Data processing method and device

Abstract

The invention discloses a data processing method and device which are applied to the field of computers, and computer equipment firstly performs truncation correction on first projection data in each truncation correction process so as to repair data outside a scanning view field in a projection domain, so that an image outside the scanning view field can be reconstructed, and meanwhile, the image outside the scanning view field can be reconstructed. The quality of data outside the scanning view in the reconstructed image is improved, and artifacts outside the scanning view in the reconstructed image are reduced. Afterwards, after the first image is determined by adopting the replaced second projection data, truncation correction is further performed on the first image, so that data outside the scanning visual field is further restored in the image domain, the quality of the data outside the scanning visual field in the reconstructed image is further improved, and artifacts outside the scanning visual field in the reconstructed image are further reduced.

Description

Technical Field

[0001] This application relates to the field of computers, specifically to a data processing method and apparatus. Background Technology

[0002] During a computed tomography (CT) scan, the object being scanned typically needs to be positioned at the center of rotation of the CT scanner to ensure that the scanning range covers the object.

[0003] However, in certain special cases, such as when the subject is excessively obese or unable to move and can only lie on the edge of a bed, the CT scan range may not cover the subject. This causes the attenuation value of the scan data to abruptly drop to 0 at the edge of the CT scanner's detector; this situation is called truncated scan data. Based on this scan data, not only is it impossible to reconstruct an image outside the scan's field of view, but the reconstructed image will also exhibit truncation artifacts, resulting in low-quality reconstructed images. Summary of the Invention

[0004] This invention proposes a data processing method and apparatus that solves the problems in related technologies where truncated scan data not only fails to reconstruct images outside the scanned field of view, but also results in truncated artifacts in the reconstructed images, leading to low image quality. The technical solution includes:

[0005] On the one hand, a data processing method is provided, which includes:

[0006] Obtain the first projection data after truncation and padding, which includes the original projection data;

[0007] Perform the truncation correction procedure at least once. The truncation correction procedure includes:

[0008] The first projection data is truncated and corrected to obtain the second projection data, and the original projection data is used to replace the data within the scanning field of view in the second projection data.

[0009] The first image is determined based on the replaced second projection data, and the first image is truncated and corrected to obtain the second image;

[0010] The third projection data is determined based on the second image, and the original projection data is used to replace the data within the scanned field of view in the third projection data;

[0011] Update the first projection data to the replaced third projection data.

[0012] Optionally, the second projection data is obtained by inputting the first projection data into a projection data truncation correction network for truncation correction.

[0013] Optionally, the method further includes:

[0014] Acquire multiple full-scan field-of-view images;

[0015] For each full-scan field-of-view image, based on the sample sub-images in the full-scan field-of-view image, the truncated first sample projection data is obtained, and the first sample projection data is extrapolated to obtain untruncated training projection data.

[0016] Obtain untruncated label projection data based on full-scan field-of-view images;

[0017] A projection data truncation correction network was trained using multiple training projection data and multiple label projection data.

[0018] The sample sub-images include portions of the sample objects in the full-scan field-of-view image. The images corresponding to the training projection data are untruncated but have artifacts, while the label images corresponding to the label projection data are untruncated and have no artifacts.

[0019] Optionally, a projection data truncation correction network is trained using multiple training projection data and multiple label projection data, including:

[0020] Multiple training projection data are input into the first initial network to obtain multiple corrected projection data output by the first initial network;

[0021] For each corrected projection data, determine the first loss value between the data outside the scan field of view in the corrected projection data and the data outside the scan field of view in the corresponding label projection data, and obtain multiple first loss values;

[0022] The first initial network is adjusted based on the mean of multiple first loss values ​​to obtain the projection data truncation correction network.

[0023] Optionally, the method also includes:

[0024] For each corrected projection data, determine the first data difference between the data of every two adjacent extended channels outside the scanning field of view in the corrected projection data, determine the second data difference between the data of every two adjacent extended channels outside the scanning field of view in the corresponding label projection data, and determine the second loss value of the first data difference and the second data difference to obtain multiple second loss values.

[0025] The first initial network is adjusted based on the mean of multiple first loss values ​​to obtain the projection data truncation correction network, including:

[0026] The first weighted loss value is obtained by weighting and summing the mean of multiple first loss values ​​and the mean of multiple second loss values; wherein the weight of the mean of multiple second loss values ​​is less than the weight of the mean of multiple first loss values.

[0027] The first initial network is adjusted based on the first weighted loss value to obtain the projection data truncation correction network.

[0028] Optionally, the first image is truncated and corrected to obtain a second image, including:

[0029] The first image is input into the image truncation correction network to perform truncation correction on the first image, thereby obtaining the second image.

[0030] Optionally, the method also includes:

[0031] Acquire multiple full-scan field-of-view images;

[0032] For each full-scan field-of-view image, based on the sample sub-images in the full-scan field-of-view image, the truncated first sample projection data is obtained, and the first sample projection data is extrapolated to obtain untruncated training projection data.

[0033] Obtain untruncated label projection data based on full-scan field-of-view images;

[0034] For each training projection data, the training projection data is truncated and corrected to obtain the second sample projection data. The data in the scanned field of view in the second sample projection data is replaced with the first sample projection data to obtain the third sample projection data. The training image without truncation and with artifacts is obtained based on the third sample projection data.

[0035] For each label projection data, obtain a label image without truncation and without artifacts based on the label projection data;

[0036] An image truncation correction network was trained using multiple training images and multiple labeled images.

[0037] Optionally, an image truncation correction network is trained using multiple training images and multiple label images, including:

[0038] Multiple training images are input into the second initial network to obtain multiple corrected images output by the second initial network;

[0039] For each corrected image, the third loss value is determined between the first data outside the scanned field of view in the corrected image and the second data outside the scanned field of view in the corresponding label image, resulting in multiple third loss values;

[0040] The second initial network is adjusted based on the mean of multiple third loss values ​​to obtain the image truncation correction network.

[0041] Optionally, the method also includes:

[0042] For each corrected image, the fourth loss value of the data in the scanned field of view in the corrected image and the data in the scanned field of view in the corresponding label image is determined, and multiple fourth loss values ​​are obtained.

[0043] The second initial network is adjusted based on the mean of multiple third loss values ​​to obtain the image truncation correction network, which includes:

[0044] The second weighted loss value is obtained by weighting and summing the mean of multiple third loss values ​​and the mean of multiple fourth loss values. The weight of the mean of multiple fourth loss values ​​is less than the weight of the mean of multiple third loss values.

[0045] The second initial network is adjusted based on the second weighted loss value to obtain the image truncation correction network.

[0046] Optionally, the third loss value is determined by a weighted sum of the average absolute value error loss value of the first data and the second data, and the structural similarity loss value of the first data and the second data.

[0047] Optionally, the training projection data is truncated and corrected to obtain the second sample projection data, including:

[0048] The training projection data is input into the projection data truncation correction network to obtain the second sample projection data.

[0049] Optionally, before updating the first projection data to the replaced third projection data, the method further includes:

[0050] The data outside the scanning field of view in the replaced third projection data are processed to achieve a smooth transition.

[0051] On the other hand, a computer device is provided, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the computer program, it implements the data processing method described above.

[0052] In another aspect, a data processing apparatus is provided, the apparatus comprising:

[0053] The acquisition module is used to acquire the first projection data that has been truncated and padded. The first projection data includes the original projection data.

[0054] The calibration module is used to: execute a truncation calibration process at least once, the truncation calibration process including:

[0055] The first projection data is truncated and corrected to obtain the second projection data, and the original projection data is used to replace the data within the scanning field of view in the second projection data.

[0056] The first image is determined based on the replaced second projection data, and the first image is truncated and corrected to obtain the second image;

[0057] The third projection data is determined based on the second image, and the original projection data is used to replace the data within the scanned field of view in the third projection data;

[0058] Update the first projection data to the replaced third projection data.

[0059] In summary, the embodiments of the present invention provide a data processing method and apparatus. In this method, during a truncation correction process, the computer device first performs truncation correction on the first projection data to repair data outside the scanning field of view in the projection domain. This improves the quality of the data outside the scanning field of view in the reconstructed image while reducing artifacts outside the scanning field of view. Then, after determining the first image using the replaced second projection data, the first image is further trunculated to further repair data outside the scanning field of view in the image domain. This further improves the quality of the data outside the scanning field of view in the reconstructed image while further reducing artifacts outside the scanning field of view, thereby improving the overall quality of the reconstructed image.

[0060] By performing multiple truncation correction processes, the quality of data outside the scanning field of view in the reconstructed image is effectively improved by repeatedly repairing the data in the projection domain and image domain. At the same time, the artifacts outside the scanning field of view in the reconstructed image are reduced, thereby effectively improving the quality of the reconstructed image.

[0061] Additional aspects and advantages of the invention will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of the invention. Attached Figure Description

[0062] Figure 1 This is a flowchart of a data processing method provided in an embodiment of the present invention;

[0063] Figure 2 This is a flowchart of another data processing method provided in an embodiment of the present invention;

[0064] Figure 3 This is a flowchart of a truncation correction process provided in an embodiment of the present invention;

[0065] Figure 4 It is an image reconstructed using methods provided by related technologies;

[0066] Figure 5 An image reconstructed using the method provided in the embodiments of the present invention;

[0067] Figure 6 It is an image reconstructed using a method provided by related technologies;

[0068] Figure 7 It is another image reconstructed using the method provided in the embodiments of the present invention;

[0069] Figure 8 This is a schematic diagram of the structure of a computer device provided in an embodiment of the present invention;

[0070] Figure 9 This is a block diagram of a data processing device provided in an embodiment of the present invention;

[0071] Figure 10 This is a block diagram of another data processing device provided in an embodiment of the present invention. Detailed Implementation

[0072] Embodiments of the present invention are described in detail below, examples of which are illustrated in the accompanying drawings, wherein the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and intended to explain the present invention, and should not be construed as limiting the present invention.

[0073] During a CT scan, the object being scanned usually needs to be positioned at the center of rotation of the CT scanner to ensure that the scanning range of the CT scanner can cover the object.

[0074] However, in certain special cases, due to the patient's position (e.g., the patient is unable to move and can only lie on the edge of the bed) or large size (e.g., the patient is obese), the CT scanner may not be able to cover the patient within its scanning range. This causes the attenuation value of the scan data to abruptly drop to 0 at the edge of the CT scanner's detector; this situation is called truncated scan data. Based on this scan data, not only is it impossible to reconstruct an image outside the scanning field of view, but the reconstructed image will also exhibit truncation artifacts, resulting in low-quality reconstructed images.

[0075] Furthermore, this truncation artifact can also affect diagnostic information such as CT values ​​and contrast in the reconstructed image, making diagnosis more difficult for doctors and increasing the risk of misdiagnosis.

[0076] Figure 1 This is a flowchart of a data processing method provided in an embodiment of the present invention. This method can be applied to computer devices, such as... Figure 1 As shown, the method includes:

[0077] Step 101: Obtain the first projection data of the truncated and padded parts.

[0078] The computer device can acquire truncated and padded first projection data, wherein the first projection data may include the original projection data.

[0079] Step 102: Perform the truncation correction process at least once.

[0080] After acquiring the raw projection data, the computer device can perform at least one truncation correction process, wherein the truncation correction process includes the following steps S1 to S4:

[0081] Step S1: Truncate and correct the first projection data to obtain the second projection data, and replace the data within the scanning field of view in the second projection data with the original projection data.

[0082] The computer equipment can truncate and correct the first projection data to obtain the second projection data, and replace the data within the scanned field of view in the second projection data with the original projection data.

[0083] By truncating and correcting the first projection data, data outside the scanning field of view is repaired in the projection domain, thereby improving the quality of data outside the scanning field of view in the reconstructed image and reducing artifacts outside the scanning field of view in the reconstructed image while being able to reconstruct the image outside the scanning field of view.

[0084] Step S2: Determine the first image based on the replaced second projection data, and perform truncation correction on the first image to obtain the second image.

[0085] The computer equipment can backproject the replaced second projection data to determine the first image, and then truncate and correct the first image to obtain the second image.

[0086] After determining the first image using the replaced second projection data, the first image is further truncated and corrected to further repair the data outside the scanning field of view in the image domain. This not only enables the reconstruction of the image outside the scanning field of view, but also further improves the quality of the data outside the scanning field of view in the reconstructed image and further reduces the artifacts outside the scanning field of view in the reconstructed image.

[0087] Step S3: Determine the third projection data based on the second image, and replace the data within the scanned field of view in the third projection data with the original projection data.

[0088] The computer device can orthographically project the second image to determine the third projection data, and replace the data within the scanned field of view in the third projection data with the original projection data.

[0089] Step S4: Update the first projection data to the replaced third projection data.

[0090] After receiving the replaced third projection data, the computer device can update the first projection data with the replaced third projection data.

[0091] In summary, the embodiments of the present invention provide a data processing method. In this method, during a truncation correction process, the computer device first performs truncation correction on the first projection data to repair data outside the scanning field of view in the projection domain. This improves the quality of the data outside the scanning field of view in the reconstructed image while reducing artifacts outside the scanning field of view. Then, after determining the first image using the replaced second projection data, the first image is further trunculated to further repair data outside the scanning field of view in the image domain. This further improves the quality of the data outside the scanning field of view in the reconstructed image while further reducing artifacts outside the scanning field of view, thereby improving the overall quality of the reconstructed image.

[0092] By performing multiple truncation correction processes, the quality of data outside the scanning field of view in the reconstructed image is effectively improved by repeatedly repairing the data in the projection domain and image domain. At the same time, the artifacts outside the scanning field of view in the reconstructed image are reduced, thereby effectively improving the quality of the reconstructed image.

[0093] Figure 2 This is a flowchart of another data processing method provided in an embodiment of the present invention. This method can be applied to computer devices, such as... Figure 2 As shown, the method may include:

[0094] Step 201: Obtain the first projection data of the truncated and padded data.

[0095] The computer device can acquire truncated and padded first projection data, which includes the original projection data.

[0096] Computer equipment can acquire truncated raw data, perform interpolation on the raw data to obtain truncated and padded raw data, and then use this truncated and padded raw data as the original projected data. For example, the interpolation process can include linear interpolation, polynomial interpolation, or spline function interpolation.

[0097] Raw data can include data from multiple scanning channels within the scanning field of view. In the case of the first execution of the truncation correction process, the raw data is interpolated to supplement the raw data with data from multiple extended channels outside the scanning field of view, thereby filling in the missing data in the raw data and obtaining untrunculated raw projection data.

[0098] The computer device can acquire this biometric data from the CT scanner. Alternatively, the computer device can pre-store the biometric data. This biometric data can be data obtained from scanning the object.

[0099] Step 202: Perform the truncation correction process at least once.

[0100] After acquiring the first projection data, the computer device can execute at least one truncation correction procedure, wherein, reference Figure 3 The truncation correction process may include the following steps 2021 to 2026:

[0101] Step 2021: Input the first projection data into the projection data truncation correction network to truncate and correct the first projection data to obtain the second projection data.

[0102] The computer device can input the first projection data into the projection data truncation correction network so that the projection data truncation correction network can truncate and correct the first projection data to obtain the second projection data.

[0103] It should be noted that, in the case of the first execution of the truncation correction process, the first projection data is the original projection data.

[0104] If the truncation correction process is not terminated, since the first projection data is updated to the replaced third projection data, the first projection data is the updated first projection data, which is also the replaced third projection data, when the truncation correction process is not executed for the first time.

[0105] In this embodiment of the invention, when the truncation correction process is executed for the first time, although the extrapolation method is used to supplement the missing data outside the scanning field of view in the raw data, the quality of the data supplemented by the extrapolation method may be poor. Therefore, the image reconstructed based on this raw data may still have artifacts, and the quality of the reconstructed image is still poor.

[0106] By truncating and correcting the first projection data, and further repairing the data outside the scanning field of view in the first projection data, the quality of the data outside the scanning field of view in the first projection data can be effectively improved.

[0107] In this embodiment of the invention, the method for a computer device to acquire projection data truncation correction network may include the following steps A1 to A4:

[0108] Step A1: Acquire multiple full-scan field-of-view images.

[0109] Computer equipment can acquire multiple full-field-of-view images. These full-field-of-view images are determined based on full-field-of-view scanning data, during which the scanned object is completely within the scanning field of view of the CT equipment.

[0110] Step A2: For each full-scan field-of-view image, obtain the truncated first sample projection data based on the sample sub-image in the full-scan field-of-view image, and perform extrapolation processing on the first sample projection data to obtain untruncated training projection data.

[0111] After acquiring multiple full-scan field-of-view images, the computer device, for each full-scan field-of-view image, obtains truncated first sample projection data based on sample sub-images within the full-scan field-of-view image. Then, extrapolation processing is performed on the first sample projection data to obtain untruncated training projection data.

[0112] The sample sub-image includes a portion of the sample object from the full-scan field-of-view image; the sample object can be the object being scanned. The image corresponding to this training projection data is an untruncated image with artifacts. The full-scan field-of-view image is determined based on the full-scan field-of-view scan data; during the acquisition of the full-scan field-of-view scan data, the scanned object is completely within the scanning field of view of the CT equipment.

[0113] As an optional implementation, the computer device can translate the sample object in the full-scan field-of-view image so that part of the sample object goes out of the scan field of view, and then orthographically project the sample sub-image within the first cropping box to obtain the truncated first sample projection data.

[0114] The first cropping frame can be circular. The center point of the first cropping frame overlaps with the center point of the full-scan field-of-view image, and the radius of the first cropping frame is smaller than the radius of the inscribed circle of the full-scan field-of-view image.

[0115] As an alternative implementation, the computer device can orthographically project the sample sub-image within the second cropping box in the full-scan field-of-view image to obtain the truncated first sample projection data. The second cropping box may include a portion of the sample object. The second cropping box can be circular, and its radius is smaller than the radius of the inscribed circle of the full-scan field-of-view image.

[0116] Step A3: Obtain untruncated label projection data based on the full-scan field-of-view image.

[0117] A computer device can acquire uncrunted label projection data based on a full-scan field-of-view image, thereby obtaining a first dataset. This first dataset may include multiple training projection data sets and multiple label projection data sets corresponding one-to-one with the training projection data sets. The label image corresponding to each label projection data set is an uncrunted and artifact-free image.

[0118] If the first sample projection data is determined based on an optional implementation method in step A2 above, the computer device can perform a line orthographic projection on the translated full-scan field-view image after translating the sample objects in the full-scan field-view image to obtain untruncated label projection data.

[0119] If the first sample projection data is determined based on another optional implementation of step A2 above, the computer device can perform orthographic projection on the full-scan field-of-view image to obtain untruncated label projection data.

[0120] After acquiring multiple training projection data and multiple label projection data, the computer device can perform at least one of the following operations on at least one training projection data and at least one label projection data corresponding to the at least one training projection data: random flipping operation, random rotation operation, and adding random noise operation, thereby enriching the projection data in the first dataset.

[0121] Step A4: Train the projection data truncation correction network using multiple training projection data and multiple label projection data.

[0122] After acquiring multiple training projection data and multiple label projection data, the computer device can use these multiple training projection data and multiple label projection data to train a projection data truncation correction network.

[0123] In this embodiment of the invention, the computer device can use a deep learning algorithm to train the multiple training projection data and the multiple label projection data to obtain a projection data truncation correction network.

[0124] Optionally, the process of training a projection data truncation correction network using multiple training projection data and multiple label projection data may include:

[0125] Step A41: Input multiple training projection data into the first initial network to obtain multiple corrected projection data output by the first initial network.

[0126] The first initial network can be a neural network, for example, the first initial network can be a UNet network.

[0127] Step A42: For each corrected projection data, determine the first loss value between the data outside the scanning field of view in the corrected projection data and the data outside the scanning field of view in the corresponding label projection data, and obtain multiple first loss values.

[0128] For each corrected projection data, the computer device can input the scan field-out data in the corrected projection data and the scan field-out data in the corresponding label projection data into the loss function to obtain the first loss value output by the loss function.

[0129] Step A43: Adjust the first initial network based on the mean of multiple first loss values ​​to obtain the projection data truncation correction network.

[0130] After obtaining multiple first loss values, the computer device can adjust the first initial network based on the average of the multiple first loss values ​​to obtain the projection data truncation correction network.

[0131] Adjusting the first initial network based on the mean of multiple first loss values ​​can make the inference results of the first initial network closer to the target.

[0132] For each corrected projection data, the computer device can determine the first data difference between the data of every two adjacent extended channels outside the scanning field of view in the corrected projection data, determine the second data difference between the data of every two adjacent extended channels outside the scanning field of view in the corresponding label projection data, and determine the second loss value of the first data difference and the second data difference, thus obtaining multiple second loss values.

[0133] Wherein, both the first data difference and the second data difference can be the absolute value of the difference between the data in the two extended channels. Furthermore, the number of first data differences and the number of second data differences can both be at least one.

[0134] After obtaining multiple first loss values ​​and multiple second loss values, the computer device can perform a weighted summation of the average of the multiple first loss values ​​and the average of the multiple second loss values ​​to obtain a first weighted loss value. Based on this first weighted loss value, the first initial network is adjusted to obtain the projection data truncation correction network.

[0135] The weight of the average of multiple second loss values ​​is less than the weight of the average of multiple first loss values. For example, the weight of the average of multiple second loss values ​​can be 0.5.

[0136] By taking into account the mean of multiple second loss values ​​during the adjustment of the first initial network, abrupt changes between expansion channels can be prevented, which could introduce artifacts.

[0137] In this embodiment of the invention, since the projection data truncation correction network is trained using a large amount of training projection data and label projection data, the projection data truncation correction network is used to truncate and correct the first projection data. This not only improves the efficiency of truncation correction, but also effectively improves the accuracy of truncation correction of the first projection data.

[0138] Step 2022: Replace the data within the scanned field of view in the second projection data with the original projection data.

[0139] After the computer equipment truncates and corrects the processed first projection data to obtain the second projection data, the original projection data can be used to replace the data within the scanned field of view in the second projection data.

[0140] During the truncation and correction process of the processed first projection data, the data within the scanning field of view in the processed first projection data may be affected, resulting in a decrease in the quality of the data within the scanning field of view in the obtained second projection data. In this embodiment of the invention, the original projection data is used to replace the data within the scanning field of view in the second projection data, which can ensure the quality of the data within the scanning field of view in the second projection data.

[0141] Step 2023: Determine the first image based on the replaced second projection data.

[0142] After replacing the data within the scanned field of view in the second projection data with the original projection data, the computer device can determine the first image based on the replaced second projection data.

[0143] Optionally, the computer device backprojects the replaced second projection data to obtain the first image.

[0144] Step 2024: Input the first image into the image truncation correction network to perform truncation correction on the first image and obtain the second image.

[0145] After obtaining the first image, the computer device can input the first image into the image truncation correction network to perform truncation correction on the first image and obtain the second image.

[0146] In this embodiment of the invention, the method for a computer device to acquire an image truncation correction network may include the following steps B1 to B4:

[0147] Step B1: For each training projection data, the training projection data is truncated and corrected to obtain the second sample projection data. The data within the scanned field of view in the second sample projection data is replaced with the first sample projection data to obtain the third sample projection data. The training image without truncation and with artifacts is obtained based on the third sample projection data.

[0148] For each training projection data, the computer device can input the training projection data into the projection data truncation correction network to obtain the second sample projection data. The data within the scanned field of view in the second sample projection data is replaced with the first sample projection data to obtain the third sample projection data. Based on the third sample projection data, a training image without truncation but with artifacts is obtained, thus obtaining a training image that has been repaired once but still has artifacts.

[0149] Optionally, the computer device can backproject the third sample projection data to obtain untruncated training images with artifacts.

[0150] Training projection data can be obtained using the methods described in A1 and A2 above.

[0151] Step B2: For each label projection data, obtain a label image without truncation and without artifacts based on the label projection data.

[0152] For each label projection data, the computer device can backproject the label projection data to obtain a label image without truncation and artifacts, thereby obtaining a second dataset, which may include multiple training images and label images corresponding one-to-one with the multiple training images.

[0153] After acquiring the second dataset, the computer device can perform at least one of the following operations on at least one training image and at least one label image corresponding to the at least one training image: random flipping, random rotation, and adding random noise, thereby enriching the images in the second dataset.

[0154] The label projection data can be obtained using the method described in A3 above.

[0155] Step B3: Train an image truncation correction network using multiple training images and multiple labeled images.

[0156] After acquiring multiple training images and multiple label images, the computer device can use these images to train an image truncation correction network.

[0157] In this embodiment of the invention, the computer device can use a deep learning algorithm to train the multiple training images and multiple label images to obtain an image truncation correction network.

[0158] In this embodiment of the invention, the process of training an image truncation correction network using multiple training images and multiple labeled images may include:

[0159] Step B31: Input multiple training images into the second initial network to obtain multiple corrected images output by the second initial network.

[0160] A computer device can input multiple training images into a second initial network to obtain multiple corrected images output by the second initial network. The second initial network can be a neural network, for example, a UNet network.

[0161] Step B32: For each corrected image, determine the third loss value of the first data outside the scanning field of view in the corrected image and the third loss value of the second data outside the scanning field of view in the corresponding label image, and obtain multiple third loss values.

[0162] For each corrected image, the computer device can input the first data and the corresponding second data into the loss function to obtain the third loss value output by the loss function.

[0163] Optionally, the third loss value is determined by a weighted sum of the average absolute value error loss value of the first data and the second data, and the structural similarity loss value of the first data and the second data.

[0164] The third loss value is Loss = L1(data_efov, label_efov) + SSIM(data_efov, label_efov).

[0165] `data_efov` represents the first data set, and `label_efov` represents the second data set. `L1(data_efov, label_efov)` represents the mean absolute error loss between the first and second data sets, and `SSIM(data_efov, label_efov)` represents the structural similarity loss between the first and second data sets.

[0166] Step B33: Adjust the second initial network based on the mean of multiple third loss values ​​to obtain the image truncation correction network.

[0167] After obtaining multiple third loss values, the computer device can adjust the parameters of the second initial network based on the average of the multiple third loss values ​​to obtain the image truncation correction network.

[0168] In this embodiment of the invention, the process of training an image truncation correction network using multiple training images and multiple labeled images after the computer device obtains multiple third loss values ​​may further include:

[0169] Step B34: For each corrected image, determine the fourth loss value of the data in the scanned field of view in the corrected image and the data in the scanned field of view in the corresponding label image, and obtain multiple fourth loss values.

[0170] Therefore, the computer device can perform a weighted summation of the mean of multiple third loss values ​​and the mean of multiple fourth loss values ​​to obtain a second weighted loss value, and adjust the second initial network based on the second weighted loss value to obtain the image truncation correction network. The weight of the mean of the multiple fourth loss values ​​is less than the weight of the mean of the multiple third loss values; for example, the weight of the multiple fourth loss values ​​can be 0.5.

[0171] Optionally, the computer device may determine a fourth loss value between a first mean of the data within the scanned field of view in the corrected image and a second mean of the data within the scanned field of view in the corresponding label image.

[0172] The fourth loss value can satisfy: 0.5×L1(mean(data_sfov), mean(label_sfov)), where mean(data_sfov) represents the first mean and mean(label_sfov) represents the second mean. L1(mean(data_sfov), mean(label_sfov)) represents the fourth loss value based on the first and second means.

[0173] In this embodiment of the invention, since the image truncation correction network is trained using a large number of training images and labeled images, using the image truncation correction network to perform truncation correction on the first image can improve the efficiency of truncation correction and effectively improve the accuracy of truncation correction on the first image.

[0174] Step 2025: Determine the third projection data based on the second image, and replace the data within the scanned field of view in the third projection data with the original projection data.

[0175] After the computer device performs truncation correction on the first image to obtain the second image, it can perform orthographic projection on the second image to determine the third projection data, and can use the original projection data to replace the data within the scanned field of view in the third projection data.

[0176] During the truncation correction process of the first image, the data within the scanned field of view in the first image may be affected, resulting in a decrease in the quality of the data within the scanned field of view in the obtained second image. In this embodiment of the invention, after determining the third projection data based on the second image, the data within the scanned field of view in the third projection data is replaced with the original projection data, thus ensuring the quality of the data within the scanned field of view in the third projection data.

[0177] Since replacing the data within the scanning field of view in the third projection data with the original projection data will introduce new artifacts, the computer device can also perform smooth transition processing on the data outside the scanning field of view in the replaced third projection data after obtaining the replaced third projection data, thereby eliminating the newly introduced artifacts and realizing the transition adjustment at the junction of the scanning field of view and the extended field of view in the replaced third projection data.

[0178] Optionally, the computer device can multiply the data outside the scanning field of view in the replaced third projection data by a target coefficient, thereby performing a smooth transition on the data outside the scanning field of view in the replaced third projection data. The target coefficient can be determined based on the ratio of a first coefficient to a second coefficient, where the first coefficient is the data of the first scanning channel within the scanning field of view in the replaced third projection data, and the second coefficient is the data of the target extended channel outside the scanning field of view in the replaced third projection data. The target extended channel is adjacent to the first scanning channel.

[0179] Alternatively, computer equipment can use spline interpolation to smooth the transition of data outside the scanning field of view in the replaced third projection data.

[0180] Step 2026: If the truncation correction process has not been terminated, update the first projection data to the replaced third projection data.

[0181] If the truncation correction process is not terminated, the computer device can update the first projection data with the replaced third projection data, thereby obtaining new first projection data, and then execute the truncation correction process again.

[0182] In the case of smoothing the transition of data outside the scanning field of view in the replaced third projection data, the replaced third projection data in step 2026 can be the third projection data after smoothing the transition.

[0183] The computer device can detect whether the number of iterations in the truncation correction process has reached the total number of iterations. If the total number of iterations has not been reached, it can be determined that the truncation correction process has not terminated. The computer device can pre-store the total number of iterations.

[0184] Related techniques are based on projection data (sine curves), using quadratic curve extrapolation to detect projection values ​​on the channels, resulting in reconstructed images with relatively clear boundaries. This corrects artifacts caused by truncation of projection values ​​within the field of view due to the patient being outside the scanning field, leading to inflated CT values ​​in the reconstructed image. Furthermore, the computational load is kept within a reasonable range. However, these techniques rely on the accuracy of predicting the number of extended channels. When the truncation is significant, such as truncating to areas like lung air, this method is difficult to perform well and is unsuitable for situations with high noise levels causing large data fluctuations.

[0185] In this embodiment of the invention, during the execution of a truncation correction process, a projection data truncation correction network is first used to truncate and correct the first projection data to repair data outside the scanning field of view in the projection domain. This improves the quality of the data outside the scanning field of view in the reconstructed image while reducing artifacts outside the scanning field of view. Then, after determining the first image using the replaced second projection data, an image truncation correction network is further used to truncate and correct the first image to further repair data outside the scanning field of view in the image domain. This further improves the quality of the data outside the scanning field of view in the reconstructed image while further reducing artifacts outside the scanning field of view, thereby improving the overall quality of the reconstructed image.

[0186] Furthermore, a projection data truncation correction network is used to truncate and correct the first projection data, and an image truncation correction network is used to truncate and correct the first image. This, combined with multiple iterations to continuously repair data outside the scanning field of view and adjusting the transition at the junction of the scanning and extended fields of view, significantly improves the quality of data outside the scanning field of view in the reconstructed image and significantly reduces artifacts outside the scanning field of view, thus greatly improving the overall image quality. Compared to related techniques, this method achieves better and more stable image quality. Moreover, it exhibits strong robustness and can adapt to data with varying degrees of truncation and noise levels.

[0187] Step 203: If the truncation correction process is terminated, reconstruct the image using the replaced third projection data.

[0188] If the truncation correction process is terminated, the computer equipment can reconstruct the image using the replaced third projection data.

[0189] If the number of iterations reaches the total number of iterations, the computer equipment can determine that the truncation correction process is terminated.

[0190] Figure 4 It is an image reconstructed using methods provided by related technologies. Figure 5 This is an image reconstructed using the method provided in the embodiments of the present invention. Figure 4 and Figure 5 It can be seen that the images reconstructed using the methods provided by related technologies have more artifacts, while the images reconstructed using the methods provided in the embodiments of the present invention have fewer artifacts.

[0191] Figure 6 It is an image reconstructed using a method provided by related technologies. Figure 7 This is an image reconstructed using another method provided in an embodiment of the present invention. From Figure 6 and Figure 7It can be seen that the images reconstructed using the methods provided by related technologies have more artifacts, while the images reconstructed using the methods provided in the embodiments of the present invention have fewer artifacts.

[0192] In summary, the embodiments of the present invention provide a data processing method. In this method, during a truncation correction process, the computer device first performs truncation correction on the first projection data to repair data outside the scanning field of view in the projection domain. This improves the quality of the data outside the scanning field of view in the reconstructed image while reducing artifacts outside the scanning field of view. Then, after determining the first image using the replaced second projection data, the first image is further trunculated to further repair data outside the scanning field of view in the image domain. This further improves the quality of the data outside the scanning field of view in the reconstructed image while further reducing artifacts outside the scanning field of view, thereby improving the overall quality of the reconstructed image.

[0193] By performing multiple truncation correction processes, the quality of data outside the scanning field of view in the reconstructed image is effectively improved by repeatedly repairing the data in the projection domain and image domain. At the same time, the artifacts outside the scanning field of view in the reconstructed image are reduced, thereby effectively improving the quality of the reconstructed image.

[0194] This invention provides a computer-readable storage medium storing a computer program thereon, which, when executed by a processor, implements the data processing method described in the above embodiments. For example, Figure 1 or Figure 2 The data processing method shown.

[0195] Figure 8 This is a schematic diagram of the structure of a computer device provided in an embodiment of the present invention, such as... Figure 8 As shown, the computer device includes a memory 801, a processor 802, and a computer program stored in the memory 801 and executable on the processor 802. When the processor 802 executes the computer program, it implements the data processing method described in the above embodiments. For example, Figure 1 or Figure 2 The data processing method shown.

[0196] Figure 9 This is a block diagram of a data processing device provided in an embodiment of the present invention, such as... Figure 9 As shown, the device includes:

[0197] The acquisition module 901 is used to acquire the first projection data that has been truncated and padded, the first projection data including the original projection data;

[0198] The calibration module 902 is configured to: execute at least one truncation calibration procedure, the truncation calibration procedure including:

[0199] The first projection data is truncated and corrected to obtain the second projection data, and the original projection data is used to replace the data within the scanning field of view in the second projection data.

[0200] The first image is determined based on the replaced second projection data, and the first image is truncated and corrected to obtain the second image;

[0201] The third projection data is determined based on the second image, and the original projection data is used to replace the data within the scanned field of view in the third projection data;

[0202] Update the first projection data to the replaced third projection data.

[0203] Optionally, the second projection data is obtained by inputting the first projection data into a projection data truncation correction network for truncation correction. (See reference) Figure 10 The device also includes: a first network training module 903, used for:

[0204] Acquire multiple full-scan field-of-view images;

[0205] For each full-scan field-of-view image, based on the sample sub-images in the full-scan field-of-view image, the truncated first sample projection data is obtained, and the first sample projection data is extrapolated to obtain the training projection data.

[0206] Obtain untruncated label projection data based on full-scan field-of-view images;

[0207] A projection data truncation correction network was trained using multiple training projection data and multiple label projection data.

[0208] The sample sub-images include portions of the sample objects in the full-scan field-of-view image. The images corresponding to the training projection data are untruncated but have artifacts, while the label images corresponding to the label projection data are untruncated and have no artifacts.

[0209] Optionally, the first network training module 903 is used for:

[0210] Multiple training projection data are input into the first initial network to obtain multiple corrected projection data output by the first initial network;

[0211] For each corrected projection data, determine the first loss value between the data outside the scan field of view in the corrected projection data and the data outside the scan field of view in the corresponding label projection data, and obtain multiple first loss values;

[0212] The first initial network is adjusted based on the mean of multiple first loss values ​​to obtain the projection data truncation correction network.

[0213] Optionally, the first network training module 903 is used for:

[0214] For each corrected projection data, determine the first data difference between the data of every two adjacent extended channels outside the scanning field of view in the corrected projection data, determine the second data difference between the data of every two adjacent extended channels outside the scanning field of view in the corresponding label projection data, and determine the second loss value of the first data difference and the second data difference to obtain multiple second loss values.

[0215] The first weighted loss value is obtained by weighting and summing the mean of multiple first loss values ​​and the mean of multiple second loss values; wherein the weight of the mean of multiple second loss values ​​is less than the weight of the mean of multiple first loss values.

[0216] The first initial network is adjusted based on the first weighted loss value to obtain the projection data truncation correction network.

[0217] Optional, the calibration module 902 is used for:

[0218] The first image is input into the image truncation correction network to perform truncation correction on the first image, thereby obtaining the second image.

[0219] refer to Figure 10 The device also includes: a second network training module 904, used for:

[0220] Acquire multiple full-scan field-of-view images;

[0221] For each of the full-scan field-of-view images, based on the sample sub-images in the full-scan field-of-view image, truncated first sample projection data is obtained, and the first sample projection data is extrapolated to obtain untruncated training projection data.

[0222] Untruncated label projection data is obtained based on the full-scan field-of-view image;

[0223] For each training projection data, the training projection data is truncated and corrected to obtain the second sample projection data. The data in the scanned field of view in the second sample projection data is replaced with the first sample projection data to obtain the third sample projection data. The training image without truncation and with artifacts is obtained based on the third sample projection data.

[0224] For each label projection data, obtain a label image without truncation and without artifacts based on the label projection data;

[0225] An image truncation correction network was trained using multiple training images and multiple labeled images.

[0226] Optionally, the second network training module 904 is used for:

[0227] Multiple training images are input into the second initial network to obtain multiple corrected images output by the second initial network;

[0228] For each corrected image, the third loss value is determined between the first data outside the scanned field of view in the corrected image and the second data outside the scanned field of view in the corresponding label image, resulting in multiple third loss values;

[0229] The second initial network is adjusted based on the mean of multiple third loss values ​​to obtain the image truncation correction network.

[0230] Optionally, the second network training module 904 is used for:

[0231] For each corrected image, the fourth loss value of the data in the scanned field of view in the corrected image and the data in the scanned field of view in the corresponding label image is determined, and multiple fourth loss values ​​are obtained.

[0232] The second weighted loss value is obtained by weighting and summing the mean of multiple third loss values ​​and the mean of multiple fourth loss values. The weight of the mean of multiple fourth loss values ​​is less than the weight of the mean of multiple third loss values.

[0233] The second initial network is adjusted based on the second weighted loss value to obtain the image truncation correction network.

[0234] Optionally, the third loss value is determined by a weighted sum of the average absolute value error loss value of the first data and the second data, and the structural similarity loss value of the first data and the second data.

[0235] The second network training module 904 is used for:

[0236] The training projection data is input into the projection data truncation correction network to obtain the second sample projection data.

[0237] Optionally, the calibration module 902 is also used for:

[0238] Before updating the first projection data to the replaced third projection data, a smooth transition process is performed on the data outside the scanning field of view in the replaced third projection data.

[0239] In summary, the embodiments of the present invention provide a data processing apparatus. During a truncation correction process, the apparatus first performs truncation correction on the first projection data to repair data outside the scanning field of view in the projection domain. This improves the quality of the data outside the scanning field of view in the reconstructed image while reducing artifacts outside the scanning field of view. Then, after determining the first image using the replaced second projection data, the first image is further trunculated to further repair data outside the scanning field of view in the image domain. This further improves the quality of the data outside the scanning field of view in the reconstructed image while further reducing artifacts outside the scanning field of view, thereby improving the overall quality of the reconstructed image.

[0240] By performing multiple truncation correction processes, the quality of data outside the scanning field of view in the reconstructed image is effectively improved by repeatedly repairing the data in the projection domain and image domain. At the same time, the artifacts outside the scanning field of view in the reconstructed image are reduced, thereby effectively improving the quality of the reconstructed image.

[0241] It should be noted that the logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as a sequenced list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by, or in conjunction with, an instruction execution system, apparatus, or device (such as a computer-based system, a processor-included system, or other system that can fetch and execute instructions from, an instruction execution system, apparatus, or device). For the purposes of this specification, "computer-readable medium" can be any means that can contain, store, communicate, propagate, or transmit programs for use by, or in conjunction with, an instruction execution system, apparatus, or device. More specific examples (a non-exhaustive list) of computer-readable media include: an electrical connection having one or more wires (electronic device), a portable computer disk drive (magnetic device), random access memory (RAM), read-only memory (ROM), erasable and editable read-only memory (EPROM or flash memory), fiber optic devices, and portable optical disc read-only memory (CDROM). Alternatively, the computer-readable medium may be paper or other suitable media on which the program can be printed, since the program can be obtained electronically, for example, by optically scanning the paper or other medium, followed by editing, interpreting, or otherwise processing as necessary, and then stored in a computer memory.

[0242] It should be understood that various parts of the present invention can be implemented in hardware, software, firmware, or a combination thereof. In the above embodiments, multiple steps or methods can be implemented in software or firmware stored in memory and executed by a suitable instruction execution system. For example, if implemented in hardware, as in another embodiment, it can be implemented using any one or a combination of the following techniques known in the art: discrete logic circuits having logic gates for implementing logical functions on data signals, application-specific integrated circuits (ASICs) having suitable combinational logic gates, programmable gate arrays (PGAs), field-programmable gate arrays (FPGAs), etc.

[0243] In the description of this specification, references to terms such as "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of the invention. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples.

[0244] In the description of this invention, it should be understood that the terms "center," "longitudinal," "lateral," "length," "width," "thickness," "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "top," "bottom," "inner," "outer," "clockwise," "counterclockwise," "axial," "radial," and "circumferential" indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are used only for the convenience of describing this invention and simplifying the description, and are not intended to indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on this invention.

[0245] Furthermore, the terms "first," "second," etc., used in the embodiments of this invention are for descriptive purposes only and should not be construed as indicating or implying relative importance, or implicitly specifying the number of technical features indicated in this embodiment. Therefore, features defined with terms such as "first" and "second" in the embodiments of this invention can explicitly or implicitly indicate that the embodiment includes at least one of those features. In the description of this invention, the word "multiple" means at least two or more, such as two, three, four, etc., unless otherwise explicitly specified in the embodiments.

[0246] In this invention, unless otherwise explicitly specified or limited in the embodiments, the terms "installation," "connection," "joining," and "fixing" appearing in the embodiments should be interpreted broadly. For example, a connection can be a fixed connection, a detachable connection, or an integral part; it can also be a mechanical connection, an electrical connection, etc. Of course, it can also be a direct connection, or an indirect connection through an intermediate medium, or it can be the internal communication of two components, or the interaction between two components. Those skilled in the art can understand the specific meaning of the above terms in this invention based on the specific implementation.

[0247] In this invention, unless otherwise explicitly specified and limited, "above" or "below" the second feature can mean that the first feature is in direct contact with the second feature, or that the first feature is in indirect contact with the second feature through an intermediate medium. Furthermore, "above," "over," and "on top" of the second feature can mean that the first feature is directly above or diagonally above the second feature, or simply that the first feature is at a higher horizontal level than the second feature. "Below," "below," and "under" the second feature can mean that the first feature is directly below or diagonally below the second feature, or simply that the first feature is at a lower horizontal level than the second feature.

[0248] Although embodiments of the present invention have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting the present invention. Those skilled in the art can make changes, modifications, substitutions and variations to the above embodiments within the scope of the present invention.

Claims

1. A data processing method, characterized in that, The method includes: Obtain the first projection data that has been truncated and padded, the first projection data including the original projection data; Perform at least one truncation correction procedure, the truncation correction procedure including: The first projection data is truncated and corrected to obtain the second projection data, and the original projection data is used to replace the data within the scanning field of view in the second projection data; The first image is determined based on the replaced second projection data, and the first image is truncated and corrected to obtain the second image; The third projection data is determined based on the second image, and the original projection data is used to replace the data within the scanned field of view in the third projection data; Update the first projection data with the replaced third projection data.

2. The method according to claim 1, characterized in that, The second projection data is obtained by inputting the first projection data into a projection data truncation correction network for truncation correction.

3. The method according to claim 2, characterized in that, The method further includes: Acquire multiple full-scan field-of-view images; For each of the full-scan field-of-view images, based on the sample sub-images in the full-scan field-of-view image, truncated first sample projection data is obtained, and the first sample projection data is extrapolated to obtain untruncated training projection data. Untruncated label projection data is obtained based on the full-scan field-of-view image; The projection data truncation correction network is trained using multiple training projection data and multiple label projection data. The sample sub-image includes a portion of the sample object in the full-scan field-of-view image, the image corresponding to the training projection data is an image without truncation but with artifacts, and the label image corresponding to the label projection data is an image without truncation and without artifacts.

4. The method according to claim 3, characterized in that, The projection data truncation correction network is trained using multiple training projection data and multiple label projection data, including: Multiple training projection data are input into a first initial network to obtain multiple corrected projection data output by the first initial network; For each of the corrected projection data, a first loss value is determined between the data outside the scan field of view in the corrected projection data and the data outside the scan field of view in the corresponding label projection data, and multiple first loss values ​​are obtained. The first initial network is adjusted based on the mean of multiple first loss values ​​to obtain the projection data truncation correction network.

5. The method according to claim 4, characterized in that, The method further includes: For each of the corrected projection data, a first data difference is determined between the data of every two adjacent extended channels outside the scanning field of view in the corrected projection data, a second data difference is determined between the data of every two adjacent extended channels outside the scanning field of view in the corresponding label projection data, and a second loss value is determined for the first data difference and the second data difference, resulting in a plurality of second loss values; The first initial network is adjusted based on the mean of multiple first loss values ​​to obtain the projection data truncation correction network, including: The first weighted loss value is obtained by weighting and summing the mean of multiple first loss values ​​and the mean of multiple second loss values; wherein the weight of the mean of multiple second loss values ​​is less than the weight of the mean of multiple first loss values. The first initial network is adjusted based on the first weighted loss value to obtain the projection data truncation correction network.

6. The method according to any one of claims 1 to 5, characterized in that, The first image is truncated and corrected to obtain the second image, which includes: The first image is input into the image truncation correction network to perform truncation correction on the first image, thereby obtaining the second image.

7. The method according to claim 6, characterized in that, The method further includes: Acquire multiple full-scan field-of-view images; For each of the full-scan field-of-view images, based on the sample sub-images in the full-scan field-of-view image, truncated first sample projection data is obtained, and the first sample projection data is extrapolated to obtain untruncated training projection data. Untruncated label projection data is obtained based on the full-scan field-of-view image; For each training projection data, the training projection data is truncated and corrected to obtain second sample projection data. The data in the scanned field of view in the second sample projection data is replaced with the first sample projection data to obtain third sample projection data. A training image without truncation and with artifacts is obtained based on the third sample projection data. For each of the label projection data, obtain a label image without truncation and without artifacts based on the label projection data; The image truncation correction network is trained using multiple training images and multiple label images.

8. The method according to claim 7, characterized in that, The image truncation correction network is trained using multiple training images and multiple label images, including: Multiple training images are input into a second initial network to obtain multiple corrected images output by the second initial network; For each of the corrected images, a third loss value is determined between the first data outside the scanned field of view in the corrected image and the second data outside the scanned field of view in the corresponding label image, resulting in multiple third loss values; The second initial network is adjusted based on the mean of multiple third loss values ​​to obtain the image truncation correction network.

9. The method according to claim 8, characterized in that, The method further includes: For each of the corrected images, a fourth loss value is determined between the data within the scanned field of view in the corrected image and the data within the scanned field of view in the corresponding label image, resulting in multiple fourth loss values; The second initial network is adjusted based on the mean of multiple third loss values ​​to obtain the image truncation correction network, including: The average of the multiple third loss values ​​and the average of the multiple fourth loss values ​​are weighted and summed to obtain a second weighted loss value, wherein the weight of the average of the multiple fourth loss values ​​is less than the weight of the average of the multiple third loss values. The second initial network is adjusted based on the second weighted loss value to obtain the image truncation correction network.

10. The method according to claim 9, characterized in that, The third loss value is determined by weighted summation of the average absolute value error loss value of the first data and the second data, and the structural similarity loss value of the first data and the second data.

11. The method according to claim 7, characterized in that, The training projection data is truncated and corrected to obtain the second sample projection data, including: The training projection data is input into the projection data truncation correction network to obtain the second sample projection data.

12. The method according to any one of claims 1 to 11, characterized in that, Before updating the first projection data to the replaced third projection data, the method further includes: The data outside the scanning field of view in the replaced third projection data are subjected to smooth transition processing.

13. A computer device, characterized in that, It includes a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the computer program, it implements the data processing method according to any one of claims 1 to 12.

14. A data processing apparatus, characterized in that, The device includes: The acquisition module is used to acquire the first projection data that has been truncated and padded, the first projection data including the original projection data; The correction module is configured to: execute at least one truncation correction process, the truncation correction process including: The first projection data is truncated and corrected to obtain the second projection data, and the original projection data is used to replace the data within the scanning field of view in the second projection data; The first image is determined based on the replaced second projection data, and the first image is truncated and corrected to obtain the second image; The third projection data is determined based on the second image, and the original projection data is used to replace the data within the scanned field of view in the third projection data; Update the first projection data with the replaced third projection data.

Images