System and method for predicting truncated images, method for preparing data and medium therefor

Abstract

The present invention relates to a system and method for predicting a truncated image, a method for preparing data, and a medium thereof. The method for preparing data includes a virtual simulation step for performing a virtual simulation on image data to simultaneously obtain a virtual distorted image with data truncation and a virtual precision standard image without data truncation. The method for predicting a truncated image predicts a truncated image based on a trained learning network, and the trained learning network is obtained by data training based on a data set consisting of a virtual distorted image and a virtual precision standard image obtained by the method for preparing data. A system corresponding to the method and a recording medium recording instructions that can implement the method are also provided.

Description

Technical Field

[0001] The present invention relates to the field of medical imaging, and in particular to a technology for preparing data for predicting a truncated image in computer tomography (CT) imaging and predicting the truncated image based on the data. Background Art

[0002] During the CT scan, a detector is used to collect data of X-rays after they pass through the patient's body, and then these collected X-ray data are processed to obtain projection data. These projection data can be used to reconstruct slice images. Complete projection data can reconstruct accurate slice images for diagnosis.

[0003] However, if the patient is large or has a special posture, some parts of the patient's body will exceed the scanning domain, and the detector will not be able to collect complete projection data. This is called data truncation. This data truncation will cause truncation artifacts and ultimately lead to distorted reconstructed images. Distorted images are definitely not ideal in radiotherapy, because doctors must know the lines of the skin and the CT number along the beam when diagnosing, so that they can accurately determine the radiation dose that should be applied to the patient, but distorted images cannot accurately reflect the above information. In this way, how to restore the image outside the scanning domain (which we call a truncated image) is a problem that must be solved.

[0004] There are several traditional methods for dealing with the above truncation problem. They predict the truncated projection data through some mathematical models, such as using water phantom to predict the truncated part. However, the quality of the truncated images restored by these traditional methods varies with different actual conditions, and the performance is not ideal. In addition, different users often involve different sets of patients, and the image data they need also has a focus, but traditional methods have never involved classifying image data.

[0005] In recent years, a new technology has emerged, in which truncated images are predicted by artificial intelligence (AI). Using AI to predict truncated images undoubtedly has huge advantages that traditional technologies cannot match. However, the performance of AI depends on its input data. Hospitals or research institutions and departments continue to accumulate raw image data every day, but we obviously cannot simply input all of these accumulated raw image data into the AI ​​network for learning, because the performance of AI depends on the quality of the input data, not the quantity. On the other hand, to predict truncated images through AI, there must be an input data set corresponding to the distorted image and a refined standard data set corresponding to the refined standard image without data truncation to be input into the AI ​​network at the same time, but such images are not easy to obtain, or the data type is relatively single. Summary of the invention

[0006] An object of the present invention is to overcome the above and / or other problems in the prior art, and to obtain sufficient and reliable data for algorithm verification, thereby greatly helping to improve the accuracy of prediction of truncated images.

[0007] According to a first aspect of the present invention, a method for preparing data for predicting a truncated image is provided, which includes a virtual simulation step for virtually simulating image data to simultaneously obtain a virtual distorted image with data truncation and a virtual precise standard image without data truncation.

[0008] Preferably, before the virtual simulation step, the method further comprises an adaptive classification step for adaptively classifying the acquired image data according to predefined features, and the virtual simulation step is for virtually simulating the classified image data.

[0009] The predefined feature may include an image data type. Further, the predefined feature may also include a possibility corresponding to the image data type.

[0010] The image data type may be an anatomical part of a patient.

[0011] Preferably, the virtual simulation step further includes: receiving an original image without data truncation; virtually translating (offsetting) a portion of the original image corresponding to the target object to partially move it out of the scanning domain, thereby obtaining a virtual precision standard image; performing simulated scanning on the virtual precision standard image and performing virtual data acquisition to generate virtual truncated data; and performing image reconstruction processing on the virtual truncated data to obtain a virtual distorted image.

[0012] Preferably, the virtual simulation step further includes: receiving an original image without data truncation, the original image being used as a virtual precision standard image; keeping the original image within the scanning domain, and performing forward projection processing on the original image to obtain a sinusoidal graph of the original image; cutting off the left channel and the right channel of the sinusoidal graph, and filling them with filling values ​​to generate virtual truncated data; and performing image reconstruction processing on the virtual truncated data to obtain a virtual distorted image.

[0013] More preferably, the above filling value is an edge filling value.

[0014] According to a second aspect of the present invention, there is provided a method for predicting a truncated image, comprising the following steps: predicting a truncated image based on a trained learning network, wherein the trained learning network is obtained by data training based on a data set consisting of a virtual distorted image and a virtual precise standard image obtained by adopting the above-mentioned method for preparing data.

[0015] According to a third aspect of the present invention, there is provided a system for predicting a truncated image, comprising: a virtual simulation device for virtually simulating image data to simultaneously obtain a virtual distorted image with data truncation and a virtual precise standard image without data truncation; and a prediction device for predicting the truncated image based on a trained learning network, wherein the trained learning network is obtained by data training based on a data set consisting of the virtual distorted image and the virtual precise standard image.

[0016] Preferably, the system further comprises an adaptive classifier for adaptively classifying the collected image data according to predefined features, and the virtual simulation device is used for virtually simulating the classified image data.

[0017] Preferably, the virtual simulation device is further configured to: receive an original image without data truncation; virtually translate (offset) a portion of the original image corresponding to the target object to partially move it out of the scanning domain, thereby obtaining a virtual precision standard image; simulate scanning the virtual precision standard image and perform virtual data acquisition to generate virtual truncated data; and perform image reconstruction processing on the virtual truncated data to obtain a virtual distorted image.

[0018] Preferably, the virtual simulation device is further configured to: receive an original image without data truncation, the original image being used as a virtual precision standard image; keep the original image within the scanning domain, and perform forward projection processing on the original image to obtain a sinusoidal graph of the original image; cut off the left channel and the right channel of the sinusoidal graph, and fill them with filling values ​​to generate virtual truncated data; and perform image reconstruction processing on the virtual truncated data to obtain a virtual distorted image.

[0019] According to a fourth aspect of the present invention, there is also provided a computer-readable storage medium, on which instructions recorded can implement the above-mentioned method and system.

[0020] According to the method for preparing data for predicting truncated images of the present invention, it is possible to intelligently simulate real data truncation to obtain the distorted image and precise standard image required for the prediction of the truncated image; not only that, it is also possible to additionally intelligently classify the image data into different categories, thereby being more conducive to the above-mentioned distorted image and precise standard image.

[0021] The method and system for predicting a truncated image according to the present invention predicts the truncated image based on a trained learning network, and the trained learning network is obtained by data training based on a data set consisting of the above-mentioned virtual distorted images and virtual precise standard images, thereby being able to obtain prediction results quickly and more accurately.

[0022] Other features and aspects will become apparent from the following detailed description and accompanying drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] The present invention may be better understood by describing exemplary embodiments of the present invention in conjunction with the accompanying drawings, in which:

[0024] Figure 1 is a flow chart of a method for preparing data for predicting a truncated image according to an exemplary embodiment of the present invention;

[0025] Figure 2 yes Figure 1 A flowchart of a first embodiment of the virtual simulation step in the method shown;

[0026] Figure 3 yes Figure 1 A schematic diagram of a first embodiment of a virtual simulation step in the illustrated method;

[0027] Figure 4 yes Figure 1 A flow chart of a second embodiment of the virtual simulation step in the method shown;

[0028] Figure 5 yes Figure 1 A schematic diagram of a second embodiment of the virtual simulation step in the illustrated method;

[0029] Figure 6 is a flow chart of an alternative embodiment of a method for preparing data for predicting a truncated image according to an exemplary embodiment of the present invention;

[0030] Figure 7 Shows Figure 6 An example of data classification in the adaptive classification step of the method shown;

[0031] Figure 8 is a flowchart of a method for predicting a truncated image according to an exemplary embodiment of the present invention;

[0032] Fig. 9 is a schematic block diagram of a system for predicting a truncated image according to an exemplary embodiment of the present invention; and

[0033] Fig.10 An example of a system for predicting a truncated image according to an exemplary embodiment of the present invention is shown. DETAILED DESCRIPTION

[0034] The specific embodiments of the present invention will be described below. It should be noted that in the specific description of these embodiments, in order to provide a concise description, it is impossible for this specification to provide a detailed description of all the features of the actual embodiments. It should be understood that in the actual implementation of any embodiment, just as in the process of any engineering project or design project, in order to achieve the specific goals of the developer and to meet system-related or business-related restrictions, various specific decisions are often made, and this will also change from one embodiment to another. In addition, it can also be understood that although the efforts made in this development process may be complex and lengthy, for ordinary technicians in the field related to the content disclosed by the present invention, some changes such as design, manufacturing or production based on the technical content disclosed in this disclosure are just conventional technical means, and should not be understood as insufficient content of this disclosure.

[0035] Unless otherwise defined, the technical or scientific terms used in the claims and the specification shall have the usual meaning understood by persons with ordinary skills in the technical field to which the invention belongs. The words "first", "second" and similar words used in the patent application specification and the claims of the present invention do not indicate any order, quantity or importance, but are only used to distinguish different components. "One" or "one" and other similar words do not indicate a quantitative limitation, but indicate the existence of at least one. "Include" or "comprises" and other similar words mean that the elements or objects appearing before "include" or "comprises" include the elements or objects listed after "include" or "comprises" and their equivalent elements, and do not exclude other elements or objects. "Connected" or "connected" and other similar words are not limited to physical or mechanical connections, nor are they limited to direct or indirect connections.

[0036] According to an embodiment of the present invention, a method for preparing data for predicting a truncated image is provided.

[0037] refer to Figure 1 , Figure 1 is a flow chart of a method 10 for preparing data for predicting a truncated image according to an exemplary embodiment of the present invention. The method 10 may include step 200.

[0038] like Figure 1 As shown, in step 200 (virtual simulation step), the image data is virtually simulated to simultaneously obtain a virtual distorted image with data truncation and a virtual precise standard image without data truncation.

[0039] Method 10 requires preparing data for predicting truncated images, specifically, preparing data for truncated image prediction based on AI. In this way, two data sets need to be prepared for AI network learning: one is the input data set (i.e., the distorted image caused by truncation); the other is the precise standard data set (i.e., without any truncation).

[0040] However, in actual clinical cases, because data truncation has occurred (for example, the target object exceeds the scanning domain), that is, the truncated part of the data has been lost, and therefore, accurate standard data cannot be obtained. However, the method 10 for preparing data for predicting a truncated image according to an exemplary embodiment of the present invention can simulate the above data truncation in an intelligent manner, thereby obtaining a distorted image and an accurate standard image at the same time.

[0041] First embodiment of virtual simulation

[0042] refer to Figure 2 In the first embodiment according to the present invention, the above step 200 may further include the following sub-steps 210 to 216.

[0043] Specifically, in sub-step 210, an original image without data truncation is received. The original image is usually a slice image, but it is not excluded that a projection image may be provided sometimes. However, even if a projection image is provided, it can be converted into a slice image by back-projection processing.

[0044] refer to Figure 3 In sub-step 212, the portion (grayish white portion) corresponding to the target object in the original image is virtually translated (offset) to partially move out of the scanning domain, thereby obtaining a virtual precision standard image. It should be noted that in this process, the original slice image itself remains stationary. The scanning domain itself is a circle with the focus as the center, which can be expressed as, for example, DFOV50, where "50" is the diameter of the scanning domain, and its unit is cm. When the scanning domain is DFOV50, partially moving out of the scanning domain means partially passing over the edge of DFOV50.

[0045] Next, if Figure 2 As shown, in sub-step 214, the virtual precision standard image is simulated scanned and virtual data acquisition is performed to generate virtual truncated data. As described above, the virtual precision standard image obtained in sub-step 212 is a portion of the image corresponding to the target object in the original image that has moved out of the scanning area (passed the edge of DFOV50), and this portion of the image will not be scanned, that is, virtual data truncation occurs. Through sub-step 214, virtual truncated data can be obtained, which is projection data.

[0046] Finally, in sub-step 216, the virtual truncation data is subjected to image reconstruction processing to obtain a virtual distorted image. The image reconstruction processing may specifically be a back-projection processing. The obtained virtual distorted image is a slice image.

[0047] from Figure 3 It can be seen that compared with the original slice image, the final virtual precision standard image only has the part of the image corresponding to the target object (the grayish-white part) translated, but the data of the grayish-white part in the virtual precision standard image is complete without any truncation. Compared with the original slice image, the final virtual distorted image not only has the part of the image corresponding to the target object (the grayish-white part) translated, but the data of the grayish-white part in the virtual distorted image is obviously incomplete, and data truncation has occurred. Therefore, through only a virtual simulation process, the input data and precision standard data required for AI network learning can be obtained from slice images without data truncation.

[0048] Second embodiment of virtual simulation

[0049] The virtual simulation process according to the present invention can also be implemented in another way.

[0050] Specific as Figure 4 As shown, in the second embodiment according to the present invention, the above step 200 may further include the following sub-steps 220 to 226.

[0051] In sub-step 220, an original image without data truncation is received, and the original image is used as a virtual precision standard image. As mentioned above, the original image is usually a slice image, but it is not excluded that a projection image may be provided sometimes, and in this case, it can be converted into a slice image by back-projection processing. Figure 5 It can be seen that the virtual precision standard image not only has complete data, but also is within the range of the scanning domain (eg, DFOV50).

[0052] Next, in sub-step 222, the original image (i.e., the virtual precision standard image) is kept in the scanning domain, and the original image is subjected to forward projection processing to obtain a sinogram of the original image. As mentioned above, the original image does not contain any truncation, and the data therein is complete. Therefore, through forward projection processing, complete projection data can be obtained from the original slice image, that is, complete sinogram data can be obtained. For details, see Figure 5 .

[0053] Then, in sub-step 224, the left channel and the right channel of the sinogram are cut off and filled with fill values ​​to generate virtual truncated data. Figure 5As shown, the original left and right channels in the sinusoidal diagram are replaced by fill values, thereby simulating truncation.

[0054] The above-mentioned "padding value" is padding, whose attributes define the space between the element border and the element content. The padding abbreviation attribute sets all inner margin attributes in one declaration, setting all inner margin attributes of the current or specified element. This attribute can have 1 to 4 values. Using the padding attribute alone is to set all inner margin attributes of an element in one declaration. The abbreviation padding attribute can also be used. Once a value is changed, the corresponding distance of padding will change.

[0055] The padding value involved in the present invention is the padding attribute value, and the edge padding attribute value is preferably selected, that is, the edge padding value (for example, the value of the outermost channel of the untruncated part) is selected to fill the original left channel and right channel of the sine graph. However, in actual operation, other padding values ​​(for example, 0) can also be selected for filling according to needs.

[0056] Finally, in sub-step 226, the virtual truncation data is subjected to image reconstruction processing to obtain a virtual distorted image. Similar to the first embodiment of the present invention, the image reconstruction processing may be a back-projection processing, and the obtained virtual distorted image is a slice image. Figure 5 As shown, the final virtual distorted image is obviously incomplete in data compared with the original slice image, especially the part of the image corresponding to the target object (the grayish-white part), that is, data truncation has occurred.

[0057] Therefore, a method different from that of the first embodiment described above is adopted to virtually simulate data truncation, and in this process, the input data (distorted image) and precise standard data (precise standard image) required for AI network learning are obtained simultaneously.

[0058] Although two embodiments of virtual simulation are described above, this does not mean that the present invention can only use these two methods to simulate data truncation. The method for preparing data for prediction of truncated images according to the present invention can also use other methods to virtually simulate data truncation to simultaneously obtain the input data (distorted image) and precise standard data (precise standard image) required for AI network learning.

[0059] The virtual simulation process according to the present invention completely solves the problem of not being able to obtain accurate standard data in the traditional technology, and can also obtain sufficient input data very quickly and conveniently.

[0060] Optionally, the method 10 for preparing data for predicting a truncated image according to an exemplary embodiment of the present invention may further include step 100 (adaptive classification step) before step 200. For details, see Figure 6 .

[0061] Step 100 adaptively classifies the collected image data according to predefined features. The image data may be a projection image or a slice image. If it is a projection image, it can be converted into a slice image by conventional back-projection processing. Step 200 further performs virtual simulation on the image data classified in step 100.

[0062] Through the above step 100, the method 10 can intelligently classify the image data into different categories before performing virtual simulation on the image data.

[0063] refer to Figure 7 It can be seen that in the above step 100, the predefined feature may include an image data type. Specifically, the image data type may be an anatomical part of a patient. Further, the predefined feature may also include a possibility corresponding to the image data type.

[0064] For example, the above-mentioned anatomical parts may include shoulders, chests, abdomens, pelvises and special models (phantoms). The models described here can usually be body models, but sometimes they can also be water models.

[0065] In clinical practice, different users (such as hospitals, patients or research institutions, etc.) may have different patient data sets, so they may need to provide different input data sets for their AI network training. The adaptive classification step 100 according to the example of the present invention can configure the data input, such as the data type, the possibility of the data, etc., which will greatly improve the ability of the AI ​​network.

[0066] Back to Figure 6 In the method shown, when performing the virtual simulation step 200, there is a non-truncated (no data truncation), complete slice image as the original simulation input. This slice image comes from a data set containing all slice images of the human body from head to toe. This huge data set can be a data set collected and made public by various clinical medical institutions (such as hospitals), or it can be a data set collected by the user himself. The adaptive classification step 100 according to the example of the present invention is to adaptively classify the huge data set based on which part of the image the user needs to use.

[0067] For example, see Figure 7Among the patients in a certain hospital, there may be more patients with shoulder diseases, so there are more patients who are irradiated with CT due to shoulder diseases, so the possibility that the slice image obtained is the shoulder part is relatively large, for example, there is a 40% possibility; the next is more patients with pelvis, so there are more patients who are irradiated with CT due to pelvic diseases, and accordingly, the possibility that the slice image obtained is the pelvic part is slightly smaller than the possibility of the shoulder, for example, there is a 30% possibility; the next is more patients with chest and abdomen, so the number of patients who are irradiated with CT due to chest and abdomen is relatively reduced, and accordingly, the possibility that the slice image obtained is the chest and abdomen is even smaller, for example, both are 20%. At the same time, the hospital also has patients who have been installed with special models, so the possibility that the slice image obtained is a special model must also be considered, for example, 10%. According to the patient distribution of the above hospital, special data types (body parts) can be classified and their corresponding possibilities can be optimized. Specifically, the following two features can be pre-defined for data classification:

[0068] Feature 1: Special disease sites (such as Figure 2 shown, shoulders, chest, abdomen, pelvis, special models);

[0069] Feature 2: Special possibilities (such as Figure 2 As shown, 40%, 20%, 20%, 30%, and 10% correspond to the above-mentioned diseased parts respectively).

[0070] Based on the above two features, the huge image data set collected by the hospital can be adaptively classified. As a result, the input data set fed into the AI ​​learning network can be selected in a targeted manner.

[0071] For another example, for some specialized hospitals, such as chest hospitals, almost all patients have chest involvement, so the "special patient site" of feature 1 can be set to "chest", and the "special possibility" of feature 2 can be set to "100%". For another example, even in a general hospital, if the slice images obtained are all from a special department (such as neurosurgery), the "special patient site" of feature 1 can be set to "brain", and the "special possibility" of feature 2 can be set to "100%".

[0072] In short, features can be predefined according to various actual situations and specific needs. Although two features are defined in the above example, more features can be defined as needed, and the features are not limited to the type of data and the corresponding possibilities. In addition, even the data type is not limited to human body parts, but can also be other structures or objects that can be imaged.

[0073] Obviously, the above method for preparing data for predicting truncated images can greatly improve the quality of input data used in AI learning networks because it includes the adaptive classification step as described above.

[0074] Furthermore, with the above data prepared, the truncated image can be predicted by AI. Figure 8 , which shows a flow chart of a method 80 for predicting a truncated image according to an exemplary embodiment of the present invention.

[0075] like Figure 8 As shown, a method 80 for predicting a truncated image according to an exemplary embodiment of the present invention may include step 810 .

[0076] In step 810, a truncated image is predicted based on a trained learning network, wherein the trained learning network is obtained by data training based on a data set consisting of a virtual distorted image and a virtual precise standard image obtained by adopting the above-mentioned method for preparing data.

[0077] The above method 80 obtains a data set for AI network learning through a data preparation method according to an exemplary embodiment of the present invention. In this process, not only can the special patient set and doctor habits for system calibration for different users be matched through adaptive data classification, but also the necessary AI training data (i.e., input data and precision standard data) can be obtained intelligently at the same time. The data set obtained in this way is undoubtedly very helpful for AI network learning and can greatly improve the accuracy of predicting truncated images through AI.

[0078] In addition, the present invention also provides a system for predicting a truncated image.

[0079] Fig. 9 A system 900 for predicting a truncated image according to an exemplary embodiment of the present invention is shown. The system 900 may include a virtual simulation device 910 and a prediction device 920. The virtual simulation device 910 is used to perform virtual simulation on image data to simultaneously obtain a virtual distorted image with data truncation and a virtual precision standard image without data truncation. The prediction device 920 is used to predict the truncated image based on a trained learning network, wherein the trained learning network is obtained by data training based on a data set consisting of the virtual distorted image and the virtual precision standard image.

[0080] Further, Fig.10 An example of a system for predicting a truncated image according to an exemplary embodiment of the present invention is shown.

[0081] As Fig.10As shown, the exemplary CT system 1000 is configured to predict a truncated image and / or prepare data for predicting a truncated image. Specifically, the CT system 1000 is configured to image a target object; if the obtained image has truncation, the image with truncation is predicted to restore the truncated image data. The target object may be a patient or any other object to be imaged, and in this example, a patient is taken as an example.

[0082] In one embodiment, the CT system 1000 includes a gantry 1002 on which an X-ray source 1004 and a detector array 1008 are disposed opposite to each other. The detector array 1008 is composed of a plurality of detector elements 2002. The X-ray source 1004 is used to project X-rays 1006 that penetrate a patient toward the detector array 1008. The detector array 1008 collects attenuated X-ray beam data, which is pre-processed as projection data of a target volume of the patient.

[0083] In one embodiment, the CT system 1000 includes a control mechanism 2008. The control mechanism 2008 may include an X-ray controller 2010 for providing power and timing signals to the X-ray source 1004. The control mechanism 2008 may also include a gantry motor controller 2012 for controlling the rotation speed and / or position of the gantry 1002 based on imaging requirements. In addition, the control mechanism 2008 may also include a patient bed controller 2026 for moving a patient bed 2028 to position a patient (not shown) in an appropriate position within the gantry 1002.

[0084] In one embodiment, the CT system 1000 further includes a data acquisition system (DAS) 2014 for sampling and digitizing analog data received from the detector elements 2002 .

[0085] In one embodiment, the CT system 100 further includes a computing device 2016, and the data sampled and digitized by the DAS 2014 will be transmitted to the computer or computing device 2016 for processing. The computing device 2016 can communicate with the operator console 2020 to facilitate the operator's operation, and can also be connected to a display 2032 to facilitate the operator's observation. In addition, the computing device 2016 can also be connected to a picture archiving and communication system (PACS) 2024.

[0086] In one example, computing device 2016 stores data on a storage device, such as computer-readable storage medium 2018 ( Fig.10The computer-readable storage medium 2018 may include a hard disk drive, a floppy disk drive, a compact disk read / write (CD-R / W) drive, a digital versatile disk (DVD) drive, a flash drive, and / or a solid-state storage device, etc.

[0087] In addition, the computing device 2016 may also be used to provide commands and parameters to one or more of the DAS 2014 , the X-ray controller 2010 , the gantry motor controller 2012 , and the patient bed controller 2026 to control system operations, such as data acquisition and / or processing.

[0088] The CT system 1000 may further include an image reconstructor 2030, which reconstructs an image based on the sampled and digitized X-ray data using a suitable image reconstruction method. For example, the image reconstructor 2030 may use, for example, filtered back projection (FBP) to reconstruct an image of the target volume of the patient. Fig.10 2000 as a separate entity, but in some embodiments, the image reconstructor 2030 may be formed as part of the computing device 2016. Alternatively, the image reconstructor 2030 may not be present in the CT system 1000; or, the computing device 2016 may perform one or more functions of the image reconstructor 2030. In addition, the image reconstructor 2030 may be located locally or remotely, and may be operatively connected to the CT system 1000 using a wired or wireless network.

[0089] In one embodiment, the image reconstructor 2030 stores the reconstructed image in a storage device or computer-readable storage medium 2018. Alternatively, the image reconstructor 2030 transmits the reconstructed image to a computing device 2016 to generate patient information for diagnosis.

[0090] In one embodiment, the CT system 1000 further includes a truncated image prediction device, which can receive the truncated image from the computing device 2016 or the image reconstructor 2030 and predict and / or restore the truncated image. The truncated image prediction device can be formed as a part of the computing device 2016 or the image reconstructor 2030 ( Fig.10 , which is part of computing device 2016).

[0091] In one embodiment, the CT system 1000 may further include a data preparation device for preparing data for truncated image prediction, which may also be formed as a part of the computing device 2016 or the image reconstructor 2030 ( Fig.10 , which is part of computing device 2016).

[0092] It should be noted that various methods and processes further described in this specification may be stored in the form of executable instructions in a computer-readable storage medium of the computing device 2016 and / or the image reconstructor 2030 in the CT system 1000. For example, a truncated image prediction device and a data preparation device may include executable instructions of a computer-readable storage medium, and may use the method described in this specification to predict a truncated image / prepare data.

[0093] Some exemplary embodiments have been described above. However, it should be understood that various modifications may be made to the exemplary embodiments described above without departing from the spirit and scope of the present invention. For example, if the described techniques are performed in a different order and / or if the components in the described systems, architectures, devices, or circuits are combined in different ways and / or replaced or supplemented by other components or their equivalents, suitable results may be achieved, and accordingly, these modified other implementations also fall within the scope of protection of the claims.

Claims

1. A method for preparing data for predicting a truncated image, comprising the steps of: The virtual simulation step is used to perform virtual simulation on the image data to simultaneously obtain a virtual distorted image with data truncation and a virtual accurate standard image without data truncation. in, The virtual simulation step further comprises: Receive the original image without data truncation; While the original image itself remains stationary, a portion of the original image corresponding to the target object is virtually translated to partially move out of the scanning area, thereby obtaining a virtual precision standard image; Performing simulated scanning on the virtual precision standard image and performing virtual data acquisition to generate virtual truncation data; and The virtual truncation data is subjected to image reconstruction processing to obtain a virtual distorted image.

2. The method according to claim 1, characterized in that Before the virtual simulation step, the method further includes an adaptive classification step for adaptively classifying the collected image data according to predefined features, and the virtual simulation step is for virtually simulating the classified image data.

3. The method according to claim 2, characterized in that The predefined characteristics include image data type.

4. The method according to claim 3, characterized in that The predefined features also include the corresponding possibilities of the image data types.

5. The method according to claim 3, characterized in that The image data type is an anatomical part of a patient.

6. A method for predicting a truncated image, comprising the steps of: The truncated image is predicted based on a trained learning network, wherein the trained learning network is obtained by data training based on a data set consisting of a virtual distorted image and a virtual precise standard image obtained by adopting the method according to any one of claims 1 to 5.

7. A system for predicting a truncated image, comprising: A virtual simulation device, used for performing virtual simulation on image data to simultaneously obtain a virtual distorted image with data truncation and a virtual precise standard image without data truncation; as well as, A prediction device is used to predict the truncated image based on a trained learning network, wherein the trained learning network is obtained by data training based on a data set consisting of the virtual distorted image and the virtual precise standard image. Wherein, the virtual simulation device is further configured as follows: Receive the original image without data truncation; While the original image itself remains stationary, a portion of the original image corresponding to the target object is virtually translated to partially move out of the scanning area, thereby obtaining a virtual precision standard image; Performing simulated scanning on the virtual precision standard image and performing virtual data acquisition to generate virtual truncation data; and The virtual truncation data is subjected to image reconstruction processing to obtain a virtual distorted image.

8. The system according to claim 7, characterized in that It also includes an adaptive classifier for adaptively classifying the collected image data according to predefined features, and the virtual simulation device is used for virtually simulating the classified image data.

9. A computer-readable storage medium having encoded instructions recorded thereon, which, when executed, performs the method according to any one of claims 1 to 6.

Images