiliary plain scan CT scan and a hepatobiliary enhanced CT scan; Determine a first abnormal area of ​​the hepatobiliary plain scan CT and / or a second abnormal area of ​​the hepatobiliary enhanced CT according to the first hepatobiliary CT and the historical hepatobiliary CT; Obtaining abnormal region information according to the first abnormal region and / or the second abnormal region; wherein the abnormal region information includes the position, shape and density of the liver and gallbladder in the first abnormal region and / or the second abnormal region; Determining the patient's hepatobiliary text features according to the abnormal region information and the hepatobiliary feature library; wherein the hepatobiliary text features are used to describe the features of the first abnormal region and / or the second abnormal region; According to the first abnormal area and / or the second abnormal area, superimposing the hepatobiliary plain scan CT and the hepatobiliary enhanced CT to obtain a second hepatobiliary CT; A diagnosis report is obtained according to the hepatobiliary text features and the second hepatobiliary CT.

[0006] The above technical solutions in the embodiments of the present application have at least the following technical effects: The method for generating a hepatobiliary CT diagnosis report provided in the embodiment of the present application obtains historical diagnosis information and the patient's first hepatobiliary CT; determines the first abnormal area of ​​the hepatobiliary plain scan CT and / or the second abnormal area of ​​the hepatobiliary enhanced CT according to the first hepatobiliary CT and the historical hepatobiliary CT; obtains abnormal area information according to the first abnormal area and / or the second abnormal area; determines the patient's hepatobiliary text features according to the abnormal area information and the hepatobiliary feature library; according to the first abnormal area and / or the second abnormal area, the hepatobiliary plain scan CT and the hepatobiliary enhanced CT are superimposed to obtain the second hepatobiliary CT; and obtains a diagnosis report according to the hepatobiliary text features and the second hepatobiliary CT. Therefore, the method for generating a hepatobiliary CT diagnosis report provided in the embodiment of the present application realizes the rapid and accurate generation of the hepatobiliary CT diagnosis report by combining the historical diagnosis information and the current first hepatobiliary CT image and utilizing feature extraction and feature matching techniques. This method can not only improve the diagnostic efficiency, but also help improve the accuracy of the diagnosis, provide a powerful auxiliary tool for doctors, and also provide patients with a better medical service experience.

[0007] In a possible implementation manner of the first aspect, the hepatobiliary enhanced CT includes hepatobiliary CT in the arterial phase, portal venous phase, and equilibrium phase, the second abnormal region includes abnormal regions of the hepatobiliary CT in the arterial phase, portal venous phase, and equilibrium phase, and the superimposing the hepatobiliary plain scan CT and the hepatobiliary enhanced CT to obtain the second hepatobiliary CT according to the first abnormal region and / or the second abnormal region includes: spatially aligning the hepatobiliary plain scan CT and the hepatobiliary enhanced CT; Determine rendering rules corresponding to the hepatobiliary plain scan CT and the hepatobiliary enhanced CT respectively according to the first abnormal region and / or the second abnormal region; wherein the rendering rules include color mapping and / or transparency adjustment; Rendering the aligned hepatobiliary plain scan CT and the hepatobiliary enhanced CT according to the rendering rules; Obtaining multiple regional areas according to the abnormal regions of the hepatobiliary CT in the arterial phase, portal venous phase and equilibrium phase of the second abnormal region; The rendered hepatobiliary plain scan CT is used as the underlying image, and the hepatobiliary CT images of the arterial phase, portal venous phase and equilibrium phase in the rendered hepatobiliary enhanced CT are sequentially superimposed in the order of the area of ​​each region from large to small to obtain the second hepatobiliary CT.

[0008] In a possible implementation manner of the first aspect, the method further includes: Obtaining a corresponding region of interest in the historical hepatobiliary CT scan according to the historical diagnosis report; Extracting a corresponding image feature vector according to the region of interest in the historical hepatobiliary CT scan; Obtaining a corresponding text feature vector according to the historical diagnosis report; A liver and gallbladder feature library is obtained according to the image feature vector and the corresponding text feature vector.

[0009] In a possible implementation manner of the first aspect, the historical hepatobiliary CT includes a normal hepatobiliary CT and an abnormal hepatobiliary CT, the normal hepatobiliary CT includes a normal hepatobiliary plain scan CT and a normal hepatobiliary enhanced CT, and determining the first abnormal area of ​​the hepatobiliary plain scan CT and / or the second abnormal area of ​​the hepatobiliary enhanced CT according to the first hepatobiliary CT and the historical hepatobiliary CT includes: Obtaining an abnormality detection model based on the historical hepatobiliary CT scan; Determining whether the first hepatobiliary CT scan is abnormal based on the abnormality detection model; In the case where it is determined based on the abnormality detection model that the first hepatobiliary CT has an abnormality, a plurality of key areas are determined based on the normal hepatobiliary CT and the first hepatobiliary CT; wherein the key areas are important parts of the hepatobiliary structure and function; The first abnormal region and / or the second abnormal region is determined according to each key region of the normal hepatobiliary CT and the key region corresponding to the first hepatobiliary CT.

[0010] In a possible implementation of the first aspect, obtaining the abnormality detection model according to the historical hepatobiliary CT includes: Dividing the historical hepatobiliary CT into a training data set and a validation data set; Training multiple restricted Boltzmann machines layer by layer according to the training data set; Obtaining a DBN model according to each of the restricted Boltzmann machines; wherein the DBN model is formed by stacking a plurality of the restricted Boltzmann machines; The structure and parameters of the DBN model are adjusted according to the verification data set to obtain the anomaly detection model.

[0011] In a possible implementation manner of the first aspect, the method further includes: When it is determined based on the abnormality detection model that the first hepatobiliary CT scan has no abnormality, the diagnosis report is obtained based on the first hepatobiliary CT scan.

[0012] In a possible implementation manner of the first aspect, the hepatobiliary enhanced CT includes hepatobiliary CT in an arterial phase, a portal venous phase, and a balanced phase, and determining the first abnormal region and / or the second abnormal region according to each key region of the normal hepatobiliary CT and the key region corresponding to the first hepatobiliary CT includes: According to the key areas of the normal hepatobiliary CT and the key areas corresponding to the first hepatobiliary CT, for each key area, a first eigenvector is calculated on the hepatobiliary plain scan CT in the first hepatobiliary CT, and a second eigenvector is calculated on the corresponding normal hepatobiliary plain scan CT; Calculating a first distance between each of the first eigenvectors and the corresponding second eigenvector; In a case where the first distance is greater than a first distance threshold, determining a key area in the hepatobiliary plain scan CT corresponding to the first distance being greater than the first distance threshold as the first abnormal area; Obtaining a first change curve of corresponding CT values ​​according to the multiple key areas of the hepatobiliary enhanced CT; Obtaining a normal change curve of the corresponding CT value according to the multiple key areas of the normal hepatobiliary enhanced CT; In the case that the difference between the first change curve and the corresponding normal change curve is greater than a difference threshold, the corresponding key area in the hepatobiliary enhanced CT is determined as the second abnormal area.

[0013] In a possible implementation of the first aspect, determining the hepatobiliary text features of the patient according to the abnormal region information and the hepatobiliary feature library includes: Calculating the similarity between the abnormal region information and each of the image feature vectors in the hepatobiliary feature library to obtain a feature similarity matrix; According to the feature similarity matrix, determining the text feature vector corresponding to the image feature vector having a similarity greater than a similarity threshold as a candidate text feature; The hepatobiliary text feature is determined according to the abnormal region information and each of the candidate text features.

[0014] In a possible implementation manner of the first aspect, obtaining a diagnosis report according to the hepatobiliary text feature and the second hepatobiliary CT includes: The second hepatobiliary CT is annotated according to the hepatobiliary text features, and the diagnosis report is obtained.

[0015] In a second aspect, an embodiment of the present application provides a device for generating a hepatobiliary CT diagnostic report, comprising: An acquisition module, used to acquire historical diagnosis information and a first hepatobiliary CT scan of a patient; wherein the historical diagnosis information includes a historical diagnosis report and a corresponding historical hepatobiliary CT scan, and the first hepatobiliary CT scan includes a hepatobiliary plain scan CT scan and a hepatobiliary enhanced CT scan; An abnormal region module, used for determining a first abnormal region of the hepatobiliary plain scan CT and / or a second abnormal region of the hepatobiliary enhanced CT according to the first hepatobiliary CT and the historical hepatobiliary CT; an abnormal region information module, used to obtain abnormal region information according to the first abnormal region and / or the second abnormal region; wherein the abnormal region information includes the position, shape and density of the liver and gallbladder in the first abnormal region and / or the second abnormal region; A text feature module, used to determine the patient's hepatobiliary text features according to the abnormal region information and the hepatobiliary feature library; wherein the hepatobiliary text features are used to describe the features of the first abnormal region and / or the second abnormal region; A second hepatobiliary CT module, configured to superimpose the hepatobiliary plain scan CT and the hepatobiliary enhanced CT to obtain a second hepatobiliary CT according to the first abnormal region and / or the second abnormal region; A diagnosis report module is used to obtain a diagnosis report based on the hepatobiliary text features and the second hepatobiliary CT.

[0016] In a third aspect, an embodiment of the present application provides a device for generating a hepatobiliary CT diagnostic report, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements a method as described in any one of the first aspects above when executing the computer program.

[0017] In a fourth aspect, an embodiment of the present application provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the method described in any one of the first aspects above is implemented.

[0018] In a fifth aspect, an embodiment of the present application provides a computer program product. When the computer program product is run on a hepatobiliary CT diagnostic report generating device, the hepatobiliary CT diagnostic report generating device executes any of the methods described in the first aspect above.

[0019] It can be understood that the beneficial effects of the second to fifth aspects mentioned above can be found in the relevant description of the first aspect mentioned above, and will not be repeated here. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] In order to more clearly illustrate the technical solutions in the embodiments of the present application, 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 application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0021] Figure 1 It is a flowchart of a method for generating a hepatobiliary CT diagnostic report provided in an embodiment of the present application; Figure 2 It is a schematic diagram of the implementation flow of steps S200, S210 and S240 in the method for generating a hepatobiliary CT diagnostic report provided in an embodiment of the present application; Figure 3 It is a schematic diagram of the implementation flow of steps S400, S500 and S600 in the method for generating a hepatobiliary CT diagnosis report provided in an embodiment of the present application; Figure 4 It is a structural schematic diagram of a hepatobiliary CT diagnostic report generating device provided in an embodiment of the present application; Figure 5 It is a structural schematic diagram of a hepatobiliary CT diagnostic report generating device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0022] In the following description, specific details such as specific system structures, technologies, etc. are provided for the purpose of illustration rather than limitation, so as to provide a thorough understanding of the embodiments of the present application. However, it should be clear to those skilled in the art that the present application may also be implemented in other embodiments without these specific details. In other cases, detailed descriptions of well-known systems, devices, circuits, and methods are omitted to prevent unnecessary details from obstructing the description of the present application.

[0023] It should be understood that when used in the present specification and the appended claims, the term "comprising" indicates the presence of described features, wholes, steps, operations, elements and / or components, but does not exclude the presence or addition of one or more other features, wholes, steps, operations, elements, components and / or combinations thereof.

[0024] It should also be understood that the term “and / or” used in the specification and appended claims refers to any and all possible combinations of one or more of the associated listed items, and includes these combinations.

[0025] As used in the specification and appended claims of this application, the term "if" can be interpreted as "when" or "uponce" or "in response to determining" or "in response to detecting", depending on the context. Similarly, the phrase "if it is determined" or "if [described condition or event] is detected" can be interpreted as meaning "uponce it is determined" or "in response to determining" or "uponce [described condition or event] is detected" or "in response to detecting [described condition or event]", depending on the context.

[0026] In addition, in the description of the present application specification and the appended claims, the terms "first", "second", "third", etc. are only used to distinguish the descriptions and cannot be understood as indicating or implying relative importance.

[0027] References to "one embodiment" or "some embodiments" etc. described in the specification of this application mean that one or more embodiments of the present application include specific features, structures or characteristics described in conjunction with the embodiment. Therefore, the statements "in one embodiment", "in some embodiments", "in some other embodiments", "in some other embodiments", etc. that appear in different places in this specification do not necessarily refer to the same embodiment, but mean "one or more but not all embodiments", unless otherwise specifically emphasized in other ways. The terms "including", "comprising", "having" and their variations all mean "including but not limited to", unless otherwise specifically emphasized in other ways.

[0028] In the related art, doctors need to manually analyze a large number of CT images and data. This means that doctors need to manually analyze CT images, identify lesion areas, and write diagnostic reports based on this information. This process is not only time-consuming, but also significantly increases the workload of doctors when facing a large number of patients. In the context of tight medical resources, doctors need to deal with a large number of patients and cases, and it is difficult to have enough time and energy to carefully analyze each hepatobiliary CT image and data. This resource shortage further exacerbates the low diagnostic efficiency. Therefore, there is currently a problem of inefficiency in the generation of hepatobiliary CT diagnostic reports.

[0029] To solve the above problems, the embodiment of the present application provides a method and device for generating a hepatobiliary CT diagnosis report. In the method, by obtaining historical diagnosis information and the patient's first hepatobiliary CT; determining the first abnormal area of ​​the hepatobiliary plain scan CT and / or the second abnormal area of ​​the hepatobiliary enhanced CT according to the first hepatobiliary CT and the historical hepatobiliary CT; obtaining abnormal area information according to the first abnormal area and / or the second abnormal area; determining the patient's hepatobiliary text features according to the abnormal area information and the hepatobiliary feature library; according to the first abnormal area and / or the second abnormal area, superimposing the hepatobiliary plain scan CT and the hepatobiliary enhanced CT to obtain the second hepatobiliary CT; obtaining a diagnosis report according to the hepatobiliary text features and the second hepatobiliary CT. Therefore, the hepatobiliary CT diagnosis report generation method provided in the embodiment of the present application realizes the rapid and accurate generation of the hepatobiliary CT diagnosis report by combining the historical diagnosis information and the current first hepatobiliary CT image and utilizing feature extraction and feature matching technology. The method can not only improve the diagnostic efficiency, but also help improve the accuracy of the diagnosis, provide a powerful auxiliary tool for doctors, and also provide patients with a better medical service experience.

[0030] The hepatobiliary CT diagnostic report generation method provided in the embodiment of the present application can be applied to a hepatobiliary CT diagnostic report generation device. At this time, the hepatobiliary CT diagnostic report generation device is the executor of the hepatobiliary CT diagnostic report generation method provided in the embodiment of the present application. The embodiment of the present application does not impose any restrictions on the specific type of the hepatobiliary CT diagnostic report generation device.

[0031] For example, the hepatobiliary CT diagnostic report generating device can be a cellular phone, a mobile phone, a tablet computer, an augmented reality (AR) / virtual reality (VR) device, a laptop computer, an ultra-mobile personal computer (UMPC), a netbook, a personal digital assistant (PDA), a desktop computer, a computing device or other processing device connected to a wireless modem, a computer, a laptop computer, a handheld computing device, etc., but not limited to these.

[0032] In order to better understand the method for generating a hepatobiliary CT diagnostic report provided in the embodiment of the present application, the specific implementation process of the method for generating a hepatobiliary CT diagnostic report provided in the embodiment of the present application is exemplarily introduced below.

[0033] Figure 1 A schematic flow chart of a method for generating a hepatobiliary CT diagnosis report provided in an embodiment of the present application is shown. The method for generating a hepatobiliary CT diagnosis report includes: S100, obtaining historical diagnosis information and the patient's first hepatobiliary CT scan, wherein the historical diagnosis information includes historical diagnosis reports and corresponding historical hepatobiliary CT scans, and the first hepatobiliary CT scans include hepatobiliary plain scan CT scans and hepatobiliary enhanced CT scans.

[0034] Exemplarily, the historical diagnosis report, the corresponding historical hepatobiliary CT scan, and the first hepatobiliary CT scan of the current patient may be obtained from the electronic medical record system of the hospital.

[0035] S200, determining a first abnormal area of ​​the hepatobiliary plain scan CT and / or a second abnormal area of ​​the hepatobiliary enhanced scan CT according to the first hepatobiliary CT and the historical hepatobiliary CT.

[0036] For example, the first abnormal area of ​​the hepatobiliary plain scan CT and / or the second abnormal area of ​​the hepatobiliary enhanced CT can be obtained by comparing the first hepatobiliary CT and the historical hepatobiliary CT images through image processing techniques such as image registration and difference detection.

[0037] In one possible implementation, see Figure 2 S200, the historical hepatobiliary CT includes a normal hepatobiliary CT and an abnormal hepatobiliary CT, the normal hepatobiliary CT includes a normal hepatobiliary plain scan CT and a normal hepatobiliary enhanced CT, and the first abnormal area of ​​the hepatobiliary plain scan CT and / or the second abnormal area of ​​the hepatobiliary enhanced CT are determined according to the first hepatobiliary CT and the historical hepatobiliary CT, including: S210, obtaining an abnormality detection model based on historical hepatobiliary CT.

[0038] For example, historical hepatobiliary CT images can be preprocessed, such as normalization and enhancement, and an architecture such as ResNet or U-Net can be used as a model basis. Cross-validation and early stopping methods are used to prevent overfitting to obtain an anomaly detection model.

[0039] Optionally, see Figure 2 , S210, obtain an abnormality detection model based on historical hepatobiliary CT, including: S211, divide the historical hepatobiliary CT into a training data set and a validation data set.

[0040] For example, the historical hepatobiliary CT data can be randomly divided into a training data set and a validation data set according to a certain ratio (such as 80% training set and 20% validation set). For example, the random sampling function in the numpy library or pandas library of Python can be used to implement this.

[0041] S212, training multiple restricted Boltzmann machines layer by layer according to the training data set.

[0042] Exemplarily, the first restricted Boltzmann machine can be trained using the training data set and regarded as a feature extractor to extract high-level features from the training data set. The hidden layer output of the first restricted Boltzmann machine is used as the input of the second restricted Boltzmann machine, and the second restricted Boltzmann machine is continued to be trained, and so on, until a restricted Boltzmann machine network with a preset number of layers is constructed.

[0043] S213, obtaining a DBN model according to each restricted Boltzmann machine, wherein the DBN model is formed by stacking multiple restricted Boltzmann machines.

[0044] For example, the hidden layer output of each restricted Boltzmann machine can be used as the input of the next restricted Boltzmann machine, and they are stacked layer by layer, and a classifier (such as a softmax classifier, a support vector machine, etc.) is added to the top layer to obtain a DBN model.

[0045] S214, adjusting the structure and parameters of the DBN model according to the verification data set to obtain an anomaly detection model.

[0046] Exemplarily, the DBN model can be tested on a validation data set to obtain test results. If it is determined based on the test results that the performance of the DBN model does not reach the ideal level, then adjust the structure of the DBN model (such as increasing or decreasing the number of layers, changing the number of neurons in each layer of the restricted Boltzmann machine, etc.) and parameters (such as learning rate, regularization coefficient, etc.), retrain the DBN model and test it again on the validation data set until. For example, use cross-validation technology to evaluate the performance of the DBN model, and use grid search or random search to find the optimal model structure and parameters. When adjusting the DBN model, gradually change a parameter or structure of the DBN model and observe the change in performance on the validation set. If the performance improves, keep the change; if the performance decreases, return to the previous state. Through continuous iteration and adjustment, an anomaly detection model is obtained.

[0047] Through the above steps S211 to S214, the abnormality detection model obtained through training can more accurately identify abnormal lesions in hepatobiliary CT. The quantitative and qualitative information provided by the model can serve as an important reference for doctors' clinical decision-making and help to formulate more reasonable treatment plans.

[0048] S220, determining whether the first hepatobiliary CT scan has an abnormality based on the abnormality detection model.

[0049] Exemplarily, the first hepatobiliary CT scan may be input into an abnormality detection model, and the probability of being predicted as abnormal may be outputted. A threshold may be set, and whether the first hepatobiliary CT scan is abnormal may be determined based on whether the probability of being predicted as abnormal exceeds the threshold.

[0050] S230, when it is determined based on the abnormality detection model that the first hepatobiliary CT has an abnormality, a plurality of key areas are determined based on the normal hepatobiliary CT and the first hepatobiliary CT, wherein the key areas are important parts of the hepatobiliary structure and function.

[0051] It is understood that the key areas are the five lobes and eight segments of the liver, the gallbladder, the bile ducts, or the intrahepatic vessels.

[0052] Exemplarily, when it is determined that the first hepatobiliary CT image has an abnormality based on the abnormality detection model, key areas such as the five lobes and eight segments of the liver, the gallbladder, the bile duct and / or the intrahepatic blood vessels in the normal hepatobiliary CT image and the first hepatobiliary CT image can be identified based on the normal hepatobiliary CT image and the first hepatobiliary CT image using techniques such as image registration, segmentation, and deep learning segmentation models (such as Mask R-CNN).

[0053] S240: Determine the first abnormal region and / or the second abnormal region according to the key regions of the normal hepatobiliary CT and the key regions corresponding to the first hepatobiliary CT.

[0054] For example, the differences between the key areas of the first hepatobiliary CT and the key areas corresponding to the normal hepatobiliary CT can be compared by using a variety of image processing techniques such as morphological analysis, texture analysis, and volume measurement, and the key areas whose differences exceed a preset threshold are determined as the first abnormal area and / or the second abnormal area. For example, it can be calculated whether the volume changes of the key areas of the first hepatobiliary CT, i.e., the volume changes of the liver segments, exceed a preset threshold, and the key areas that exceed the preset threshold are determined as the first abnormal area or the second abnormal area.

[0055] Through the above steps S210 to S240, it is possible to build an accurate liver and gallbladder abnormality detection model, quickly identify abnormalities, accurately locate key areas, and carefully divide abnormality types, which is conducive to improving diagnostic accuracy, shortening diagnosis time, and improving medical efficiency.

[0056] Optionally, see Figure 2 S240, the hepatobiliary enhanced CT includes the hepatobiliary CT in the arterial phase, the portal venous phase and the equilibrium phase, and the first abnormal area and / or the second abnormal area are determined according to the key areas of the normal hepatobiliary CT and the key areas corresponding to the first hepatobiliary CT, including: S241, according to the key areas of the normal hepatobiliary CT and the key areas corresponding to the first hepatobiliary CT, for each key area, a first eigenvector is calculated on the hepatobiliary plain scan CT in the first hepatobiliary CT, and a second eigenvector is calculated on the corresponding normal hepatobiliary plain scan CT.

[0057] For example, image processing techniques (such as edge detection, texture analysis, shape analysis, etc.) can be used to extract features (including density, size, shape compactness, i.e., the ratio of perimeter to area, and edge smoothness) from each key area of ​​the normal hepatobiliary CT and the key area corresponding to the first hepatobiliary CT, and the extracted features are integrated into a feature vector. For each key area in the first hepatobiliary CT, a first feature vector is obtained; for the corresponding key area in the normal hepatobiliary CT, a second feature vector is obtained. Among them, edge smoothness can be obtained by using edge detection algorithms (such as Sobel, Canny, etc.) to detect the edges of each key area of ​​the normal hepatobiliary CT and the key area corresponding to the first hepatobiliary CT, and calculating the gray value change rate or gradient of the edge pixels.

[0058] S242: Calculate a first distance between each first eigenvector and the corresponding second eigenvector.

[0059] Exemplarily, each first eigenvector and the corresponding second eigenvector can be standardized (such as Min-Max, Z-score and MaxAbs methods), and the distance between the first eigenvector and the corresponding second eigenvector can be calculated by a distance measurement method (such as Euclidean distance, Manhattan distance or cosine similarity, etc.) to obtain multiple first distances. For example, the first eigenvector is: density = 50HU, size = 1000mm², shape compactness = 0.4, edge smoothness = 0.85, and the second eigenvector is: density = 45HU, size = 950mm², shape compactness = 0.38, edge smoothness = 0.90. The first eigenvector and the corresponding second eigenvector are Z-score standardized: After the first eigenvector is standardized: [(50-47.5) / 2.5, (1000-975) / 25, (0.4-0.39) / 0.01, ( 0.85-0.875) / 0.025]=[1,1,0.1 / 0.01,-0.1 / 0.025]=[1,1,10,-4], after the second eigenvector is standardized: [(45-47.5) / 2.5,(950-975) / 25,(0.38-0.39) / 0.01,(0.9-0.875) / 0.025]=[-1,-1,-0.1 / 0.01,0.1 / 0.025]=[-1,-1,-10,4], calculate the Euclidean distance: the first distance = ≈21.72.

[0060] S243: When the first distance is greater than the first distance threshold, determine the key area in the hepatobiliary plain scan CT scan corresponding to the first distance being greater than the first distance threshold as the first abnormal area.

[0061] It can be understood that the first distance threshold is that the first distance does not exceed a preset specified value. The preset specified value can be set by ordinary technicians in this field according to actual needs and is not limited here.

[0062] Exemplarily, a reasonable first distance threshold may be determined based on historical data and industry standards. For example, the first distance threshold is 20. If the distance between the first eigenvector and the corresponding second eigenvector is greater than the first distance threshold, the key area in the corresponding hepatobiliary plain scan CT is determined as the first abnormal area.

[0063] S244, obtaining a first change curve of corresponding CT values ​​according to multiple key areas of the hepatobiliary enhanced CT.

[0064] Exemplarily, for each key area, the CT value is measured in the arterial phase, portal venous phase and equilibrium phase of the hepatobiliary enhanced CT, and multiple first change curves are obtained by plotting the average value of the CT value of each key area in different periods.

[0065] S245, obtaining a normal change curve of the corresponding CT value according to multiple key areas of normal hepatobiliary enhanced CT.

[0066] Exemplarily, for each key area, the CT value is measured in the arterial phase, portal venous phase and equilibrium phase of normal hepatobiliary enhanced CT, and multiple normal change curves are plotted according to the average value of the CT value of each key area in different periods.

[0067] S246: When the difference between the first change curve and the corresponding normal change curve is greater than the difference threshold, the corresponding key area in the hepatobiliary enhanced CT is determined as the second abnormal area.

[0068] It can be understood that the difference threshold is that the difference between the first change curve and the corresponding normal change curve does not exceed a preset specified value. The preset specified value can be set by ordinary technicians in this field according to actual needs and is not limited here.

[0069] For example, a reasonable difference threshold may be determined based on historical data and industry standards, for example, the difference threshold is 5 HU.

[0070] Exemplarily, statistical methods (such as mean square error, correlation coefficient, etc.) can be used to calculate the difference between each first change curve and the corresponding normal change curve. If the difference between a first change curve and the corresponding normal change curve is greater than a difference threshold, the corresponding key area in the hepatobiliary enhanced CT is determined as the second abnormal area.

[0071] Through the above steps S241 to S246, the characteristic vectors of the key areas in the first hepatobiliary CT and the normal hepatobiliary CT are calculated, and the distances therebetween are compared, so that the first abnormal area that is different from the normal state can be accurately identified, and by analyzing the CT value change curves of the key areas in the hepatobiliary enhanced CT and the normal hepatobiliary enhanced CT, the dynamic changes of the second abnormal area can be monitored, which helps doctors to have a more comprehensive understanding of the patient's hepatobiliary condition.

[0072] In one possible implementation, see Figure 2 , S200, the method further includes: S201, when it is determined based on the abnormality detection model that the first hepatobiliary CT scan has no abnormality, a diagnosis report is obtained based on the first hepatobiliary CT scan.

[0073] It is understood that the diagnosis report includes the CT image and the corresponding explanatory text. If the abnormality detection model determines that there is no abnormality in the first hepatobiliary CT, a diagnosis report including the first hepatobiliary CT and the explanatory text (ie, "no abnormality found") is generated.

[0074] Through the above step S201, the abnormality detection model can quickly analyze the first hepatobiliary CT, and automatically generate a diagnosis report when it is determined that there is no abnormality, thereby reducing the time for doctors to review the first hepatobiliary CT, which is conducive to improving the overall quality and efficiency of medical services and providing patients with more timely and accurate diagnosis services.

[0075] S300, obtaining abnormal region information according to the first abnormal region and / or the second abnormal region, wherein the abnormal region information includes the position, shape and density of the liver and gallbladder in the first abnormal region and / or the second abnormal region.

[0076] Exemplarily, the position (i.e., center coordinate value), shape (including area and perimeter) and density (i.e., average CT value) of the liver and gallbladder in the first abnormal region and / or the second abnormal region can be obtained by using an image processing algorithm (such as an edge detection algorithm and a region growing algorithm). For example, if the center coordinate value of a certain abnormal region in the first abnormal region and / or the second abnormal region is (50, 60), the area is 100 square millimeters, the perimeter is 40 millimeters, and the average CT value is 55HU, then the abnormal region information includes [(50, 60), 100, 40, 55].

[0077] In one possible implementation, see Figure 3 , the method further comprises: S001, obtaining a region of interest in the corresponding historical hepatobiliary CT scan according to the historical diagnosis report.

[0078] Exemplarily, image processing software (such as MATLAB, OpenCV in Python, SimpleITK and pydicom, etc.) can be used to pre-process historical hepatobiliary CT images, such as denoising and contrast enhancement, and semi-automatic segmentation algorithms (such as threshold segmentation, region growing, level set method, etc.) can be used to locate and mark the hepatobiliary region as the region of interest in the CT image based on the historical diagnosis report.

[0079] S002, extracting corresponding image feature vectors according to the region of interest in the historical hepatobiliary CT.

[0080] For example, image processing techniques (such as edge detection algorithms and region growing algorithms, etc.) can be used to extract features of regions of interest in historical hepatobiliary CT. The features of the region of interest include position (i.e., center coordinate value), morphology (including area and perimeter), and density (i.e., average CT value). For example, if the center coordinate value of the region of interest is (30, 40), the area is 200 square millimeters, the perimeter is 60 millimeters, and the average CT value is 65HU, then the image feature vector is [(30, 40), 200, 60, 65].

[0081] S003, obtaining a corresponding text feature vector according to the historical diagnosis report.

[0082] Exemplarily, the historical diagnosis report can be preprocessed, including removing stop words, punctuation marks, performing stem extraction or lemmatization, and applying the NLP model to convert the text into a numerical feature vector. For example, the historical diagnosis report contains keywords such as "enlarged liver", "blurred edges", and "uneven density". Using the TF-IDF method, each keyword will be assigned a weight based on its frequency of occurrence in the document and its inverse document frequency in the entire corpus: "liver": 0.3, "enlarged": 0.4, "edge": 0.2, "blurred": 0.5, "density": 0.3, "uneven": 0.6, and these weights are combined to obtain a text feature vector representing the diagnosis report [0.3, 0.4, 0.2, 0.5, 0.3, 0.6].

[0083] S004, obtaining a liver and gallbladder feature library according to the image feature vector and the corresponding text feature vector.

[0084] Exemplarily, the image feature vector and the corresponding text feature vector can be combined to form a hepatobiliary feature library containing image and text information. The hepatobiliary feature library can be organized in the form of a database, a CSV file, or any form suitable for storing and retrieving large amounts of numerical data. For example, there are 1,000 historical hepatobiliary CT images and their corresponding diagnostic reports, 5 image features are extracted from each image, and each report is converted into 10 text features. Then the hepatobiliary feature library will be a matrix of 1,000 rows and 15 columns, where each row represents a comprehensive feature vector of a patient (5 image features + 10 text features).

[0085] Through the above steps S001 to S004, the image feature vector is combined with the text feature vector, which provides doctors with a decision-making assistance tool that can interpret and analyze hepatobiliary CT images more quickly, helping to reduce the time required for diagnosis.

[0086] S400, determining the patient's hepatobiliary text features according to the abnormal region information and the hepatobiliary feature library, wherein the hepatobiliary text features are used to describe the features of the first abnormal region and / or the second abnormal region.

[0087] Exemplarily, the similarity between the information of the first abnormal region and / or the second abnormal region in the abnormal region information and each image feature vector in the hepatobiliary feature library can be calculated, and the text feature vector corresponding to the image feature vector with the greatest similarity can be determined as the hepatobiliary text feature.

[0088] In one possible implementation, see Figure 3 S400, determining the patient's hepatobiliary text features according to the abnormal region information and the hepatobiliary feature library, including: S410, calculating the similarity between the abnormal region information and each image feature vector in the hepatobiliary feature library to obtain a feature similarity matrix.

[0089] For example, the similarity can be calculated by distance measurement methods such as cosine similarity and Euclidean distance, and each image feature vector in the hepatobiliary feature library is traversed to calculate the similarity between the abnormal region information and each image feature vector, and the result is stored in a two-dimensional matrix, namely, a feature similarity matrix. Among them, each row of the feature similarity matrix represents a feature vector in the hepatobiliary feature library, each column represents a dimension of the abnormal region feature, and the value in the feature similarity matrix represents the similarity size.

[0090] S420: According to the feature similarity matrix, determine the text feature vector corresponding to the image feature vector having a similarity greater than a similarity threshold as a candidate text feature.

[0091] It can be understood that the similarity threshold is that the similarity does not exceed a preset specified value, and the preset specified value can be set by ordinary technicians in the field according to actual needs, and is not limited here.

[0092] For example, a reasonable similarity threshold can be determined based on historical data and industry standards. For example, the similarity threshold is set to 0.8, the feature similarity matrix is ​​traversed, all image feature vectors with similarities greater than 0.8 are found, and the text feature vectors corresponding to the image feature vectors with similarities greater than 0.8 are used as candidate text features.

[0093] S430, determining the liver and gallbladder text features according to the abnormal region information and each candidate text feature.

[0094] For example, the abnormal region information and candidate text features can be combined to generate hepatobiliary text features using natural language processing (NLP) technology using keywords and phrases. For example, the abnormal region information is: the center coordinate value is (50, 60), the area is 100 square millimeters, the perimeter is 40 millimeters, and the average CT value is 55HU. The candidate text features include "right lobe of liver", "irregular shape" and "unclear boundaries". Then the hepatobiliary text feature is "there is an abnormality in the right lobe of liver (50, 60), 100mm 2 Irregular shape and size, circumference of 40 mm with unclear borders, average CT value of 55 HU".

[0095] Through the above steps S410 to S430, by accurately matching the abnormal area with the hepatobiliary feature library and performing comprehensive analysis in combination with text features, the diagnostic accuracy of hepatobiliary diseases can be improved, which helps doctors to judge the condition more accurately.

[0096] S500: According to the first abnormal region and / or the second abnormal region, the hepatobiliary plain scan CT and the hepatobiliary enhanced CT are superimposed to obtain a second hepatobiliary CT.

[0097] Exemplarily, the hepatobiliary plain scan CT and the hepatobiliary enhanced CT may be superimposed by image fusion technology (such as Alpha fusion, pyramid fusion, wavelet fusion, discrete cosine transform fusion, etc.) to obtain a second hepatobiliary CT image.

[0098] In one possible implementation, see Figure 3 S500, the hepatobiliary enhanced CT includes the hepatobiliary CT in the arterial phase, the portal venous phase and the equilibrium phase, the second abnormal area includes the abnormal area of ​​the hepatobiliary CT in the arterial phase, the portal venous phase and the equilibrium phase, and according to the first abnormal area and / or the second abnormal area, the hepatobiliary plain scan CT and the hepatobiliary enhanced CT are superimposed to obtain the second hepatobiliary CT, including: S510, spatially aligning the hepatobiliary plain scan CT and the hepatobiliary enhanced CT.

[0099] Exemplarily, the feature points or edge information of the hepatobiliary plain scan CT and the hepatobiliary enhanced CT can be extracted by rigid body transformation, affine transformation or non-rigid transformation, and image registration (such as iterative nearest point algorithm, mutual information maximization method, etc.) can be performed based on the feature points or edge information of the hepatobiliary plain scan CT and the hepatobiliary enhanced CT, that is, the optimal transformation parameters between the two images are found so that the transformed enhanced CT image is completely aligned with the plain scan CT image in space.

[0100] S520: Determine rendering rules corresponding to the hepatobiliary plain scan CT and the hepatobiliary enhanced CT respectively according to the first abnormal region and / or the second abnormal region, wherein the rendering rules include color mapping and / or transparency adjustment.

[0101] Exemplarily, for the first abnormal area in the hepatobiliary plain scan CT and / or the second abnormal area in the hepatobiliary CT in the arterial phase, portal venous phase and equilibrium phase in the hepatobiliary enhanced CT, different color mapping schemes can be used to highlight different abnormal areas, and the rendering rules can be dynamically adjusted according to information such as the area size, shape and position of the abnormal area. For example, the larger the area of ​​the abnormal area, the greater the transparency; the irregularly shaped abnormal area uses color gradient or edge processing to highlight its outline; the abnormal area located at the edge of the liver uses a color with greater contrast.

[0102] S530: Render the aligned hepatobiliary plain scan CT and hepatobiliary enhanced CT according to various rendering rules.

[0103] Exemplarily, the aligned hepatobiliary plain scan CT and hepatobiliary enhanced CT can be rendered according to various rendering rules, and each pixel point is traversed to determine whether it belongs to an abnormal area based on its position, intensity and other information. For pixels in different areas, corresponding color mapping and transparency adjustment are performed according to the corresponding rendering rules.

[0104] S540, obtaining multiple area areas according to the abnormal regions of the hepatobiliary CT in the arterial phase, portal venous phase, and equilibrium phase in the second abnormal region.

[0105] Exemplarily, an image segmentation algorithm can be used to separate the abnormal areas of hepatobiliary CT in the arterial phase, portal venous phase and equilibrium phase in the second abnormal area from the hepatobiliary CT, extract the contours of each abnormal area, calculate the size of the area surrounded by the contour, and obtain the areas of multiple regions.

[0106] S550, using the rendered hepatobiliary plain scan CT as the bottom image, and superimposing the hepatobiliary CT images of the arterial phase, portal venous phase, and equilibrium phase in the rendered hepatobiliary enhanced CT in descending order of area to obtain a second hepatobiliary CT.

[0107] Exemplarily, the rendered hepatobiliary plain scan CT can be used as the underlying image, and the abnormal areas of the hepatobiliary CT in the arterial phase, portal venous phase, and equilibrium phase are sorted according to the size of the area, and the abnormal areas of the hepatobiliary CT in the arterial phase, portal venous phase, and equilibrium phase are partially superimposed on the underlying image in order from large to small using an image fusion algorithm (such as pixel-level image fusion, image pyramid fusion algorithm, and wavelet transform fusion algorithm, etc.). The brightness of the hepatobiliary CT in the arterial phase, portal venous phase, and equilibrium phase can be adjusted as needed, for example, the larger the area of ​​the abnormal area, the higher the brightness.

[0108] Through the above steps S510 to S550, the rendered plain scan CT of the liver and gallbladder is used as the bottom image, and the liver and gallbladder CT images of different periods in the rendered enhanced CT of the liver and gallbladder are sequentially superimposed in order of area from large to small, thereby obtaining a second liver and gallbladder CT image containing comprehensive information. Through the superposition process, the doctor can more intuitively see the specific manifestations of the abnormal area under plain scan and enhancement, thereby more accurately judging the nature and scope of the lesion, which helps the doctor to comprehensively analyze and compare the lesions at different periods, and improves the comprehensiveness and accuracy of the diagnosis.

[0109] S600, obtain a diagnosis report based on the hepatobiliary text features and the second hepatobiliary CT.

[0110] It can be understood that the second hepatobiliary CT and hepatobiliary text features are integrated to obtain a diagnosis report, so that the diagnosis report contains both intuitive image assistance and detailed text description.

[0111] In one possible implementation, see Figure 3 , S600, based on the hepatobiliary text features and the second hepatobiliary CT, a diagnosis report is obtained, including: S610, annotate the second hepatobiliary CT according to the hepatobiliary text features, and obtain a diagnosis report.

[0112] Exemplarily, the hepatobiliary text feature can be associated with the abnormal area in the second hepatobiliary CT image to obtain a diagnosis report. For example, the hepatobiliary text feature is "there is an abnormality in the right lobe of the liver (50, 60), with an irregular shape of 100 mm2, a perimeter of 40 mm and unclear boundaries, and an average CT value of 55 HU", then at the coordinates (50, 60), a text annotation is added next to the second hepatobiliary CT image, where the text annotation is "100 mm2 irregular shape, a perimeter of 40 mm and unclear boundaries, and an average CT value of 55 HU", and the second hepatobiliary CT with the added text is determined as a diagnosis report.

[0113] Through the above step S610, the hepatobiliary text features are fused with the second hepatobiliary CT, so that the text information and the image information are mutually supplemented and verified, which helps doctors to understand the condition more comprehensively and improve the accuracy of diagnosis.

[0114] It should be understood that the size of the serial numbers of the steps in the above embodiments does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present application.

[0115] Corresponding to the method for generating a hepatobiliary CT diagnostic report described in the above embodiment, the embodiment of the present application also provides a hepatobiliary CT diagnostic report generating device, and each module of the device can implement each step of the hepatobiliary CT diagnostic report generating method. Figure 4 A structural block diagram of a hepatobiliary CT diagnostic report generating device provided in an embodiment of the present application is shown. For ease of explanation, only the parts related to the embodiment of the present application are shown.

[0116] Reference Figure 4 , the device comprises: An acquisition module, used to acquire historical diagnosis information and a first hepatobiliary CT scan of a patient; wherein the historical diagnosis information includes a historical diagnosis report and a corresponding historical hepatobiliary CT scan, and the first hepatobiliary CT scan includes a hepatobiliary plain scan CT scan and a hepatobiliary enhanced CT scan; An abnormal region module, used for determining a first abnormal region of the hepatobiliary plain scan CT and / or a second abnormal region of the hepatobiliary enhanced CT according to the first hepatobiliary CT and the historical hepatobiliary CT; an abnormal region information module, used to obtain abnormal region information according to the first abnormal region and / or the second abnormal region; wherein the abnormal region information includes the position, shape and density of the liver and gallbladder in the first abnormal region and / or the second abnormal region; A text feature module, used to determine the patient's hepatobiliary text features according to the abnormal region information and the hepatobiliary feature library; wherein the hepatobiliary text features are used to describe the features of the first abnormal region and / or the second abnormal region; A second hepatobiliary CT module, configured to superimpose the hepatobiliary plain scan CT and the hepatobiliary enhanced CT to obtain a second hepatobiliary CT according to the first abnormal region and / or the second abnormal region; A diagnosis report module is used to obtain a diagnosis report based on the hepatobiliary text features and the second hepatobiliary CT.

[0117] In some embodiments, the hepatobiliary enhanced CT includes hepatobiliary CT in the arterial phase, portal venous phase, and equilibrium phase, the second abnormal area includes an abnormal area of ​​hepatobiliary CT in the arterial phase, portal venous phase, and equilibrium phase, and the second hepatobiliary CT module includes: An alignment module, used for spatially aligning the hepatobiliary plain scan CT and the hepatobiliary enhanced CT; A rendering rule module, used to determine the rendering rules corresponding to the hepatobiliary plain scan CT and the hepatobiliary enhanced CT respectively according to the first abnormal area and / or the second abnormal area; wherein the rendering rules include color mapping and / or transparency adjustment; A rendering module, used for rendering the aligned hepatobiliary plain scan CT and the hepatobiliary enhanced CT according to the rendering rules; A regional area module, used for obtaining multiple regional areas according to the abnormal areas of the hepatobiliary CT in the arterial phase, portal venous phase and equilibrium phase in the second abnormal area; The superposition module is used to use the rendered hepatobiliary plain scan CT as the underlying image, and to superimpose the hepatobiliary CT in the arterial phase, portal venous phase and equilibrium phase in the rendered hepatobiliary enhanced CT in the order of the area of ​​each region from large to small to obtain the second hepatobiliary CT.

[0118] In some embodiments, the hepatobiliary CT diagnostic report generating device further comprises: A region of interest module, used to obtain the corresponding region of interest in the historical hepatobiliary CT according to the historical diagnosis report; An image feature vector module, used for extracting a corresponding image feature vector according to a region of interest in the historical hepatobiliary CT scan; A text feature vector module, used to obtain a corresponding text feature vector according to the historical diagnosis report; The hepatobiliary feature module is used to obtain a hepatobiliary feature library according to the image feature vector and the corresponding text feature vector.

[0119] In some embodiments, the historical hepatobiliary CT includes normal hepatobiliary CT and abnormal hepatobiliary CT, the normal hepatobiliary CT includes normal hepatobiliary plain scan CT and normal hepatobiliary enhanced CT, and the abnormal area module includes: An abnormality detection model module, used to obtain an abnormality detection model based on the historical hepatobiliary CT; an abnormality detection module, used to determine whether the first hepatobiliary CT scan has an abnormality based on the abnormality detection model; A key area module, used for determining a plurality of key areas according to the normal hepatobiliary CT and the first hepatobiliary CT when it is determined that the first hepatobiliary CT has an abnormality based on the abnormality detection model; wherein the key areas are important parts of the hepatobiliary structure and function; The abnormal region generating module is used to determine the first abnormal region and / or the second abnormal region according to the key regions of the normal hepatobiliary CT and the key regions corresponding to the first hepatobiliary CT.

[0120] In some embodiments, the anomaly detection model module includes: A data set module, used for dividing the historical hepatobiliary CT into a training data set and a validation data set; A training module, used for training multiple restricted Boltzmann machines layer by layer according to the training data set; A DBN model module, used to obtain a DBN model according to each of the restricted Boltzmann machines; wherein the DBN model is formed by stacking multiple restricted Boltzmann machines; The adjustment module is used to adjust the structure and parameters of the DBN model according to the verification data set to obtain the anomaly detection model.

[0121] In some embodiments, the hepatobiliary CT diagnostic report generating device further comprises: A diagnosis report generation module is used to obtain the diagnosis report according to the first hepatobiliary CT when it is determined that the first hepatobiliary CT has no abnormality based on the abnormality detection model.

[0122] In some embodiments, the hepatobiliary enhanced CT includes hepatobiliary CT in the arterial phase, portal venous phase, and equilibrium phase, and the abnormal region generation module includes: A feature vector module, for calculating, based on each key area of ​​the normal hepatobiliary CT and the key area corresponding to the first hepatobiliary CT, for each key area, a first feature vector on the hepatobiliary plain scan CT in the first hepatobiliary CT, and a second feature vector on the corresponding normal hepatobiliary plain scan CT; A first distance module, used for calculating a first distance between each of the first feature vectors and the corresponding second feature vector; A first abnormal region module, configured to determine, when the first distance is greater than a first distance threshold, a key region in the hepatobiliary plain scan CT corresponding to the first distance being greater than the first distance threshold as the first abnormal region; A first change curve module, used to obtain a first change curve of corresponding CT values ​​according to a plurality of key areas of the hepatobiliary enhanced CT; A normal change curve module, used for obtaining a normal change curve of a corresponding CT value according to a plurality of key areas of the normal hepatobiliary enhanced CT; The second abnormal area module is used to determine the corresponding key area in the hepatobiliary enhanced CT as the second abnormal area when the difference between the first change curve and the corresponding normal change curve is greater than a difference threshold.

[0123] In some embodiments, the text feature module includes: A feature similarity matrix module is used to calculate the similarity between the abnormal region information and each of the image feature vectors in the hepatobiliary feature library to obtain a feature similarity matrix; A candidate text feature module, configured to determine, according to the feature similarity matrix, the text feature vector corresponding to the image feature vector having a similarity greater than a similarity threshold as a candidate text feature; The hepatobiliary text feature module is used to determine the hepatobiliary text feature according to the abnormal area information and each of the candidate text features.

[0124] In some embodiments, the diagnostic reporting module comprises: A labeling module is used to label the second hepatobiliary CT according to the hepatobiliary text features and obtain the diagnosis report.

[0125] It should be noted that the information interaction, execution process and other contents between the above-mentioned modules are based on the same concept as the method embodiment of the present application. Their specific functions and technical effects can be found in the method embodiment part and will not be repeated here.

[0126] The technicians in the relevant field can clearly understand that for the convenience and simplicity of description, only the division of the above-mentioned functional units and modules is used as an example for illustration. In practical applications, the above-mentioned function allocation can be completed by different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above. The functional units and modules in the embodiment can be integrated in a processing unit, or each unit can exist physically separately, or two or more units can be integrated in one unit. The above-mentioned integrated unit can be implemented in the form of hardware or in the form of software functional units. In addition, the specific names of the functional units and modules are only for the convenience of distinguishing each other, and are not used to limit the scope of protection of this application. The specific working process of the units and modules in the above-mentioned device can refer to the corresponding process in the aforementioned method embodiment, which will not be repeated here.

[0127] The present application also provides a device for generating a hepatobiliary CT diagnosis report. Figure 5 This is a schematic diagram of the structure of a device for generating a hepatobiliary CT diagnostic report provided in an embodiment of the present application. Figure 5 As shown, the hepatobiliary CT diagnostic report generating device 6 of this embodiment includes: at least one processor 60 ( Figure 5 Only one is shown), at least one memory 61 ( Figure 5Only one is shown in the figure) and a computer program 62 stored in the at least one memory 61 and executable on the at least one processor 60. When the processor 60 executes the computer program 62, the hepatobiliary CT diagnosis report generation device 6 implements the steps in any of the above-mentioned hepatobiliary CT diagnosis report generation method embodiments, or implements the functions of each module / unit in the above-mentioned device embodiments.

[0128] Exemplarily, the computer program 62 may be divided into one or more modules / units, which are stored in the memory 61 and executed by the processor 60 to complete the present application. The one or more modules / units may be a series of computer program instruction segments capable of completing specific functions, which are used to describe the execution process of the computer program 62 in the hepatobiliary CT diagnostic report generation device 6.

[0129] The hepatobiliary CT diagnosis report generation device 6 can be a computing device such as a desktop computer, a notebook, a PDA, and a cloud server. The hepatobiliary CT diagnosis report generation device can include, but is not limited to, a processor 60 and a memory 61. Those skilled in the art can understand that Figure 5 It is only an example of the hepatobiliary CT diagnostic report generating device 6 and does not constitute a limitation on the hepatobiliary CT diagnostic report generating device 6. It may include more or fewer components than shown in the figure, or a combination of certain components, or different components, for example, it may also include input and output devices, network access devices, buses, etc.

[0130] The processor 60 may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor may be a microprocessor or any conventional processor, etc.

[0131] In some embodiments, the memory 61 may be an internal storage unit of the hepatobiliary CT diagnosis report generation device 6, such as a hard disk or memory of the hepatobiliary CT diagnosis report generation device 6. In other embodiments, the memory 61 may also be an external storage device of the hepatobiliary CT diagnosis report generation device 6, such as a plug-in hard disk, a smart memory card (Smart Media Card, SMC), a secure digital (Secure Digital, SD) card, a flash card (Flash Card), etc. equipped on the hepatobiliary CT diagnosis report generation device 6. Further, the memory 61 may also include both the internal storage unit of the hepatobiliary CT diagnosis report generation device 6 and an external storage device. The memory 61 is used to store an operating system, an application program, a boot loader (BootLoader), data, and other programs, such as the program code of the computer program, etc. The memory 61 may also be used to temporarily store data that has been output or is to be output.

[0132] An embodiment of the present application further provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps in any of the above method embodiments are implemented.

[0133] An embodiment of the present application provides a computer program product. When the computer program product runs on a hepatobiliary CT diagnosis report generation device, the hepatobiliary CT diagnosis report generation device implements the steps in any of the above-mentioned method embodiments.

[0134] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the present application implements all or part of the processes in the above-mentioned embodiment method, which can be completed by instructing the relevant hardware through a computer program. The computer program can be stored in a computer-readable storage medium, and the computer program can implement the steps of the above-mentioned various method embodiments when executed by the processor. Among them, the computer program includes computer program code, and the computer program code can be in source code form, object code form, executable file or some intermediate form. The computer-readable medium may at least include: any entity or device that can carry the computer program code to the hepatobiliary CT diagnostic report generation device, recording medium, computer memory, read-only memory (ROM, Read-Only Memory), random access memory (RAM, RandomAccess Memory), electric carrier signal, telecommunication signal and software distribution medium. For example, a USB flash drive, a mobile hard disk, a magnetic disk or an optical disk. In some jurisdictions, according to legislation and patent practice, computer-readable media cannot be electric carrier signals and telecommunication signals.

[0135] In the above embodiments, the description of each embodiment has its own emphasis. For parts that are not described or recorded in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0136] Those of ordinary skill in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.

[0137] In the embodiments provided in the present application, it should be understood that the disclosed hepatobiliary CT diagnostic report generation device and method can be implemented in other ways. For example, the above-described hepatobiliary CT diagnostic report generation device embodiment is only schematic. For example, the division of the modules or units is only a logical function division. There may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be an indirect coupling or communication connection through some interfaces, devices or units, which can be electrical, mechanical or other forms.

[0138] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0139] The embodiments described above are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the aforementioned embodiments, a person skilled in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features may be replaced by equivalents. Such modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present application, and should all be included in the protection scope of the present application.

Claims

1. A method for generating a hepatobiliary CT diagnostic report, characterized in that: include: Acquire historical diagnosis information and the patient's first hepatobiliary CT scan; wherein the historical diagnosis information includes a historical diagnosis report and a corresponding historical hepatobiliary CT scan, and the first hepatobiliary CT scan includes a hepatobiliary plain scan CT scan and a hepatobiliary enhanced CT scan; Determine a first abnormal area of ​​the hepatobiliary plain scan CT and / or a second abnormal area of ​​the hepatobiliary enhanced CT according to the first hepatobiliary CT and the historical hepatobiliary CT; Obtaining abnormal region information according to the first abnormal region and / or the second abnormal region; wherein the abnormal region information includes the position, shape and density of the liver and gallbladder in the first abnormal region and / or the second abnormal region; Determining the patient's hepatobiliary text features according to the abnormal region information and the hepatobiliary feature library; wherein the hepatobiliary text features are used to describe the features of the first abnormal region and / or the second abnormal region; According to the first abnormal area and / or the second abnormal area, superimposing the hepatobiliary plain scan CT and the hepatobiliary enhanced CT to obtain a second hepatobiliary CT; A diagnosis report is obtained according to the hepatobiliary text features and the second hepatobiliary CT.

2. The method for generating a hepatobiliary CT diagnostic report according to claim 1, characterized in that: The hepatobiliary enhanced CT includes hepatobiliary CT in the arterial phase, portal venous phase and equilibrium phase, the second abnormal area includes abnormal areas of the hepatobiliary CT in the arterial phase, portal venous phase and equilibrium phase, and the superposition of the hepatobiliary plain scan CT and the hepatobiliary enhanced CT to obtain the second hepatobiliary CT according to the first abnormal area and / or the second abnormal area includes: spatially aligning the hepatobiliary plain scan CT and the hepatobiliary enhanced CT; Determine rendering rules corresponding to the hepatobiliary plain scan CT and the hepatobiliary enhanced CT respectively according to the first abnormal region and / or the second abnormal region; wherein the rendering rules include color mapping and / or transparency adjustment; Rendering the aligned hepatobiliary plain scan CT and the hepatobiliary enhanced CT according to the rendering rules; Obtaining multiple regional areas according to the abnormal regions of the hepatobiliary CT in the arterial phase, portal venous phase and equilibrium phase of the second abnormal region; The rendered hepatobiliary plain scan CT is used as the underlying image, and the hepatobiliary CT images of the arterial phase, portal venous phase and equilibrium phase in the rendered hepatobiliary enhanced CT are sequentially superimposed in the order of the area of ​​each region from large to small to obtain the second hepatobiliary CT.

3. The method for generating a hepatobiliary CT diagnostic report according to claim 1, characterized in that: The method further comprises: Obtaining a corresponding region of interest in the historical hepatobiliary CT scan according to the historical diagnosis report; Extracting a corresponding image feature vector according to the region of interest in the historical hepatobiliary CT scan; Obtaining a corresponding text feature vector according to the historical diagnosis report; A liver and gallbladder feature library is obtained according to the image feature vector and the corresponding text feature vector.

4. The method for generating a hepatobiliary CT diagnosis report according to claim 1, characterized in that: The historical hepatobiliary CT includes a normal hepatobiliary CT and an abnormal hepatobiliary CT, the normal hepatobiliary CT includes a normal hepatobiliary plain scan CT and a normal hepatobiliary enhanced CT, and determining the first abnormal area of ​​the hepatobiliary plain scan CT and / or the second abnormal area of ​​the hepatobiliary enhanced CT according to the first hepatobiliary CT and the historical hepatobiliary CT includes: Obtaining an abnormality detection model based on the historical hepatobiliary CT scan; Determining whether the first hepatobiliary CT scan is abnormal based on the abnormality detection model; In the case where it is determined based on the abnormality detection model that the first hepatobiliary CT has an abnormality, a plurality of key areas are determined based on the normal hepatobiliary CT and the first hepatobiliary CT; wherein the key areas are important parts of the hepatobiliary structure and function; The first abnormal region and / or the second abnormal region is determined according to each key region of the normal hepatobiliary CT and the key region corresponding to the first hepatobiliary CT.

5. The method for generating a hepatobiliary CT diagnosis report according to claim 4, characterized in that: The abnormality detection model is obtained according to the historical hepatobiliary CT scan, including: Dividing the historical hepatobiliary CT into a training data set and a validation data set; Training multiple restricted Boltzmann machines layer by layer according to the training data set; Obtaining a DBN model according to each of the restricted Boltzmann machines; wherein the DBN model is formed by stacking a plurality of the restricted Boltzmann machines; The structure and parameters of the DBN model are adjusted according to the verification data set to obtain the anomaly detection model.

6. The method for generating a hepatobiliary CT diagnosis report according to claim 4, characterized in that: The method further comprises: When it is determined based on the abnormality detection model that the first hepatobiliary CT scan has no abnormality, the diagnosis report is obtained based on the first hepatobiliary CT scan.

7. The method for generating a hepatobiliary CT diagnosis report according to claim 4, characterized in that: The hepatobiliary enhanced CT includes hepatobiliary CT in the arterial phase, portal venous phase, and equilibrium phase, and determining the first abnormal region and / or the second abnormal region according to the key regions of the normal hepatobiliary CT and the key regions corresponding to the first hepatobiliary CT includes: According to the key areas of the normal hepatobiliary CT and the key areas corresponding to the first hepatobiliary CT, for each key area, a first eigenvector is calculated on the hepatobiliary plain scan CT in the first hepatobiliary CT, and a second eigenvector is calculated on the corresponding normal hepatobiliary plain scan CT; Calculating a first distance between each of the first eigenvectors and the corresponding second eigenvector; In a case where the first distance is greater than a first distance threshold, determining a key area in the hepatobiliary plain scan CT corresponding to the first distance being greater than the first distance threshold as the first abnormal area; Obtaining a first change curve of corresponding CT values ​​according to the multiple key areas of the hepatobiliary enhanced CT; Obtaining a normal change curve of the corresponding CT value according to the multiple key areas of the normal hepatobiliary enhanced CT; In the case that the difference between the first change curve and the corresponding normal change curve is greater than a difference threshold, the corresponding key area in the hepatobiliary enhanced CT is determined as the second abnormal area.

8. The method for generating a hepatobiliary CT diagnosis report according to claim 3, characterized in that: The determining of the patient's hepatobiliary text features according to the abnormal region information and the hepatobiliary feature library includes: Calculating the similarity between the abnormal region information and each of the image feature vectors in the hepatobiliary feature library to obtain a feature similarity matrix; According to the feature similarity matrix, determining the text feature vector corresponding to the image feature vector having a similarity greater than a similarity threshold as a candidate text feature; The hepatobiliary text feature is determined according to the abnormal region information and each of the candidate text features.

9. The method for generating a hepatobiliary CT diagnosis report according to claim 1, characterized in that: The obtaining of a diagnosis report according to the hepatobiliary text features and the second hepatobiliary CT comprises: The second hepatobiliary CT is annotated according to the hepatobiliary text features, and the diagnosis report is obtained.

10. A device for generating a hepatobiliary CT diagnostic report, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, the method according to any one of claims 1 to 9 is implemented.

Images