Imaging method and apparatus

Abstract

This application provides an imaging method and system, as well as a non-transitory computer-readable storage medium. The imaging method includes preprocessing projection data to obtain a predicted image of a truncated portion; forward projecting the predicted image to obtain predicted projection data of the truncated portion; and reconstructing an image by combining the projection data obtained from the forward projection with the projection data of the untruncated portion.

Description

Technical Field

[0001] This invention relates to image processing, and more particularly to an imaging method and apparatus, as well as a non-transitory computer-readable storage medium. Background Technology

[0002] During computed tomography (CT) scanning, detectors acquire X-ray data as it passes through the object being examined. This acquired X-ray data is then processed to obtain projection data. This projection data can be used to reconstruct CT images. Complete projection data allows for the reconstruction of accurate CT images for diagnostic purposes.

[0003] However, if the object being detected is large or positioned in a special way, some parts of the object will extend beyond the scanning area, and the detector will be unable to acquire complete projection data. This is called data truncation. Typically, mathematical models, such as water models, can be used to predict the projection data or image of the truncated portion. However, the image quality of the truncated portion predicted by these traditional methods varies depending on different actual situations, and their performance is not ideal.

[0004] Furthermore, since image reconstruction utilizes methods such as filtered back projection (FBP) to reconstruct the image, the untruncated portion of the image or projection data is also affected by data truncation. During the filtering process, data error channels are contaminated, causing artifacts such as CT value drift within the scan domain, resulting in distorted and inaccurate reconstructed images. Even if a size smaller than or equal to the scan domain is used for reconstruction, the reconstructed CT image cannot have the correct CT values. Summary of the Invention

[0005] This invention provides an imaging method and apparatus, as well as a non-transitory computer-readable storage medium.

[0006] An exemplary embodiment of the present invention provides an imaging method, the method comprising: preprocessing projection data to obtain a predicted image of a truncated portion; forward projecting the predicted image to obtain predicted projection data of the truncated portion; and reconstructing an image based on the predicted projection data and the projection data of the untruncated portion.

[0007] Optionally, preprocessing the projection data to obtain a predicted image of the truncated portion includes processing the truncated portion of the projection data based on the untruncated portion of the projection data to reconstruct an initial image of the truncated portion; and calibrating the initial image based on a trained learning network to obtain a predicted image of the truncated portion.

[0008] Furthermore, filling the truncated portion of the projection data includes filling the truncated portion with the projection data information of the boundary of the untruncated portion.

[0009] Furthermore, calibrating the initial image based on the trained learning network to obtain a predicted image of the truncated portion includes converting the pixels of the initial image of the truncated portion from polar coordinates to rectangular coordinates to obtain a pixel matrix of the initial image; calibrating the pixel matrix based on the trained learning network; and converting the calibrated pixel matrix from rectangular coordinates to polar coordinates to obtain a predicted image of the truncated portion.

[0010] Optionally, the trained learning network is trained based on a virtual distorted image and a control image. Further, the trained learning network is trained based on a pixel matrix obtained by coordinate transformation of the virtual distorted image and a pixel matrix obtained by coordinate transformation of the control image. Further, the method for acquiring the virtual distorted image and the control image includes receiving an original image without data truncation; virtually translating a portion of the original image corresponding to a target object to partially move it out of the scanning domain, thereby obtaining a control image; virtually scanning and acquiring virtual data from the control image to generate virtual truncated projection data; and reconstructing the virtual truncated projection data to obtain a virtual distorted image. Optionally, the method for acquiring the virtual distorted image and the control image includes: receiving an original image without data truncation, the original image being used as a control image; keeping the original image within the scanning domain and performing orthographic projection on the original image to obtain a projection of the original image; filling the channels on both sides of the projection to generate virtual truncated projection data; and reconstructing the virtual truncated projection data to obtain a virtual distorted image.

[0011] An exemplary embodiment of the present invention also provides a non-transitory computer-readable storage medium for storing a computer program that, when executed by a computer, causes the computer to perform the imaging method described above.

[0012] An exemplary embodiment of the present invention also provides an imaging system, the system including a prediction device, an image processing device, and an image reconstruction device. The prediction device is used to preprocess projection data to obtain a predicted image of a truncated portion. The image processing device is configured to forward project the predicted image to obtain predicted projection data of the truncated portion. The image reconstruction device is used to perform image reconstruction based on the predicted projection data and the projection data of the untruncated portion.

[0013] Optionally, the prediction device includes a preprocessing device and a control device. The preprocessing device processes the truncated portion of the projection data based on the untruncated portion to obtain an initial image of the truncated portion. The control device calibrates the initial image based on a trained learning network to obtain a predicted image of the truncated portion. Further, the control device includes a transformation module, a calibration module, and an inverse transformation module. The transformation module converts the pixels of the initial image of the truncated portion from polar coordinates to Cartesian coordinates to obtain a pixel matrix of the initial image. The calibration module calibrates the pixel matrix based on the trained learning network. The inverse transformation module converts the calibrated pixel matrix from Cartesian coordinates to polar coordinates to obtain the predicted image of the truncated portion.

[0014] Other features and aspects will become clear from the following detailed description, accompanying drawings, and claims. Attached Figure Description

[0015] The invention can be better understood by describing exemplary embodiments of the invention in conjunction with the accompanying drawings, in which:

[0016] Figure 1 A flowchart of an imaging method according to some embodiments of the present invention is shown;

[0017] Figure 2 It shows according to Figure 1 The flowchart shown illustrates the step of preprocessing the projection data to obtain the truncated predicted image in the method described.

[0018] Figure 3 It shows according to Figure 2 The flowchart shows the steps of calibrating an initial image using a trained learning network to obtain a predicted image of the truncated portion in the method shown.

[0019] Figure 4 It shows according to Figure 2 The method shown includes some intermediate images of the projection data preprocessed. Image (1) shows the projection with data truncation, image (2) shows the simulated projection obtained by filling the truncated portion of the projection data in (1), and image (3) shows the CT image obtained by image reconstruction of the simulated projection in (2).

[0020] Figure 5 It shows according to Figure 3Some intermediate images in the steps shown, wherein image (1) shows the rectangular coordinate pixel matrix corresponding to the truncated part, image (2) shows the predicted pixel matrix obtained by the learning network after transforming the rectangular coordinate image in (1), and image (3) shows the predicted image obtained after performing an inverse coordinate transformation on the predicted pixel matrix in (2).

[0021] Figure 6 A flowchart is shown for a method of preparing training data for a learning network according to some embodiments of the present invention;

[0022] Figure 7 A flowchart is shown for a method of preparing training data for a learning network according to other embodiments of the present invention;

[0023] Figure 8 A schematic diagram of a CT system according to some embodiments of the present invention is shown;

[0024] Figure 9 A schematic diagram of an imaging system according to some embodiments of the present invention is shown; and

[0025] Figure 10 It shows according to Figure 9 A schematic diagram of the prediction device in the system shown. Detailed Implementation

[0026] The following describes specific embodiments of the present invention. It should be noted that, in order to provide a concise description, this specification cannot exhaustively describe all features of the actual embodiments. It should be understood that, in the actual implementation of any embodiment, just as in any engineering or design project, various specific decisions are often made to achieve the developer's specific goals and to meet system-related or business-related constraints, and this can change from one embodiment to another. Furthermore, it is understood that although the efforts made in this development process may be complex and lengthy, for those skilled in the art related to the content disclosed in this invention, some design, manufacturing, or production modifications based on the technical content disclosed herein are merely conventional technical means and should not be construed as insufficient content of this disclosure.

[0027] Unless otherwise defined, the technical or scientific terms used in the claims and description shall have the ordinary meaning understood by one of ordinary skill in the art to which this invention pertains. The terms "first," "second," and similar terms used in the patent application description and claims of this invention do not indicate any order, quantity, or importance, but are merely used to distinguish different components. The terms "an" or "a" and similar terms do not indicate a quantity limitation, but rather indicate the presence of at least one. The terms "comprising" or "including" and similar terms mean that the element or object preceding "comprising" or "including" encompasses the element or object listed following "comprising" or "including" and its equivalents, and do not exclude other elements or objects. The terms "connected" or "linked" and similar terms are not limited to physical or mechanical connections, nor are they limited to direct or indirect connections.

[0028] As used in this invention, the term "object being detected" can include any object that is imaged. The terms "projection data" and "projection image" have the same meaning.

[0029] In some embodiments, during CT scans, when the subject is large or in a specific posture, parts of the image may extend beyond the CT scan field of view (FOV), resulting in truncated projection data and distorted reconstructed images. The imaging methods and systems in some embodiments of this invention, based on artificial intelligence, can more accurately predict the image of the truncated portion and, based on the predicted image of the truncated portion, improve the quality of the image of the untruncated portion, thus providing better evidence for diagnosis and / or treatment. It should be noted that, from the perspective of those skilled in the art or related fields, this description should not be construed as limiting the invention solely to CT systems. In fact, the methods and systems described herein for acquiring predicted images of truncated portions can be reasonably applied to other imaging fields, both medical and non-medical, such as X-ray systems, PET systems, SPECT systems, MR systems, or any combination thereof.

[0030] As discussed in this paper, artificial intelligence (which includes deep learning techniques, also known as deep machine learning, hierarchical learning, or deep structured learning) employs artificial neural networks for learning. Deep learning methods are characterized by using one or more network architectures to extract or model a class of data of interest. Deep learning methods can be accomplished using one or more processing layers (e.g., convolutional layers, input layers, output layers, normalization layers, etc., with different functional layers depending on the deep learning network model), where the configuration and number of layers allow deep learning networks to handle complex information extraction and modeling tasks. Specific parameters of the network (also known as “weights” and “biases”) are typically estimated through a so-called learning process (or training process), although in some embodiments, the learning process itself can be extended to learning elements of the network architecture. The parameters learned or trained typically result in (or output) a network corresponding to different levels of layers; therefore, extracting or modeling different aspects of the initial data or the output of the previous layer can often represent the hierarchical structure or cascade of layers. In image processing or reconstruction, this can be characterized as different layers relative to different feature levels or resolutions in the data. Therefore, processing can be layered; earlier or higher-level layers might correspond to extracting “simple” features from the input data, followed by layers that combine these simple features into features exhibiting higher complexity. In practice, each layer (or more specifically, each “neuron” within a layer) can employ one or more linear and / or nonlinear transformations (so-called activation functions) to process the input data into an output data representation. The number of “neurons” can be constant across multiple layers, or it can vary from layer to layer.

[0031] As discussed herein, as part of the initial training of a deep learning process for solving a specific problem, a training dataset with known input values ​​(e.g., an input image or a pixel matrix of an image after coordinate transformation) and known or desired values ​​can be used to obtain the final output of the deep learning process (e.g., a target image or a pixel matrix of an image after coordinate transformation) or the individual layers of the deep learning process (assuming a multi-layer network architecture). In this way, the deep learning algorithm can process the known or training dataset (in a supervised or guided manner or in an unsupervised or unguided manner) until the mathematical relationship between the initial data and the desired output is identified and / or the mathematical relationship between the input and output of each layer is identified and characterized. The learning process typically utilizes (partially) the input data and creates a network output for that input data. The created output is then compared with the desired (target) output of the dataset, and the network parameters (weights and biases) are iteratively updated using the difference between the generated and desired outputs. One such update / learning mechanism uses the stochastic gradient descent (SGD) method to update the network parameters; of course, those skilled in the art will understand that other methods known in the art can also be used. Similarly, a separate validation dataset can be used, where the input and desired target values ​​are known, but only the initial values ​​are provided to the trained deep learning algorithm, and then the output is compared with the output of the deep learning algorithm to validate the previous training and / or prevent overtraining.

[0032] Figure 1 A flowchart of an imaging method 100 according to some embodiments of the present invention is shown. For example... Figure 1 As shown, the imaging method 100 includes steps 101, 102 and 103.

[0033] In step 101, the projection data is preprocessed to obtain a predicted image of the truncated portion. In some embodiments, the truncated portion can be predicted using conventional methods, such as padding the truncated portion of the projection data. In other embodiments, the truncated portion can be predicted based on deep learning, and specific methods will be discussed later. Figure 2-5 To elaborate.

[0034] In step 102, the predicted image is forward-projected to obtain the predicted projection data of the truncated portion. In some embodiments, the predicted projection data of the truncated portion can be obtained by forward-projecting (or orthographically projecting) the predicted image.

[0035] In step 103, image reconstruction is performed based on the predicted projection data and the projection data of the uncropped portion. In some embodiments, since image reconstruction is performed using, for example, filtered back projection (FBP), during the image reconstruction process, the projection data of the uncropped portion obtained by forward projection is affected by filtered back projection, such as convolution calculation, to improve the image quality of the uncropped portion.

[0036] Figure 2 Shown according to Figure 1 The flowchart shows the step 101 of the method 100 shown, which involves preprocessing the projection data to obtain a predicted image of the truncated portion. Figure 3 Shown according to Figure 2 The flowchart of step 120 in method 101, which involves a trained learning network calibrating an initial image to obtain a predicted image of the truncated portion, is shown. Figure 4 Shown according to Figure 2 The method shown includes some intermediate images of the projection data preprocessed. Image (1) shows a projection with data truncation, image (2) shows a simulated projection obtained by filling the truncated portion of the projection in (1), and image (3) shows a CT image obtained by image reconstruction of the simulated projection in (2). Figure 5 Shown according to Figure 3 Some intermediate images of the method shown are shown, wherein image (1) shows the rectangular coordinate pixel matrix corresponding to the truncated part, image (2) shows the predicted pixel matrix obtained by learning the rectangular coordinate pixel matrix in (1) through the learning network, and image (3) shows the predicted image obtained by performing an inverse coordinate transformation on the predicted pixel matrix in (2).

[0037] like Figure 2 As shown, the method 101 for preprocessing projection data to obtain a predicted image of the truncated portion includes steps 110 and 120.

[0038] In step 110, the truncated portion of the projection data is processed based on the untruncated portion of the projection data to obtain an initial image of the truncated portion.

[0039] In some embodiments, the preprocessed initial image can be used as the predicted image of the truncated portion. In some embodiments, preprocessing the projection data includes padding the truncated portion of the projection data. Data obtained by acquiring X-ray data passing through the object being inspected can be used to obtain... Figure 4The projection data (i.e., projection) 210 shown in the middle image (1) is truncated because part of the object being detected exceeds the scanning range of the CT system. This truncated portion can be simulated through preprocessing of the projection data. Preferably, filling the truncated portion of the projection data includes filling the truncated portion with the projection data information of the untruncated boundary (outermost channel). Further, the projection data information of the untruncated boundary (outermost channel) of each channel is filled with the truncated portion of the corresponding channel, such as... Figure 4 As shown in image (2), the simulated projection data 220 is obtained by filling.

[0040] Although some embodiments of the present invention employ a filling method to simulate the truncated portion, those skilled in the art should understand that the truncated portion can be simulated in other ways, such as calculating the projection data information of the truncated portion based on the projection data information of the untruncated portion using a mathematical model. Furthermore, although some embodiments of the present invention employ filling the truncated portion with the projection data information of the untruncated portion boundary, those skilled in the art should understand that the embodiments of the present invention are not limited to this filling method, and the truncated portion can also be filled in other ways, such as filling all or part of the truncated portion with a specific tissue or a preset CT value (e.g., 0) or projection data information. Further, if the obtained projection data has missing data, the same filling method can also be used to fill the data. Moreover, although some embodiments of the present invention include filling the truncated portion in the preprocessing, those skilled in the art should understand that preprocessing does not only include this step, but can also include any data preprocessing operation performed before image reconstruction.

[0041] In some embodiments, after preprocessing the projection data, image reconstruction is performed on the filled projection data to obtain a CT image. For example... Figure 4 As shown in image (3), CT image 230 is a pair Figure 4 The image reconstruction is obtained by reconstructing the projection data of image (2) 220 (including the untruncated portion and the simulated truncated portion). The CT image 230 includes the untruncated portion image 231 and the truncated portion image 232. That is, the portion within the scanning domain 80 is the truncated portion image 231, and the portion outside the scanning domain 80 is the truncated portion image 232. The boundary of the scanning domain 80 is the intersection 234 of the untruncated portion 231 and the truncated portion 232. The image reconstruction algorithm may include, for example, back projection (FBP) reconstruction method, adaptive statistical iterative reconstruction (ASIR) method, common gradient (CG) method, maximum likelihood expectation maximization (MLEM) method, model-based iterative reconstruction (MBIR) method, etc.

[0042] In some embodiments, step 110 further includes extracting the CT image obtained from image reconstruction to obtain an initial image of the truncated portion, such as... Figure 4 The image (3) shows portion 232. In some embodiments, although the initial image of the truncated portion in this invention represents a portion of the image outside the scanning domain, those skilled in the art should know that image extraction is not limited to extracting only the portion of the image outside the scanning domain, but may also include at least a portion of the image that is not truncated. In practice, adjustments can be made as needed.

[0043] In step 120, the initial image of the truncated portion is calibrated based on the trained learning network to obtain a predicted image of the truncated portion.

[0044] By inputting a truncated virtual distorted image (known input value) and a corresponding image (expected output value), or the pixel matrix corresponding to the truncated virtual distorted image (known input value) and the corresponding pixel matrix corresponding to the corresponding image (expected output value), a learning network is constructed or trained based on deep learning methods to obtain the mathematical relationship between the known input value and the expected output value. Based on this, in practical operation, when a known truncated image (e.g., the initial truncated image 232 mentioned above) is input, the expected truncated image (i.e., expected output value - corresponding image) can be obtained based on the learning network. The training, construction, and data preparation of the learning network will be discussed in conjunction with... Figure 6 and Figure 7 Further description. In some embodiments, the training of the learning network in this invention is based on some network models known in the art (e.g., Unet, etc.).

[0045] Please refer to Figure 3 As shown, in some embodiments, the initial image of the truncated portion is calibrated based on the trained learning network to obtain a predicted image of the truncated portion (step 120), which includes steps 121, 122 and 123.

[0046] In step 121, the pixels of the truncated initial image are transformed from polar coordinates to Cartesian coordinates to obtain the pixel matrix of the initial image. For example, by transforming the initial image of the truncated portion (such as...) Figure 4 By performing a coordinate transformation from polar coordinates to rectangular coordinates on part 232 of the image (3), we can obtain the following: Figure 5 The rectangular coordinate pixel matrix 242 is shown in image (1).

[0047] In step 122, the pixel matrix is ​​calibrated based on the trained learning network, for example, by... Figure 5 After the pixel matrix 242 in image (1) is input into the learning network, the following can be obtained: Figure 5 The image (2) shows the calibrated control pixel matrix 252.

[0048] In step 123, the calibrated pixel matrix is ​​transformed from Cartesian coordinates to polar coordinates to obtain the predicted image of the truncated portion. For example, by transforming the calibrated pixel matrix (such as...) Figure 5 Image (2) 252 in the image (2) can be transformed from rectangular coordinates to polar coordinates to obtain the predicted image 262 of the truncated portion (as shown in the image). Figure 5 Image (3) is shown in the image.

[0049] Figure 6 A flowchart of a learning network training data preparation method 300 according to some embodiments of the present invention is shown. Figure 6 As shown, the training data preparation method 300 for the learning network includes steps 310, 320, 330 and 340.

[0050] In step 310, the original image without data truncation is acquired. In some embodiments, the initial image is reconstructed based on projection data, where all parts of the detected object are within the scanning domain, meaning there is no data truncation in the obtained projection data.

[0051] In step 320, a portion of the original image corresponding to the target object is virtually shifted (offset) to partially move it out of the scan domain, thereby obtaining a control image. In some embodiments, the control image represents an untruncated, complete CT image.

[0052] In step 330, the comparison image is virtually scanned and virtual data is acquired to generate virtual truncation projection data. In some embodiments, based on a virtual sampling system, the comparison image is virtually scanned. Since some target objects have moved out of the scanning field, this part of the image will not be scanned, which is equivalent to virtual truncation.

[0053] In step 340, image reconstruction is performed on the virtual truncated projection data to obtain a virtual distorted image. In some embodiments, the virtual distorted image represents a distorted image with data truncation.

[0054] Optionally, the data preparation method further includes step 350, in which coordinate transformation (e.g., from polar coordinates to rectangular coordinates) is performed on the virtual distorted image and the control image to obtain a virtual distorted pixel matrix and a virtual control pixel matrix, respectively. The aforementioned pixel matrices are used as known input values ​​and expected output values ​​of the learning network to construct or train the learning network to better obtain the predicted image of the truncated portion.

[0055] Figure 7A flowchart illustrating a method for preparing training data for a learning network according to other embodiments of the present invention is shown. Figure 7 As shown, the training data preparation method 400 for the learning network includes steps 410, 420, 430 and 440.

[0056] In step 410, the original image without data truncation is obtained and used as a control image.

[0057] In step 420, projection data corresponding to the original image is obtained. In some embodiments, the original image (i.e., the reference image) is kept within the scanning field, and forward projection is performed on the original image to obtain the projection (i.e., projection data) of the original image. In other embodiments, the original projection data obtained from the initial scan can be directly used as the projection (i.e., projection data).

[0058] In step 430, the channels on both sides of the projection data are filled to generate virtual truncated projection data. The channels on both sides include the top and bottom channels, and may also include the left and right channels, depending on the orientation of the projected image. In some embodiments, the filling in step 430 is the same as the filling method used in preprocessing the initial image (step 110 described earlier). Similarly, when other non-filling methods are used for preprocessing the initial image, the same method is used in step 430.

[0059] In step 440, image reconstruction is performed on the virtual truncated projection data to obtain a virtual distorted image.

[0060] Optionally, the data preparation method further includes step 450, in which coordinate transformation (e.g., from polar coordinates to rectangular coordinates) is performed on the virtual distorted image and the control image to obtain a virtual distorted pixel matrix and a virtual control pixel matrix, respectively. The aforementioned pixel matrices are used as known input values ​​and expected output values ​​of the learning network to construct or train the learning network to better obtain the predicted image of the truncated portion.

[0061] Although the above describes two embodiments of training data preparation for learning networks, this does not mean that the present invention can only use these two methods to simulate data truncation. Other methods can also be used to simulate data truncation to simultaneously obtain the virtual distorted image (known input value) and the control image (expected output value) required for AI network learning.

[0062] The proposed embodiment of image reconstruction based on the forward projection of the predicted image of the truncated portion can better improve the image quality of the untruncated portion and reduce the impact of artifacts caused by data truncation. Furthermore, by using an AI-based method to acquire the predicted image of the truncated portion, the projection data and / or image of the portion exceeding the scanning domain (i.e., the truncated portion) can be predicted more accurately, and the lines and boundaries of the detected object can be restored more precisely, providing strong evidence for radiotherapy and a powerful foundation for improving the image quality of the untruncated portion. By filling the truncated portion to obtain the initial image of the truncated portion (preprocessing step 110) or the virtual truncated data of the truncated portion (step 430 in data preparation), the learning network can be constructed better and more accurately, which is beneficial to improving the accuracy of the truncated portion prediction. In addition, by transforming the initial image of the truncated portion from an initial annular image to a matrix image in rectangular coordinates, the accuracy of the learning network prediction can also be improved, making the predicted image more precise.

[0063] Figure 8 A schematic diagram of a CT system 10 according to some embodiments of the present invention is shown. Figure 8 As shown, system 10 includes a rack 12 on which an X-ray source 14 and a detector array 18 are disposed opposite each other. The detector array 18 consists of multiple detectors 20 and a data acquisition system (DAS) 26. The DAS 26 is used to convert the sampled analog attenuated data received by the multiple detectors 20 into digital signals for subsequent processing. In some embodiments, system 10 is used to acquire projection data of the object under test at different angles. Therefore, the components on rack 12 are used to rotate around a rotation center 24 to acquire projection data. During rotation, the X-ray radiation source 14 projects X-rays 16 that penetrate the object under test toward the detector array 18. The attenuated X-ray beam data is preprocessed and used as projection data of the target volume of the object. Based on this projection data, an image of the object under test can be reconstructed. The reconstructed image can display the internal features of the object under test, including, for example, lesions, size, and shape of body tissue structures. The rotation center 24 of the rack also defines the center of the scanning domain 80.

[0064] System 10 further includes an image reconstruction module 50. As described above, the DAS 26 samples and digitizes projection data acquired by multiple detectors 20. The image reconstruction module 50 then performs high-speed image reconstruction based on the sampled and digitized projection data. In some embodiments, the image reconstruction module 50 stores the reconstructed image in a storage device or mass storage device 46. Alternatively, the image reconstruction module 50 transmits the reconstructed image to a computer 40 to generate patient information for diagnosis and evaluation.

[0065] although Figure 8The image reconstruction module 50 is illustrated as a separate entity, but in some embodiments, the image reconstruction module 50 may be part of the computer 40. Alternatively, the image reconstruction module 50 may not be present in the system 10, or the computer 40 may perform one or more functions of the image reconstruction module 50. Furthermore, the image reconstruction module 50 may be located locally or remotely and may be connected to the system 10 using wired or wireless networks. In some embodiments, computing resources centralized in a cloud network may be used for the image reconstruction module 50.

[0066] In some embodiments, system 10 includes a control mechanism 30. The control mechanism 30 may include an X-ray controller 34 for providing power and timing signals to the X-ray radiation source 14. The control mechanism 30 may also include a gantry controller 32 for controlling the rotational speed and / or position of the gantry 12 based on imaging requirements. The control mechanism 30 may also include a bed controller 36 for driving a bed 28 to a suitable position to position the object to be inspected within the gantry 12 to acquire projection data of the target volume of the object. Further, the bed 28 includes a drive mechanism, and the bed controller 36 can control the bed 28 by controlling the drive mechanism.

[0067] In some embodiments, system 10 further includes computer 40, to which images reconstructed by data sampled and digitized by DAS26 and / or image reconstruction module 50 are transmitted for processing. In some embodiments, computer 40 stores data and / or images in a storage device such as mass storage 46. Mass storage 46 may include hard disk drives, floppy disk drives, optical disc read / write (CD-R / W) drives, digital universal disk (DVD) drives, flash drives, and / or solid-state storage devices, etc. In some embodiments, computer 40 transmits reconstructed images and / or other information to display 42, which is communicatively connected to computer 40 and / or image reconstruction module 50. In some embodiments, computer 40 may connect to local or remote displays, printers, workstations, and / or similar devices, for example, such devices in medical institutions or hospitals, or remote devices via one or more configured wires or wireless networks such as the Internet and / or VPNs.

[0068] Furthermore, computer 40 can provide commands and parameters to DAS 26 and control mechanisms 30 (including rack controller 32, X-ray controller 34, and bed controller 36) based on user-provided and / or system-defined commands to control system operations, such as data acquisition and / or processing. In some embodiments, computer 40 controls system operations based on user input; for example, computer 40 can receive user input, including commands, scan protocols, and / or scan parameters, through an operator console 48 connected to it. Operator console 48 may include a keyboard (not shown) and / or a touchscreen to allow users to input / select commands, scan protocols, and / or scan parameters. Although Figure 7 Only one operator console 48 is shown as an example, but the computer 40 can connect to more consoles, for example, for inputting or outputting system parameters, requesting medical examinations, and / or viewing images.

[0069] In some embodiments, system 10 may include or be connected to an image storage and transmission system (PACS) (not shown). In some embodiments, the PACS is further connected to remote systems such as radiology information systems, hospital information systems, and / or internal or external networks (not shown) to allow operators located in different locations to provide commands and parameters, and / or access image data.

[0070] The methods or processes further described below may be stored as non-volatile memory on the computing device of the executable instruction storage system 10. For example, computer 40 may include executable instructions in non-volatile memory and may automatically execute part or all of the scanning process using the methods described herein, such as selecting a suitable protocol, determining suitable parameters, etc. Similarly, image reconstruction module 50 may include executable instructions in non-volatile memory and may perform image reconstruction tasks using the methods described herein.

[0071] Computer 40 can be configured and / or arranged for use in different ways. For example, in some implementations, a single computer 40 may be used; in other implementations, multiple computers 40 are configured to work together (e.g., based on a distributed processing configuration) or individually, each computer 40 being configured to handle specific aspects and / or functions, and / or process data for generating models that are only used for a specific medical imaging system 10. In some implementations, computer 40 may be local (e.g., co-located with one or more medical imaging systems 10, such as within the same facility and / or the same local network); in other implementations, computer 40 may be remote and therefore accessible only via a remote connection (e.g., via the Internet or other available remote access technologies). In a particular implementation, computer 40 may be configured in a cloud-like manner and may be accessed and / or used in a manner substantially similar to that used for accessing and using other cloud-based systems.

[0072] Once data (e.g., a trained learning network) is generated and / or configured, it can be copied and / or loaded into the medical imaging system 10, which can be done in various ways. For example, it can be loaded via a directed connection or link between the medical imaging system 10 and the computer 40. In this regard, communication between different components can be accomplished using available wired and / or wireless connections and / or according to any suitable communication (and / or network) standards or protocols. Alternatively or additionally, data can be loaded into the medical imaging system 10 indirectly. For example, data can be stored on a suitable machine-readable medium (e.g., a flash memory card, etc.) and then loaded into the medical imaging system 10 (on-site, such as by the system's user or authorized personnel) using that medium, or data can be downloaded to an electronic device capable of local communication (e.g., a laptop computer, etc.) and then used on-site (e.g., by the system's user or authorized personnel) to upload the data to the medical imaging system 10 via a direct connection (e.g., a USB connector, etc.).

[0073] Figure 9 An imaging system 500 according to some embodiments of the present invention is shown. For example... Figure 9 As shown, the imaging system 500 includes a prediction device 501, an image processing device 502, and an image reconstruction device 503.

[0074] The prediction device 501 is used to preprocess the projection data to obtain a predicted image of the truncated portion. In some embodiments, the prediction device 501 is used to predict the truncated portion using conventional methods, such as filling in the truncated portion of the projection data. In other embodiments, the prediction device 501 can also be used to predict the truncated portion based on artificial intelligence, specifically in conjunction with... Figure 10 To elaborate further.

[0075] Image processing device 502 is used to perform forward projection on the predicted image to obtain predicted projection data of the truncated portion. Image reconstruction device 503 is used to perform image reconstruction based on the predicted projection data and the projection data of the untruncated portion. In some embodiments, image reconstruction device 503 is as follows: Figure 8 The image reconstruction module 50 in the CT system 10 shown.

[0076] Figure 10 Shown according to Figure 9 The prediction device 501 in the system 500 shown. For example... Figure 10 As shown, the prediction device 501 includes a preprocessing device 510 and a control device 520.

[0077] The preprocessing unit 510 processes the truncated portion of the projection data based on the untruncated portion to obtain an initial image of the truncated portion. The control unit 520 calibrates the initial image based on a trained learning network to obtain a predicted image of the truncated portion.

[0078] In some embodiments, the preprocessing apparatus 510 includes a filling module (not shown). The filling module is used to fill the truncated portion of the projection data. Further, the filling module is further used to fill the truncated portion with the projection data information of the boundary of the untruncated portion (outermost channel), preferably, to fill the truncated portion of the corresponding channel with the projection data information of the boundary of the untruncated portion (outermost channel) of each channel.

[0079] In some embodiments, the preprocessing apparatus 510 further includes an image reconstruction module (not shown). The image reconstruction module is used to perform image reconstruction on the preprocessed (e.g., filled) projection data. In some embodiments, the image reconstruction module is as follows: Figure 8 The image reconstruction module 50 in the CT system 10 shown. In other embodiments, the preprocessing device 510 and the image reconstruction module 50 of the CT system are connected by a wired or wireless connection (including direct connection or indirect connection through a computer, etc.). After the filling module has filled the truncated portion of the projection data, the filling module (or the preprocessing device 510) can send the filled projection data to the image reconstruction module 50. After the image reconstruction module 50 completes the reconstruction, it sends the reconstructed CT image to the image extraction module (or the preprocessing device 510).

[0080] Optionally, the preprocessing device 510 further includes an image extraction module (not shown in the figure). The image extraction module is used to extract from the reconstructed CT image to obtain an initial image of the truncated portion.

[0081] In some embodiments, the control device 520 includes a transformation module 521, a calibration module 522, and an inverse transformation module 523.

[0082] The transformation module 521 is used to transform the pixels of the initial image of the truncated portion from polar coordinates to rectangular coordinates to obtain the pixel matrix of the initial image.

[0083] The calibration module 522 is used to calibrate the pixel matrix based on the trained learning network. In some embodiments, the calibration module 522 and the training module 540 are connected via a wired or wireless connection (including a direct connection or an indirect connection via a computer, etc.), and the training module 540 is configured to calibrate the pixel matrix based on the trained learning network. Figure 6 and Figure 7The method shown prepares virtual distorted images and control images in the learning network, and learns or trains based on an existing deep learning network model (e.g., Unet) to construct the learning network. In some embodiments, although Figure 10 The training module 540 is a separate module, but those skilled in the art will understand that the training module 540 can be integrated into the control device or other modules or devices.

[0084] The inverse transformation module 523 is used to transform the calibrated pixel matrix from rectangular coordinates to polar coordinates to obtain the predicted image of the truncated portion.

[0085] The present invention may also provide a non-transitory computer-readable storage medium for storing an instruction set and / or a computer program that, when executed by a computer, causes the computer to perform the imaging method described above. The computer executing the instruction set and / or the computer program may be a computer of a CT system or other devices / modules of a CT system. In one embodiment, the instruction set and / or the computer program may be programmed into the processor / controller of the computer.

[0086] Specifically, when this instruction set and / or computer program is executed by the computer, it causes the computer to:

[0087] Preprocess the projection data to obtain the predicted image of the truncated portion;

[0088] Forward projection is performed on the predicted image to obtain the predicted projection data of the truncated portion; and

[0089] Image reconstruction is performed based on the predicted projection data and the projection data of the uncropped portion.

[0090] As described above, instructions can be combined into a single instruction for execution, or any instruction can be split into multiple instructions for execution. Furthermore, the execution order of instructions is not limited to that described above.

[0091] In some embodiments, preprocessing the projection data to obtain a predicted image of the truncated portion may include:

[0092] The truncated portion of the projection data is processed based on the untruncated portion to reconstruct the initial image of the truncated portion; and

[0093] The initial image is calibrated based on the trained learning network to obtain a predicted image of the truncated portion.

[0094] In some embodiments, calibrating the initial image based on a trained learning network to obtain a predicted image of the truncated portion may include:

[0095] The pixels of the initial image of the truncated portion are transformed from polar coordinates to rectangular coordinates to obtain the pixel matrix of the initial image;

[0096] The pixel matrix is ​​calibrated based on the trained learning network; and

[0097] The calibrated pixel matrix is ​​transformed from Cartesian coordinates to polar coordinates to obtain the predicted image of the truncated portion.

[0098] In some embodiments, the instruction set and / or computer program further instruct the computer to perform training of the learning network, and further includes the following instructions:

[0099] Obtain the original image without data truncation;

[0100] The corresponding target object in the original image is virtually translated to partially move out of the scanning field, thereby obtaining a comparison image;

[0101] The comparison image is virtually scanned and virtual data is acquired to generate virtual truncated projection data; and

[0102] Image reconstruction is performed on the virtual truncated projection data to obtain a virtual distorted image.

[0103] Optionally, it also includes:

[0104] Coordinate transformations are performed on the virtual distorted image and the virtual precision standard image to obtain the virtual distorted pixel matrix and the virtual precision standard pixel matrix.

[0105] In other embodiments, instructing the computer to perform training of the learning network further includes the following instructions:

[0106] Coordinate transformations are performed on the virtual distorted image and the control image to obtain the virtual distorted pixel matrix and the virtual control pixel matrix.

[0107] In other embodiments, instructing the computer to perform training of the learning network further includes the following instructions:

[0108] Obtain the original image without data truncation, which is used as a control image;

[0109] Obtain the projection data corresponding to the original image;

[0110] The two channels of the projection data are filled to generate virtual truncated projection data; and

[0111] Image reconstruction is performed on the virtual truncated projection data to obtain a virtual distorted image.

[0112] Optionally, it also includes:

[0113] Coordinate transformations are performed on the virtual distorted image and the virtual precision standard image to obtain the virtual distorted pixel matrix and the virtual precision standard pixel matrix.

[0114] As used herein, the term "computer" can include any processor-based or microprocessor-based system, including systems that use microcontrollers, reduced instruction set computers (RISC), application-specific integrated circuits (ASICs), logic circuits, and any other circuitry or processors capable of performing the functions described herein. The examples above are merely illustrative and are not intended to limit the definition and / or meaning of the term "computer" in any way.

[0115] The instruction set may include various commands that instruct a computer or processor, acting as a processor, to perform specific operations, such as methods and processes according to various embodiments. The instruction set may take the form of a software program, which may be part of one or more tangible, non-transitory computer-readable media. The software may take various forms, such as system software or application software. Furthermore, the software may take the form of a collection of independent programs or modules, a program module within a larger program, or part of a program module. The software may also include modular programming in the form of object-oriented programming. Input data processing by the processor may be in response to operator commands, previous processing results, or requests made by another processor.

[0116] Some exemplary embodiments have been described above; however, it should be understood that various modifications can be made. For example, suitable results may be achieved if the described techniques are performed in a different order and / or if components in the described system, architecture, device, or circuit are combined in a different manner and / or replaced or supplemented by other components or their equivalents. Accordingly, other embodiments also fall within the scope of the claims.

Claims

1. An imaging method, comprising: The projection data is preprocessed to obtain the predicted image of the truncated portion; The predicted image is forward-projected to obtain the predicted projection data of the truncated portion; as well as Image reconstruction is performed based on the predicted projection data and the projection data of the uncropped portion. The step of preprocessing the projection data to obtain the predicted image of the truncated portion includes: The truncated portion of the projection data is processed based on the untruncated portion to reconstruct the initial image of the truncated portion; and The initial image is calibrated based on the trained learning network to obtain a predicted image of the truncated portion.

2. The method as described in claim 1, wherein, Filling the truncated portion of the projection data involves filling the truncated portion with the projection data information of the boundary of the untruncated portion.

3. The method as described in claim 1, wherein, The trained learning network calibrates the initial image of the truncated portion to obtain a predicted image of the truncated portion, including: The pixels of the initial image of the truncated portion are transformed from polar coordinates to rectangular coordinates to obtain the pixel matrix of the initial image; The pixel matrix is ​​calibrated based on the trained learning network, and The calibrated pixel matrix is ​​transformed from Cartesian coordinates to polar coordinates to obtain the predicted image of the truncated portion.

4. The method of claim 1, wherein, The training network is trained based on virtual distorted images and control images.

5. The method of claim 4, wherein, The training network is trained based on the pixel matrix obtained by coordinate transformation of the virtual distorted image and the pixel matrix obtained by coordinate transformation of the control image.

6. The method of claim 4, wherein, The method for obtaining the virtual distorted image and the control image includes: Obtain the original image without data truncation; The corresponding target object in the original image is virtually translated to partially move out of the scanning field, thereby obtaining a comparison image; The comparison image is virtually scanned and virtual data is acquired to generate virtual truncated projection data; and Image reconstruction is performed on the virtual truncated projection data to obtain a virtual distorted image.

7. The method of claim 4, wherein, The method for obtaining the virtual distorted image and the control image includes: Obtain the original image without data truncation, which is used as a control image; Obtain the projection data corresponding to the original image; The two channels of the projection data are filled to generate virtual truncated projection data; and Image reconstruction is performed on the virtual truncated projection data to obtain a virtual distorted image.

8. An imaging system comprising: A prediction device for preprocessing projection data to obtain a predicted image of the truncated portion; An image processing apparatus for forward projection of the predicted image to obtain predicted projection data of the truncated portion; as well as An image reconstruction apparatus for reconstructing an image based on the predicted projection data and the projection data of the uncropped portion. The prediction device includes: A preprocessing apparatus for processing a truncated portion of projection data based on the untruncated portion of the projection data to obtain an initial image of the truncated portion; and A control device for calibrating the initial image based on a trained learning network to obtain a predicted image of the truncated portion.

9. The system of claim 8, wherein, The control device includes: A transformation module is used to transform the pixels of the initial image of the truncated portion from polar coordinates to rectangular coordinates to obtain the pixel matrix of the initial image; A calibration module for calibrating the pixel matrix based on a trained learning network; and The inverse transformation module is used to transform the calibrated pixel matrix from Cartesian coordinates to polar coordinates to obtain the predicted image of the truncated portion.

10. A non-transitory computer-readable storage medium for storing a computer program that, when executed by a computer, causes the computer to perform the imaging method according to any one of claims 1-7.

Images