Truncation artifact correction method, CT image correction method, device and medium

Abstract

The present application relates to a truncation artifact correction method, a CT image correction method, a device and a medium. The truncation artifact correction method comprises: obtaining truncation artifact data according to the original projection data of the scanned object; the truncation artifact data is the projection data of each channel under the projection angle of view with truncation in the original projection data; correcting the truncation artifact data through a first correction network to obtain extended projection data; generating a medical image of the scanned object according to the original projection data and the extended projection data. The method directly analyzes and corrects the projection data in the data domain, so that the corrected extended projection data is closer to the real projection data. Furthermore, when performing truncation artifact correction in the data domain, it is only necessary to correct the projection data of each channel under the projection angle of view with truncation in the data domain, which reduces the amount of data calculation and improves the correction speed of truncation artifacts, thereby obtaining a complete medical image of the scanned object.

Description

Technical Field

[0001] The present application relates to the technical field of medical image processing, and in particular to a truncation artifact correction method, a CT image correction method, a device and a medium. Background Art

[0002] In clinical diagnosis and radiotherapy, due to limitations such as detector size and on-site scanning space, if part of the scanned object is outside the scanning range, the data acquired by the detector will be incomplete, resulting in bright truncation artifacts in the reconstructed medical image. As a result, the overall structure of the scanned object cannot be determined in the reconstructed medical image, and the scanned object cannot be accurately diagnosed or radiotherapy can be performed.

[0003] In the related art, when correcting truncation artifacts in an image, the projection data at the truncation point can be supplemented by using extrapolated data in the data domain, or artificial intelligence (AI) technology can be combined in the image domain to correct truncation artifacts in medical images.

[0004] However, whether it is the method of extrapolating data in the data domain or combining AI technology in the image domain, the correction effect of truncation artifacts is poor. Summary of the invention

[0005] Based on this, it is necessary to provide a truncation artifact correction method, CT image correction method, device and medium that can effectively correct truncation artifacts to obtain a complete medical image in response to the above technical problems.

[0006] In a first aspect, the present application provides a truncation artifact correction method, the method comprising:

[0007] According to the original projection data of the scanned object, truncation artifact data is obtained; the truncation artifact data is the projection data of each channel under the projection viewing angle where truncation exists in the original projection data;

[0008] Correcting the truncation artifact data through a first correction network to obtain expanded projection data;

[0009] A medical image of the scanned object is generated according to the original projection data and the expanded projection data.

[0010] In one embodiment, the process of constructing the first correction network includes:

[0011] Acquire a plurality of first sample projection data; each of the first sample projection data includes projection data of each channel under at least one projection viewing angle with truncation;

[0012] Constructing a loss function of a first correction network according to projection data of multiple channels at each projection viewing angle in multiple first sample projection data;

[0013] According to the plurality of first sample projection data, the preset first gold standard and the loss function of the first correction network, the first initial correction network is trained to obtain the first correction network.

[0014] In one embodiment, constructing a loss function of a first correction network according to projection data of multiple channels at each projection viewing angle in a plurality of first sample projection data includes:

[0015] According to the projection data of multiple channels at each projection viewing angle in the multiple first sample projection data, the sample ideal channel cumulative sum data after removing the truncation artifacts at each projection viewing angle is obtained;

[0016] According to the accumulated data of multiple sample ideal channels, a loss function of the first correction network is constructed.

[0017] In one embodiment, according to the projection data of multiple channels at each projection viewing angle in the multiple first sample projection data, obtaining the sample ideal channel cumulative sum data after removing the truncation artifact at each projection viewing angle includes:

[0018] According to the projection data of multiple channels at each projection viewing angle in the multiple first sample projection data, initial channel cumulative sum data of the projection data of multiple channels at each projection viewing angle is obtained;

[0019] The initial channel cumulative sum data is input into the second correction network to obtain the ideal channel cumulative sum data of the samples after removing the truncation artifacts at each projection viewing angle.

[0020] In one embodiment, the process of constructing the second correction network includes:

[0021] Acquire a plurality of second sample projection data; each second sample projection data includes projection data of each channel under at least one projection viewing angle with truncation;

[0022] According to the second sample projection data, sample channel cumulative sum data of multiple channels under each projection viewing angle is obtained;

[0023] The second initial correction network is trained according to the accumulated data of multiple sample channels and a preset second gold standard to obtain a second correction network.

[0024] In one embodiment, generating a medical image of a scanned object according to the original projection data and the expanded projection data includes:

[0025] According to the original projection data and the expanded projection data, target projection data after truncation artifact correction is obtained;

[0026] The target projection data is reconstructed to obtain a medical image of the scanned object.

[0027] In one embodiment, the target projection data after truncation artifact correction is obtained according to the original projection data and the expanded projection data, including:

[0028] The expanded projection data is added to the original projection data to obtain the corrected projection data;

[0029] The corrected projection data is corrected once or iteratively corrected multiple times to obtain target projection data after truncation artifact correction.

[0030] In a second aspect, the present application also provides a CT image correction method, the method comprising:

[0031] Acquire CT projection data of the scanned object; the CT projection data includes projection data of each channel under at least one projection viewing angle with truncation;

[0032] The CT projection data is corrected by a first correction network to obtain target projection data; the portion of the target projection data corresponding to the CT projection data that is truncated under the projection viewing angle is compensated;

[0033] A CT image of the scanned object is generated based on the target projection data.

[0034] In a third aspect, the present application further provides a truncation artifact correction device, the device comprising:

[0035] An acquisition module is used to acquire truncation artifact data according to the original projection data of the scanned object; the truncation artifact data is the projection data of each channel under the projection viewing angle where truncation exists in the original projection data;

[0036] A correction module, used for correcting the truncation artifact data through a first correction network to obtain expanded projection data;

[0037] The reconstruction module is used to generate a medical image of the scanned object based on the original projection data and the expanded projection data.

[0038] In a fourth aspect, the present application further provides a CT image correction device, the device comprising:

[0039] An acquisition module, used for acquiring CT projection data of a scanned object; the CT projection data includes projection data of each channel under at least one projection viewing angle with truncation;

[0040] A correction module, used to correct the CT projection data through a first correction network to obtain target projection data; the part of the target projection data corresponding to the CT projection data that is truncated under the projection viewing angle is compensated;

[0041] The reconstruction module is used to generate a CT image of the scanned object according to the target projection data.

[0042] In a fifth aspect, the present application further provides a computer device, which includes a memory and a processor, wherein the memory stores a computer program, and the processor implements the steps of any of the above method embodiments when executing the computer program.

[0043] In a sixth aspect, the present application further provides a computer-readable storage medium having a computer program stored thereon, which implements the steps of any of the above method embodiments when the computer program is executed by a processor.

[0044] In a seventh aspect, the present application further provides a computer program product, which includes a computer program, and when the computer program is executed by a processor, the steps of any of the above method embodiments are implemented.

[0045] The above-mentioned truncation artifact correction method, CT image correction method, device and medium obtain truncation artifact data according to the original projection data of the scanned object; the truncation artifact data is the projection data of each channel under the projection angle of view where truncation exists in the original projection data; the truncation artifact data is corrected by the first correction network to obtain the extended projection data; and the medical image of the scanned object is generated according to the original projection data and the extended projection data. In this method, considering that the projection data in the data domain is closer to the tissue structure information of the scanned object itself, when performing truncation artifact correction, the artifact correction is not performed on the scanned medical image in the image domain, but the truncation artifact data in the original projection data is analyzed and corrected in the data domain, so that the corrected extended projection data is closer to the real projection data of the scanned object, and the integrity and accuracy of the projection data are ensured. Furthermore, when performing truncation artifact correction in the data domain, only the projection data of each channel under the projection angle of view where truncation exists in the data domain is corrected, and its change trend is predicted by the first correction network, so as to determine the extended projection data corresponding to the truncation area. While reducing the amount of data calculation, the correction speed of truncation artifacts is improved. Therefore, the present application corrects the projection data at the truncation artifact in the data domain, and then generates a medical image based on the corrected expanded projection data, thereby improving the efficiency of truncation artifact correction and obtaining a complete and accurate medical image of the scanned object with better image quality. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] Figure 1 A diagram showing an application environment of a truncation artifact correction method in an embodiment;

[0047] Figure 2 is a schematic flow chart of a truncation artifact correction method in one embodiment;

[0048] Figure 3 A schematic diagram of a construction process of a first correction network in one embodiment;

[0049] Figure 4 A schematic diagram of a curve of channel accumulation and data corresponding to a projection viewing angle in one embodiment;

[0050] Figure 5 is a schematic diagram of training a second correction network in one embodiment;

[0051] Figure 6 is a schematic diagram of training a first correction network in one embodiment;

[0052] Figure 7 A schematic diagram of a method for supplementing truncation artifact data in one embodiment;

[0053] Figure 8 It is a schematic diagram of the process of obtaining the expanded artifact data in one embodiment;

[0054] Fig. 9 A schematic diagram of a process for generating a medical image in one embodiment;

[0055] Fig.10 is a schematic flow chart of a truncation artifact correction method in another embodiment;

[0056] Fig.11 is a schematic flow chart of a CT image correction method in one embodiment;

[0057] Fig.12 is a structural block diagram of a truncation artifact correction device in one embodiment;

[0058] Fig.13 is a structural block diagram of a CT image correction device in one embodiment;

[0059] Fig.14 FIG. 4 is a diagram showing the internal structure of a computer device in one embodiment. DETAILED DESCRIPTION

[0060] In order to make the purpose, technical solution and advantages of the present application more clearly understood, the present application is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0061] With the development of computed tomography (CT) technology, various scanning methods have emerged. Regardless of the scanning method used, in order to obtain high-quality medical images, the projection data of the scanned object must be low-noise and complete. However, in industry and medicine, due to the limitations of detector size, on-site scanning space and other conditions, there may be situations where the scanned object is outside the detector scanning range, resulting in incomplete projection data obtained by the detector. However, the discontinuity of projection data may cause truncation artifacts at the edges of the final generated medical image, seriously affecting the image quality, and then causing the physician to be unable to determine the complete tissue structure information of the scanned object based on the medical image, and unable to make an accurate diagnosis.

[0062] There are generally two ways to deal with truncation artifacts:

[0063] (1) Suppress the truncation artifacts generated during filtering and back-projection filtering by extrapolating data in the data domain.

[0064] However, this processing method can suppress the quality of the image inside the scanning field of view (FOV) when truncation artifacts occur, but the imaging quality of the image outside the scanning field of view is poor.

[0065] (2) Combine AI technology in the image domain to process truncation artifacts. The image information learned from the image domain is limited by the training data set used to train the neural network. In addition, when AI technology is used to directly correct truncation artifacts in medical images in the image domain, although the quality of the corrected image is improved, some image details may be changed in a sense, making them slightly different from the real object. If the corrected medical image is re-projected into the data domain, it will be found that the deviation of the image details is more obvious.

[0066] Based on at least one defect in the above-mentioned truncation artifact processing method, the present application provides a truncation artifact correction method to correct the truncation artifact data in combination with a trained correction network in the data domain, thereby generating a medical image of the scanned object based on the corrected complete and continuous projection data.

[0067] The truncation artifact correction method provided in this application can be applied to Figure 1 In the application environment shown. Figure 1As shown, the application environment may include an imaging device 110, an information processing device 120, and a storage device 130, and each device may be connected and communicated with each other through a network. For example, the imaging device 110 and the information processing device 120 may be connected or communicated through a network; the imaging device 110 and the storage device 130 may also be connected or communicated through a network; the information processing device 120 and the storage device 130 may also be connected or communicated through a network.

[0068] In some embodiments, the imaging device 110 may be a non-invasive biomedical imaging device for disease diagnosis or research purposes, for example, including a single-modality scanner and / or a multi-modality scanner. Among them, the single-modality scanner may include, for example, an ultrasound scanner, an X-ray scanner, a CT scanner, a magnetic resonance imaging (MRI) scanner, an ultrasound scanner, an optical coherence tomography (OCT) scanner, an ultrasound (Ultra Sound, US) scanner, an intravascular ultrasound (IVUS) scanner, a near infrared spectrum (NIRS) scanner, a far infrared (FIR) scanner, etc., or any combination of the above scanners. The multi-modality scanner may include, for example, an X-ray imaging-magnetic resonance imaging (X-ray-MRI) scanner, a single photon emission computed tomography-magnetic resonance imaging (SPECT-MRI) scanner, a positron emission tomography-computed tomography (PET-CT) scanner, a digital subtraction angiography-magnetic resonance imaging (DSA-MRI) scanner, etc. It should be understood that the scanner provided above is for illustrative purposes only and is not intended to limit the scope of the present application.

[0069] As an example, the imaging device 110 may specifically include a frame, a detector, a detection area, a scanning bed, and a radiation source. The frame may be used to support the detector and the radiation source, the scanning bed may be used to place a scanned object for scanning; the radiation source may emit radiation to the scanned object to irradiate the scanned object; and the detector may be used to receive radiation that passes through the scanned object.

[0070] Furthermore, the imaging device 110 may also include modules and / or components for performing imaging and / or related analysis. For example, the imaging device 110 may include a processor, which may be a part of the information processing device 120 and may perform the truncation artifact correction method provided in this application.

[0071] Optionally, the imaging device 110 may further include a display screen, which may be used to observe data information of the imaging device 110 and / or the scanned object scanned by the imaging device 110. For example, medical staff may observe lesion information of the chest, bones, breast and other detection parts of the scanned object through the display screen.

[0072] In some embodiments, the imaging device 110 may also send the acquired scan data (eg, projection data of the scanned object) to the information processing device 120 via the network for further analysis, processing and display, and / or send the acquired scan data to the storage device 130 via the network for storage.

[0073] The information processing device 120 may be specifically at least one computer device other than the imaging device, such as various personal computers, laptops, smart phones, tablet computers, portable wearable devices or any combination thereof; or a single server or a server group, which may be centralized or distributed.

[0074] In some embodiments, the storage device 130 may store data, instructions and / or any other information. For example, the storage device 130 may store scan data of the scanned object acquired by the imaging device 110 and / or a medical image of the scanned object processed by the information processing device 120.

[0075] As an example, the storage device 130 may include one or more of a mass storage device, a removable storage device, a volatile read-write storage device, a read-only storage device, and the like.

[0076] It should be noted that the above description is provided for illustrative purposes only and is not intended to limit the scope of the present application. For those of ordinary skill in the art, under the guidance of the content of the present application, various changes and modifications can be made. The features, structures, methods and other features of the exemplary embodiments described in this application can be combined in various ways to obtain additional and / or alternative exemplary embodiments, and these changes and modifications will not deviate from the scope of the present application. For example, the storage device 130 can be a data storage device including a cloud computing platform (such as a public cloud, a private cloud, a community and a hybrid cloud, etc.).

[0077] Next, the technical solution of the embodiment of the present application and how the technical solution of the embodiment of the present application solves the above-mentioned technical problems will be described in detail through embodiments and in combination with the accompanying drawings. The following specific method embodiments can be combined with each other, and the same or similar concepts or processes may not be repeated in some embodiments. It should be noted that a truncation artifact correction method provided in the embodiment of the present application can be executed by Figure 1The imaging device 110 shown may also be a computer device other than the imaging device 110, and may also be specifically a truncation artifact correction device, which may be implemented as part or all of a processor by software, hardware, or a combination of software and hardware. Obviously, the described embodiments are only part of the embodiments of the present application, not all of them.

[0078] In one embodiment, Figure 2 As shown, a truncation artifact correction method is provided, which is applied to Figure 1 The information processing device is used as an example to illustrate the method, which includes the following steps:

[0079] Step 210: acquiring truncation artifact data according to the original projection data of the scanned object; the truncation artifact data is the projection data of each channel under the projection viewing angle where truncation exists in the original projection data.

[0080] The original projection data may be projection data obtained after the imaging device scans at least one part of the scanned object, and the scanned object may include a biological object and / or a non-biological object. For example, the target object may include a specific part of the human body, such as one or more parts of the head, chest, abdomen, etc. The scanned object may also be an artificial component of living or non-living organic and / or inorganic matter, which is not limited in this embodiment.

[0081] As an example, the projection data may be data collected by a detector. Generally, the detector is artificially defined to have three directions, namely, X, Y, and Z. Figure 1 As shown in the figure: the X-axis is the detector length, which reflects the number of acquisition units in each row of detectors; the Z-axis is the detector width, which reflects the number of rows of detection; and the Y-axis is the X-line direction. The original projection data in this application includes the data collected by each channel under each projection angle (View). The projection angle here is perpendicular to the direction in which the detector receives data and parallel to the direction in which the bed moves (corresponding to Figure 1 The Z-axis direction in the figure); each channel refers to the direction in which the detector receives data (see Figure 1 in the X-axis direction).

[0082] It should be noted that when scanning an object, due to the influence of detector size, on-site scanning space, and movement of the object, some channels may not collect projection data during the scanning process, or only collect part of the projection data, making the original projection data obtained after scanning discontinuous. In this case, truncation artifacts will exist in the medical image generated based on the discontinuous original projection data.

[0083] Based on this, after acquiring the original projection data of the scanned object, step 210 analyzes the original projection data, and screens out the projection data of each channel under the projection viewing angle with truncation, to obtain truncation artifact data.

[0084] It should be understood that the truncation artifact data is part of the projection data in the original projection data, not all of the projection data. Considering that when performing truncation artifact correction in the data domain, there are many data sets involved and the amount of data calculation is large, this embodiment only corrects the truncation artifact data in the original projection data, rather than optimizing the original projection data as a whole, which greatly reduces the amount of data processing and thus improves the correction speed.

[0085] Step 220: Correct the truncation artifact data through a first correction network to obtain expanded projection data.

[0086] The truncated artifact data is the partial projection data collected by each channel of the detector at the truncation point under the projection viewing angle, while the expanded projection data is the complete projection data that should be collected by each channel under the corresponding projection viewing angle in an ideal state.

[0087] In other words, the truncation artifact data is corrected by the first correction network, and the projection data not collected by each channel under the projection viewing angle at the truncation artifact is predicted by the first correction network, so as to obtain complete projection data.

[0088] In a possible implementation, the implementation process of step 220 may be: inputting the truncation artifact data into a pre-trained first correction network, and outputting the expanded projection data corresponding to the truncation artifact data through the first correction network.

[0089] Step 230: Generate a medical image of the scanned object according to the original projection data and the expanded projection data.

[0090] The medical image generated based on the original projection data and the expanded projection data is an image after eliminating truncation artifacts, and the image information is complete and accurate. Compared with the medical image directly generated based on the original projection data, or the medical image generated based on the original projection data and then corrected, the imaging quality is better.

[0091] In a possible implementation, the implementation process of step 230 may be: fusing the original projection data and the expanded projection data to obtain complete and continuous target projection data of the scanned object; and then reconstructing the image according to the target projection data to obtain a medical image of the scanned object.

[0092] It should be noted that the fusion processing includes at least two situations: one is to use the expanded projection data to replace the truncated artifact data in the original projection data; the other is to correct or supplement the truncated artifact data in the original projection data based on the corrected expanded projection data, which is not limited in this embodiment.

[0093] In the above truncation artifact correction method, truncation artifact data is obtained according to the original projection data of the scanned object; the truncation artifact data is the projection data of each channel under the projection angle of view where truncation exists in the original projection data; the truncation artifact data is corrected by the first correction network to obtain the extended projection data; and the medical image of the scanned object is generated according to the original projection data and the extended projection data. In this method, considering that the projection data in the data domain is closer to the tissue structure information of the scanned object itself, when performing truncation artifact correction, the artifact correction is not performed on the scanned medical image in the image domain, but the truncation artifact data in the original projection data is analyzed and corrected in the data domain, so that the corrected extended projection data is closer to the real projection data of the scanned object, and the integrity and accuracy of the projection data are ensured. Furthermore, when performing truncation artifact correction in the data domain, only the projection data of each channel under the projection angle of view where truncation exists in the data domain is corrected, and its change trend is predicted by the first correction network, so as to determine the extended projection data corresponding to the truncation area. While reducing the amount of data calculation, the correction speed of truncation artifacts is improved. Therefore, the present application corrects the projection data at the truncation artifact in the data domain, and then generates a medical image based on the corrected expanded projection data, thereby improving the efficiency of truncation artifact correction and obtaining a complete and accurate medical image of the scanned object with better image quality.

[0094] Based on the above embodiments, Figure 3-6 , the construction process of the first correction network is further explained.

[0095] In one embodiment, Figure 3 As shown, the construction process of the first correction network includes the following steps:

[0096] Step 310: Acquire a plurality of first sample projection data; each first sample projection data includes projection data of each channel under at least one projection viewing angle with truncation.

[0097] The first sample projection data may be sample projection data acquired from a database, or may be real-time projection data acquired from an imaging device. This embodiment does not limit the data source of the first sample projection data.

[0098] It should be noted that due to the influence of truncation, the cumulative sum of the projection data of each channel at each projection angle in the data domain will have a significant decrease in the truncation area. Figure 4 , Figure 4 (a) shows the ideal channel cumulative sum data calculated based on the complete projection data collected by each channel under the projection angle when there is no truncation area. When truncation occurs, the channel cumulative sum data of the truncation area will drop significantly, such as Figure 4 As shown in (b).

[0099] If linear interpolation is used to correct the projection data of the truncated area, the result will still be discontinuous. In some special cases, the cumulative sum of the truncated area may have abnormal changes. At this time, if linear interpolation is used to fit, the opposite correction effect will be obtained, and the imaging quality will be worse.

[0100] Based on this, the present application uses the first correction network to correct the truncated artifact data. In order to obtain a first correction network with better correction effect, it is necessary to use the projection data of each channel in the first sample projection data to train the first initial correction network, so that it has the ability to predict the change trend of the channel accumulation and data under each projection angle under an ideal state, and predict the complete projection data of each channel under each projection angle.

[0101] Step 320: construct a loss function of the first correction network according to the projection data of multiple channels at each projection viewing angle in the multiple first sample projection data.

[0102] The loss function can be used to train the first initial correction network. When the output of the first initial correction network meets the corresponding convergence condition, the training is terminated to obtain the first correction network.

[0103] It should be noted that the loss function of the first correction network can include the following two cases:

[0104] (1) The loss function of the first correction network is the sample ideal channel cumulative sum data corresponding to the projection data of multiple channels at each projection viewing angle in the first sample projection data.

[0105] The sample ideal channel cumulative sum data represents the cumulative sum of projection data of each channel after removing truncation artifacts at each projection angle in the first sample projection data. After determining the ideal channel cumulative sum data, the first initial correction network is trained using the sample ideal channel cumulative sum data as a constraint condition.

[0106] (2) The loss function of the first correction network includes the sample ideal channel accumulation data and the preset loss function.

[0107] In this way, when training the first initial correction network, after adjusting the first initial correction network according to the loss value of the preset loss function, the training result of the first initial correction network can be verified again through the sample ideal channel accumulation data. The sample ideal channel accumulation and data corresponding to the same sample projection data before and after adjustment should be the same.

[0108] Optionally, the weight coefficients of the sample ideal channel cumulative sum data and the preset loss function may be preset, and the weight proportions of the ideal sample channel cumulative sum data and the preset loss function in the loss function of the first correction network may be determined according to the weight coefficients.

[0109] Furthermore, the implementation process of constructing the first loss function based on the sample ideal channel cumulative sum data can be: according to the projection data of multiple channels at each projection angle in multiple first sample projection data, obtain the sample ideal channel cumulative sum data after removing truncation artifacts at each projection angle; based on multiple sample ideal channel cumulative sum data, construct the loss function of the first correction network.

[0110] Wherein, when obtaining the sample ideal channel cumulative sum data corresponding to the first sample projection data, it can be achieved through a specific computer algorithm, or it can be achieved by combining AI technology and a trained neural network, which is not limited in this embodiment.

[0111] In one embodiment, the implementation process of obtaining sample ideal channel cumulative sum data can be: according to the projection data of multiple channels at each projection viewing angle in multiple first sample projection data, initial channel cumulative sum data of the projection data of multiple channels at each projection viewing angle is obtained; the initial channel cumulative sum data is input into the second correction network to obtain the sample ideal channel cumulative sum data after removing the truncation artifacts at each projection viewing angle.

[0112] The network type and training method of the second correction network and the first correction network may be the same or different, and there is no limitation on this. "First" and "second" are only used to distinguish them, and are not intended to limit their functions.

[0113] In a possible implementation, the construction process of the second correction network includes: acquiring multiple second sample projection data; the second sample projection data includes projection data of each channel under at least one projection viewing angle with truncation; according to the second sample projection data, acquiring sample channel cumulative sum data of multiple channels under each projection viewing angle; according to the multiple sample channel cumulative sum data and a preset second gold standard, training the second initial correction network to obtain the second correction network.

[0114] Each second sample projection data includes projection data of each channel under at least one projection viewing angle with truncation, and the second gold standard is the annotated ideal channel cumulative sum data corresponding to the pre-calculated second sample projection data without truncation.

[0115] That is, when constructing the second correction network, the second sample projection data is input into the second initial correction network, and whether the second initial correction network has converged is determined based on the sample ideal channel accumulation data output by the second initial correction network. When the error between the sample ideal channel accumulation output by the second initial correction network and the second gold standard is less than the preset error value, the second initial correction network is considered to have converged, the training is completed, and a trained second correction network is obtained.

[0116] As an example, see Figure 5 The second correction network training schematic diagram shown in the figure, when the second sample projection data is used to train the second initial correction network, the learning status of the second initial correction network is judged in combination with the second gold standard. In addition, the loss of the second initial correction network is calculated in real time during the training process, and the network parameters of the second initial correction network are adjusted according to the loss calculation result, so that the sample ideal channel accumulation and data output by the second initial correction network can approach the second gold standard.

[0117] In addition, the trained second correction network can be used alone or combined with other networks. The second correction network can be used to predict the ideal channel accumulation and data corresponding to the projection data of multiple channels under each projection viewing angle.

[0118] Step 330: According to the plurality of first sample projection data, the preset first gold standard and the loss function of the first correction network, the first initial correction network is trained to obtain the first correction network.

[0119] Each first sample projection data includes projection data of each channel under at least one projection viewing angle with truncation, and the first gold standard is the annotated ideal channel cumulative sum data corresponding to the pre-calculated first sample projection data without truncation artifacts.

[0120] Furthermore, when constructing the first correction network, a plurality of first sample projection data are input into the first initial correction network, and whether the first initial correction network has converged is determined based on the expanded projection data output by the first initial correction network. When the error between the sample channel cumulative sum data corresponding to the sample expanded projection data output by the first initial correction network and the first gold standard is less than a preset error value, the first initial correction network has converged, the training is completed, and a trained first correction network is obtained.

[0121] During the training process, the loss function of the first correction network is needed to control the sample expansion projection data output by the first initial correction network to simultaneously satisfy the sample ideal channel cumulative sum data. That is, the sample expansion projection data predicted by the first initial correction network should match the sample ideal channel cumulative sum data under the corresponding projection viewing angle and cannot exceed the range of the sample ideal channel cumulative sum data.

[0122] As an example, see Figure 6 The first correction network training schematic diagram shown in the figure, when the first sample projection data is used to train the first initial correction network, the learning status of the first initial correction network is judged in combination with the first gold standard. At the same time, the corresponding sample channel cumulative sum data is calculated in combination with the sample expanded projection data output by the first initial correction network, and it is judged whether the sample channel cumulative sum data meets the ideal channel cumulative sum data. In addition, the loss of the first initial correction network is calculated in real time during the training process, and according to the loss calculation result, the network parameters of the first initial correction network are adjusted so that the sample channel cumulative sum data of the sample expanded projection data output by the first initial correction network can approach the first gold standard.

[0123] In this embodiment, the loss function of the first correction network is constructed by using the projection data of multiple channels under each projection viewing angle, and then the first sample projection data and the loss function of the first correction network are used to train the first initial correction network to obtain the first correction network. Therefore, when the trained first correction network is applied, it can predict the ideal channel cumulative sum data of each channel under the projection viewing angle when the truncation artifact is eliminated based on the input truncation artifact data, and predict the complete projection data of the truncation area based on the ideal channel cumulative sum data to obtain the expanded projection data. In this way, by correcting the truncation artifact data in the data domain through the first correction network, the complete projection data of the truncation area can be obtained, ensuring the integrity and accuracy of the projection data.

[0124] Based on the above embodiments, in one embodiment, the second correction network can be integrated into the first correction network as a subnetwork or intermediate layer of the first correction network, and is used to output the ideal channel cumulative sum data corresponding to the truncated area when the first correction network predicts the expanded projection data, so as to limit the prediction range of the expanded projection data.

[0125] like Figure 7 As shown in (a), the traditional method of correcting the projection data at the truncated area is to use elliptical filling in the data domain to supplement the uncollected projection data, thereby obtaining the expanded projection data. However, in some cases, the shape of the real image edge is not elliptical, so a neural network is needed to learn the data features of the truncated area, so as to predict the change trend of the projection data corresponding to the correct image edge.

[0126] See also Figure 7 In (b), after the first correction network is trained using the first sample projection data, the first correction network can be used to predict the complete projection data corresponding to the truncated artifact data, that is, to determine the expanded projection data that is more consistent with the actual information of the truncated area.

[0127] See also Figure 8 In a possible implementation, the implementation process of predicting the extended projection data by the first correction network can be: extracting truncation artifact data from the original projection data, and inputting the truncation artifact data into the first correction network. The truncation artifact data is analyzed by the first correction network to predict the ideal channel cumulative sum data corresponding to the projection data of each channel under the projection viewing angle with truncation under an ideal state. At the same time, based on the truncation artifact data, the complete projection data of each channel under the projection viewing angle with truncation is predicted, that is, the extended projection data. Furthermore, the extended projection data is corrected by the ideal channel cumulative sum data, so that the output result of the first correction network is more accurate.

[0128] It should be noted that the ideal channel cumulative sum data is the intermediate information of the first correction network, and can be directly output as the output item of the first correction network, or the ideal channel cumulative sum data may not be output and only the final expanded projection data may be output, which is not limited in this embodiment.

[0129] In this embodiment, the truncated artifact data is corrected through the trained first correction network to obtain extended projection data, thereby improving data processing efficiency; moreover, the extended projection data corrected from the data domain is more consistent with the actual situation of the truncated area of ​​the scanned object and has better imaging quality.

[0130] In one embodiment, after the expanded projection data corresponding to the truncation artifact in the original projection data is determined by the first correction network, the complete projection data can be determined according to the original projection data and the expanded projection data, thereby generating a medical image of the scanned object.

[0131] In this embodiment, if Fig. 9 As shown, the implementation process of generating a medical image of the scanned object according to the original projection data and the expanded projection data (i.e., the above step 230) may include the following steps:

[0132] Step 910: Acquire target projection data after truncation artifact correction according to the original projection data and the expanded projection data.

[0133] The target projection data is the projection data after truncation artifacts are eliminated, and is used to generate a medical image of the scanned object.

[0134] In a possible implementation, the implementation process of step 910 may be: supplementing the expanded projection data to the original projection data to obtain corrected projection data; correcting the corrected projection data once or iteratively to obtain target projection data after truncation artifact correction.

[0135] If the error between the channel cumulative sum data at each projection viewing angle in the corrected correction projection data and the ideal channel cumulative sum data is less than a preset error value, the correction is completed and the target projection data is obtained. The ideal channel cumulative sum data represents the cumulative sum of the projection data of each channel at each projection viewing angle in the original projection data after removing the truncation artifact.

[0136] As an example, the ideal channel cumulative sum data corresponding to the original projection data can be obtained through the above-mentioned second correction network. That is, the original projection data is input into the second correction network, and the second correction network predicts the ideal channel cumulative sum data under each projection viewing angle after removing the truncation artifact.

[0137] Furthermore, the implementation process of iteratively correcting the corrected projection data can be: reversely projecting the corrected projection data to obtain a first medical image with truncation artifacts removed, and then projecting the first medical image back into the data domain by forward projection to obtain new corrected projection data. Then, the new corrected projection data after this iteration is corrected again using the first correction network. Then, the corrected corrected projection data is further reversely projected to obtain a second medical image with truncation artifacts removed, and so on, and multiple iterative processes are performed until the error between the channel cumulative sum data at each projection angle in the corrected corrected projection data and the ideal channel cumulative sum data is less than a preset error value, and the corrected projection data at this time is used as the target projection data.

[0138] Step 920: reconstruct the target projection data to obtain a medical image of the scanned object.

[0139] Specifically, the target projection data is back-projected and reconstructed to obtain a medical image of the scanned object, wherein the medical image is an image after truncation artifacts are eliminated, and its tissue structure information is complete and accurate, and the edge contour is clearer.

[0140] In this embodiment, the expanded projection data is added to the original projection data to obtain the corrected projection data, and then the corrected projection data can be made closer to the true value through repeated iterations. In this way, image reconstruction is performed based on the target projection data obtained after repeated corrections to obtain a medical image with better image quality, which can clearly and comprehensively reflect the true tissue structure information of the scanned object.

[0141] Based on the above method embodiments, in one embodiment, Fig.10As shown, the present application also provides another truncation artifact correction method, which is also applied to Figure 1 The information processing device in the example is used to illustrate, and the method includes the following steps:

[0142] Step 1010: Acquire truncation artifact data and ideal channel accumulation sum data according to the original projection data of the scanned object.

[0143] The truncation artifact data is the projection data of each channel at the projection viewing angle where truncation exists in the original projection data, and the ideal channel cumulative sum data represents the cumulative sum of the projection data of each channel at each projection viewing angle in the original projection data after the truncation artifact is removed.

[0144] Furthermore, when obtaining the ideal channel accumulation and data corresponding to the original projection data, it can be achieved through the second correction network trained in the above-mentioned embodiment.

[0145] Step 1020: Correct the truncation artifact data through a first correction network to obtain expanded projection data.

[0146] Step 1030: Supplement the expanded projection data to the original projection data to obtain corrected projection data.

[0147] Step 1040: iteratively correct the corrected projection data according to the ideal channel cumulative sum data to obtain target projection data after truncation artifact correction.

[0148] In one possible implementation, the implementation process of step 1040 is: iteratively correct the corrected projection data. If the error between the channel cumulative sum data at each projection viewing angle in the corrected corrected projection data and the ideal channel cumulative sum data is less than a preset error value, the iterative correction ends and the corrected corrected projection data is used as the target projection data.

[0149] Step 1050: reconstruct the target projection data to obtain a medical image of the scanned object.

[0150] It should be noted that the implementation principle and technical effects of the steps of the truncation artifact correction method provided in this embodiment can be referred to the steps in any of the above embodiments, and will not be repeated here.

[0151] In one embodiment, based on the constructed first correction network and the second correction network, specifically in the field of CT scanning, the present application also provides a CT image correction method, see Fig.11 , this method can be applied in CT scanners and can also be applied to Figure 1 The information processing device comprises the following steps:

[0152] Step 1110: Acquire CT projection data of the scanned object; the CT projection data includes projection data of each channel under at least one projection viewing angle with truncation.

[0153] The CT projection data is the projection data obtained by scanning at least one part of the scanned object through a CT scanner. However, since the scanning field of view / scanning field of view that can be imaged by the CT scanner is smaller than the scanned object, part of the scanned object may be located outside the scanning field of view during the scanning process, and the obtained CT projection data will include projection data at the truncated part.

[0154] Step 1120: Correct the CT projection data through the first correction network to obtain target projection data; the truncated portion of the target projection data corresponding to the CT projection data under the projection viewing angle is compensated.

[0155] It should be noted that the construction method of the first correction network can refer to the above Figure 3 In the corresponding embodiment, the constructed first correction network has the ability to predict the change trend of the channel accumulation and data at each projection viewing angle under an ideal state, and to predict the complete projection data of each channel at each projection viewing angle.

[0156] Optionally, the first correction network may adopt the above-mentioned second correction network as a subnetwork or an intermediate layer to predict the ideal channel accumulation and data corresponding to the CT projection data through the second correction network.

[0157] In a possible implementation, the correction principle of the first correction network is: input the CT projection data into the first correction network, analyze the CT projection data through the first correction network, and predict the ideal channel cumulative sum data corresponding to the projection data of each channel under the projection angle of view. At the same time, based on the CT projection data, predict the complete projection data of each channel under the projection angle of view with truncation to compensate for the projection data at the truncation. Further, the compensated projection data is corrected by the ideal channel cumulative sum data, and finally the target projection data is output through the first correction network.

[0158] That is, the target projection data obtained in this step is the complete projection data obtained after compensating the projection data at the truncated portion of the CT projection data.

[0159] Step 1130: Generate a CT image of the scanned object according to the target projection data.

[0160] Specifically, the target projection data is back-projected and reconstructed to obtain a CT image of the scanned object. The CT image generated according to the target projection data is an image after truncation artifacts are eliminated, and its tissue structure information is complete and accurate, the edge contour is clearer, and the image quality is better.

[0161] In the embodiment, the projection data at the truncation in the CT projection data can be effectively compensated by the constructed first correction network, so as to output the corrected target projection data. Since the target projection data is the complete data obtained after compensation, the CT image quality generated based on the target projection data is better. In this way, the data at the truncation is effectively compensated in the data domain by the first correction network, thereby improving the efficiency of truncation artifact correction.

[0162] It should be understood that, although the various steps in the flowcharts involved in the above-mentioned embodiments are displayed in sequence according to the indication of the arrows, these steps are not necessarily executed in sequence according to the order indicated by the arrows. Unless there is a clear explanation in this article, the execution of these steps does not have a strict order restriction, and these steps can be executed in other orders. Moreover, at least a part of the steps in the flowcharts involved in the above-mentioned embodiments can include multiple steps or multiple stages, and these steps or stages are not necessarily executed at the same time, but can be executed at different times, and the execution order of these steps or stages is not necessarily to be carried out in sequence, but can be executed in turn or alternately with other steps or at least a part of the steps or stages in other steps.

[0163] Based on the same inventive concept, the embodiment of the present application also provides a truncation artifact correction device for implementing the truncation artifact correction method involved above. The implementation scheme for solving the problem provided by the device is similar to the implementation scheme recorded in the above method, so the specific limitations in one or more truncation artifact correction device embodiments provided below can refer to the limitations of the truncation artifact correction method above, and will not be repeated here.

[0164] In one embodiment, Fig.12 As shown, a truncation artifact correction device is provided, including: an acquisition module 1210, a correction module 1220 and a reconstruction module 1230, wherein:

[0165] The acquisition module 1210 is used to acquire truncation artifact data according to the original projection data of the scanned object; the truncation artifact data is the projection data of each channel under the projection viewing angle where truncation exists in the original projection data;

[0166] A correction module 1220, configured to correct the truncation artifact data through a first correction network to obtain expanded projection data;

[0167] The reconstruction module 1230 is used to generate a medical image of the scanned object according to the original projection data and the expanded projection data.

[0168] In one embodiment, the process of constructing the first correction network includes:

[0169] Acquire a plurality of first sample projection data; each of the first sample projection data includes projection data of each channel under at least one projection viewing angle where a truncation artifact exists or truncation exists;

[0170] Constructing a loss function of a first correction network according to projection data of multiple channels at each projection viewing angle in multiple first sample projection data;

[0171] According to the plurality of first sample projection data, the preset first gold standard and the loss function of the first correction network, the first initial correction network is trained to obtain the first correction network.

[0172] In one embodiment, constructing a loss function of a first correction network according to projection data of multiple channels at each projection viewing angle in a plurality of first sample projection data includes:

[0173] According to the projection data of multiple channels at each projection viewing angle in the multiple first sample projection data, the sample ideal channel cumulative sum data after removing the truncation artifacts at each projection viewing angle is obtained;

[0174] According to the accumulated data of multiple sample ideal channels, a loss function of the first correction network is constructed.

[0175] In one embodiment, according to the projection data of multiple channels at each projection viewing angle in the multiple first sample projection data, obtaining the sample ideal channel cumulative sum data after removing the truncation artifact at each projection viewing angle includes:

[0176] According to the projection data of multiple channels at each projection viewing angle in the multiple first sample projection data, initial channel cumulative sum data of the projection data of multiple channels at each projection viewing angle is obtained;

[0177] The initial channel cumulative sum data is input into the second correction network to obtain the ideal channel cumulative sum data of the samples after removing the truncation artifacts at each projection viewing angle.

[0178] In one embodiment, the process of constructing the second correction network includes:

[0179] Acquire a plurality of second sample projection data; each second sample projection data includes projection data of each channel under at least one projection viewing angle with truncation;

[0180] According to the second sample projection data, sample channel cumulative sum data of multiple channels under each projection viewing angle is obtained;

[0181] The second initial correction network is trained according to the accumulated data of multiple sample channels and a preset second gold standard to obtain a second correction network.

[0182] In one embodiment, the reconstruction module 1230 includes:

[0183] A data acquisition unit, used for acquiring target projection data after truncation artifact correction according to the original projection data and the expanded projection data;

[0184] The image reconstruction unit is used to reconstruct the target projection data to obtain a medical image of the scanned object.

[0185] In one embodiment, the data acquisition unit includes

[0186] A supplementing subunit, used to supplement the expanded projection data to the original projection data to obtain corrected projection data;

[0187] The correction subunit is used to perform one correction or multiple iterative corrections on the correction projection data to obtain target projection data after truncation artifact correction.

[0188] Each module in the above-mentioned truncation artifact correction device can be implemented in whole or in part by software, hardware or a combination thereof. Each module can be embedded in or independent of a computer device (e.g., Figure 1 The information processing device shown in any one of the information processing devices) can also be stored in the memory of the computer device in the form of software, so that the processor can call and execute the operations corresponding to the above modules.

[0189] Similarly, for the above-mentioned CT image correction method, the present application also provides a CT image correction device. In one embodiment, Fig.13 As shown, the CT image correction device 1300 includes: an acquisition module 1310, a correction module 1320 and a reconstruction module 1330, wherein:

[0190] An acquisition module 1310 is used to acquire CT projection data of a scanned object; the CT projection data includes projection data of each channel under at least one projection viewing angle with truncation;

[0191] The correction module 1320 is used to correct the CT projection data through the first correction network to obtain target projection data; the part of the target projection data corresponding to the CT projection data that is truncated under the projection viewing angle is compensated;

[0192] The reconstruction module 1330 is used to generate a CT image of the scanned object according to the target projection data.

[0193] It should be noted that the implementation solution provided by the CT image correction device to solve the problem is similar to the implementation solution recorded in the above method. Please refer to the definition of the CT image correction method above, and no further details will be given here.

[0194] In addition, each module in the above CT image correction device can be implemented in whole or in part by software, hardware or a combination thereof. Each module can be embedded in or independent of a computer device (e.g., Figure 1 The information processing device shown in any one of the information processing devices) can also be stored in the memory of the computer device in the form of software, so that the processor can call and execute the operations corresponding to the above modules.

[0195] In one embodiment, a computer device is provided, whose internal structure diagram can be as follows: Fig.12 As shown. The computer device includes a processor, a memory, a communication interface, a display screen and an input device connected through a system bus. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The communication interface of the computer device is used to communicate with an external terminal in a wired or wireless manner, and the wireless manner can be implemented through WIFI, an operator network, NFC (near field communication) or other technologies. When the computer program is executed by the processor, the truncation artifact correction method and the CT image correction method provided in the present application can be implemented. The display unit of the computer device can be a liquid crystal display or an electronic ink display, and the input device of the computer device can be a touch layer covered on the display screen, or a button, a trackball or a touchpad provided on the housing of the computer device, or an external keyboard, touchpad or mouse.

[0196] Those skilled in the art will understand that Fig.12 The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.

[0197] In one embodiment, a computer device is provided, including a memory and a processor, wherein a computer program is stored in the memory, and when the processor executes the computer program, the following steps are implemented:

[0198] According to the original projection data of the scanned object, truncation artifact data is obtained; the truncation artifact data is the projection data of each channel under the projection viewing angle where truncation exists in the original projection data;

[0199] Correcting the truncation artifact data through a first correction network to obtain expanded projection data;

[0200] A medical image of the scanned object is generated according to the original projection data and the expanded projection data.

[0201] In addition, the processor may also implement the following steps when executing the computer program:

[0202] Acquire CT projection data of the scanned object; the CT projection data includes projection data of each channel under at least one projection viewing angle with truncation;

[0203] The CT projection data is corrected by a first correction network to obtain target projection data; the portion of the target projection data corresponding to the CT projection data that is truncated under the projection viewing angle is compensated;

[0204] A CT image of the scanned object is generated based on the target projection data.

[0205] The above embodiment provides a computer device, whose implementation principle and technical effect are similar to those of the above method embodiment, and will not be repeated here.

[0206] In one embodiment, a computer-readable storage medium is provided, on which a computer program is stored, and when the computer program is executed by a processor, the following steps are implemented:

[0207] According to the original projection data of the scanned object, truncation artifact data is obtained; the truncation artifact data is the projection data of each channel under the projection viewing angle where truncation exists in the original projection data;

[0208] Correcting the truncation artifact data through a first correction network to obtain expanded projection data;

[0209] A medical image of the scanned object is generated according to the original projection data and the expanded projection data.

[0210] In addition, when the computer program is executed, the following steps may be performed:

[0211] Acquire CT projection data of the scanned object; the CT projection data includes projection data of each channel under at least one projection viewing angle with truncation;

[0212] The CT projection data is corrected by a first correction network to obtain target projection data; the portion of the target projection data corresponding to the CT projection data that is truncated under the projection viewing angle is compensated;

[0213] A CT image of the scanned object is generated based on the target projection data.

[0214] The above embodiment provides a computer-readable storage medium, whose implementation principle and technical effect are similar to those of the above method embodiment, and will not be repeated here.

[0215] In one embodiment, a computer program product is provided, the computer program product comprising a computer program, and when the computer program is executed by a processor, the following steps are implemented:

[0216] According to the original projection data of the scanned object, truncation artifact data is obtained; the truncation artifact data is the projection data of each channel under the projection viewing angle where truncation exists in the original projection data;

[0217] Correcting the truncation artifact data through a first correction network to obtain expanded projection data;

[0218] A medical image of the scanned object is generated according to the original projection data and the expanded projection data.

[0219] In addition, when the computer program is executed, the following steps may be performed:

[0220] Acquire CT projection data of the scanned object; the CT projection data includes projection data of each channel under at least one projection viewing angle with truncation;

[0221] The CT projection data is corrected by a first correction network to obtain target projection data; the portion of the target projection data corresponding to the CT projection data that is truncated under the projection viewing angle is compensated;

[0222] A CT image of the scanned object is generated based on the target projection data.

[0223] The above embodiment provides a computer program product, whose implementation principle and technical effect are similar to those of the above method embodiment, and will not be repeated here.

[0224] Those of ordinary skill in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, storage, database or other media used in the embodiments provided in this application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory or optical memory, etc. Volatile memory can include random access memory (RAM) or external cache memory. As an illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM).

[0225] The technical features of the above embodiments may be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0226] The above-mentioned embodiments only express several implementation methods of the present application, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the invention patent. It should be pointed out that, for a person of ordinary skill in the art, several variations and improvements can be made without departing from the concept of the present application, and these all belong to the protection scope of the present application. Therefore, the protection scope of the patent of the present application shall be subject to the attached claims.

Claims

1. A truncation artifact correction method, characterized in that: The method comprises: According to the original projection data of the scanned object, truncation artifact data is acquired; the truncation artifact data is the projection data of each channel under the projection viewing angle where truncation exists in the original projection data; Inputting the truncation artifact data into a pre-trained first correction network, predicting the projection data not collected by each channel under the projection viewing angle with truncation, and obtaining the expanded projection data; A medical image of the scanned object is generated according to the original projection data and the expanded projection data.

2. The method according to claim 1, characterized in that The construction process of the first correction network includes: Acquire a plurality of first sample projection data; each of the first sample projection data includes projection data of each channel under at least one projection viewing angle with truncation; Constructing a loss function of the first correction network according to the projection data of multiple channels at each projection viewing angle in the multiple first sample projection data; The first correction network is trained according to the plurality of first sample projection data, a preset first gold standard and a loss function of the first correction network to obtain the first correction network.

3. The method according to claim 2, characterized in that The step of constructing the loss function of the first correction network according to the projection data of multiple channels at each projection viewing angle in the multiple first sample projection data includes: According to the projection data of multiple channels at each projection viewing angle in the multiple first sample projection data, acquiring sample ideal channel cumulative sum data after removing truncation artifacts at each projection viewing angle; A loss function of the first correction network is constructed based on the cumulative sum data of the plurality of sample ideal channels.

4. The method according to claim 3, characterized in that: The step of obtaining the sample ideal channel cumulative sum data after removing the truncation artifacts at each projection angle according to the projection data of multiple channels at each projection angle in the multiple first sample projection data comprises: Acquire initial channel cumulative sum data of the projection data of the multiple channels at each projection viewing angle according to the projection data of the multiple channels at each projection viewing angle in the multiple first sample projection data; The initial channel cumulative sum data is input into the second correction network to obtain the sample ideal channel cumulative sum data after removing the truncation artifacts at each projection viewing angle.

5. The method according to claim 4, characterized in that The construction process of the second correction network includes: Acquire a plurality of second sample projection data; each of the second sample projection data includes projection data of each channel under at least one projection viewing angle with truncation; According to the second sample projection data, sample channel cumulative sum data of multiple channels under each projection viewing angle is acquired; The second initial correction network is trained according to the accumulated sum data of the plurality of sample channels and a preset second gold standard to obtain the second correction network.

6. The method according to any one of claims 1 to 5, characterized in that: Generating a medical image of the scanned object according to the original projection data and the expanded projection data includes: Acquire target projection data after truncation artifact correction according to the original projection data and the expanded projection data; Image reconstruction is performed on the target projection data to obtain a medical image of the scanned object.

7. The method according to claim 6, characterized in that The step of obtaining target projection data after truncation artifact correction according to the original projection data and the expanded projection data includes: Complementing the expanded projection data to the original projection data to obtain corrected projection data; The corrected projection data is corrected once or iteratively corrected multiple times to obtain the target projection data after the truncation artifact correction.

8. A CT image correction method, characterized in that: The method comprises: Acquire CT projection data of the scanned object; the CT projection data includes projection data of each channel under at least one projection viewing angle with truncation; Inputting the CT projection data into a first correction network, predicting ideal channel cumulative sum data corresponding to the projection data of each channel under the projection viewing angle, predicting complete projection data of each channel under the projection viewing angle with truncation, and compensating the projection data of each channel under the projection viewing angle with truncation based on the complete projection data to obtain compensated projection data, and correcting the compensated projection data by using the ideal channel cumulative sum data to obtain target projection data; the part of the target projection data corresponding to the truncation of the CT projection data under the projection viewing angle is compensated; A CT image of the scanned object is generated according to the target projection data.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 8 are implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 8 are implemented.

Images