Method for obtaining reconstruction data, electronic equipment and CT (Computed Tomography) machine
By constructing a pre-trained convolutional neural network model, the truncation problem in CT scan data is solved, and the image resolution exceeding the scanning field of view is improved.
Patent Information
- Application Number
- CN202510106637.7
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-01-23
- Publication Date
- 2025-05-23
AI Technical Summary
In CT scanning imaging, data truncation occurs when the object exceeds the scanning field of view, resulting in truncation artifacts of the reconstruction image, affecting diagnosis, and at the same time, the resolution of the extended reconstruction area beyond the scanning field of view is low.
By obtaining the actual reconstruction data and determining whether there is a truncation, if so, input these data into the pre-trained convolutional neural network model to obtain the standard reconstruction data. The pre-trained convolutional neural network model is constructed through multiple sets of first and second reconstruction data under different samples and scanning parameters, and can process truncated and non-truncated data under different samples information and scanning parameters.
Automatic completion of data is achieved, images without truncated artifacts are reconstructed, and resolution of extended reconstruction beyond the scanning field of view is improved.
Smart Images
Figure CN120031998A_ABST
Abstract
Description
Technical Field
[0001] The present disclosure relates to the technical field of medical devices, and in particular to a method for obtaining reconstruction data, an electronic device and a CT machine. Background Art
[0002] CT (Computed Tomography) uses an X-ray source and a highly sensitive X-ray detector to scan around the object to convert the X-rays that pass through the object into digital signals and reconstruct a three-dimensional tomographic image of the object based on a reconstruction algorithm. CT imaging is a non-destructive imaging method that can be used to inspect and test human bodies, animals, and industrial samples.
[0003] In CT scanning imaging, there is a concept of scanning field of view (usually a circle with a diameter of 500mm, with the center of rotation as the center of the circle and half of the scanning field of view as the radius), which means that only the data of objects within this field of view after X-ray exposure will be received by the detector and can be used to reconstruct the image. When the object exceeds this field of view, the collected data is usually called data "truncation", that is, the edge channel collects object data instead of air. Direct reconstruction of this "truncated" data will produce CT images with "truncation" artifacts (the edge of the image beyond the scanning field will be bright), affecting diagnosis.
[0004] At the same time, CT imaging also has a reconstruction function of "extending" the field of view. Usually, CT reconstruction can only reconstruct images that are less than or equal to the scanning field of view (because only the scanning field of view has data, and there is no data beyond the scanning field of view), but the data range of the current scanning field of view can be extended to both sides through "interpolation", so that a larger range of CT images can be reconstructed. However, the image resolution of this reconstructed image in the area beyond the scanning field of view is much lower than that in the scanning field of view, and it cannot be fully used for diagnosis. Summary of the invention
[0005] The purpose of the present disclosure is to provide a method for obtaining reconstruction data, an electronic device and a CT machine, which can solve at least one of the above-mentioned technical problems. The specific solution is as follows:
[0006] According to a specific embodiment of the present disclosure, on the one hand, the present disclosure provides a method for obtaining reconstruction data, which is used for CT scanning, and the method for obtaining reconstruction data includes: obtaining actual reconstruction data; judging whether the actual reconstruction data is truncated; if the actual reconstruction data is truncated, inputting the reconstruction data with truncation into a pre-trained convolutional neural network model to obtain standard reconstruction data; wherein the pre-trained convolutional neural network model includes: scanning a sample with a first preset number of channels and preset scanning parameters to obtain first reconstruction data; the first preset field of view is a scanning field of view covering the entire area to be scanned; scanning the same sample with a second preset number of channels and the same preset scanning parameters to obtain second reconstruction data; the second preset field of view is a scanning field of view covering part of the area to be scanned; scanning multiple different samples and corresponding preset scanning parameters to obtain multiple groups of the first reconstruction data and the second reconstruction data; constructing a pre-trained convolutional neural network model based on the multiple groups of the first reconstruction data and the second reconstruction data.
[0007] In an optional embodiment, the sample is scanned using the number of channels of a first preset field of view and preset scanning parameters to obtain first reconstructed data, including: determining the size of the first preset field of view based on the shape and size of the sample; determining the number of channels based on the size of the first preset field of view to ensure that the scanning field of view using this number of channels covers the entire area to be scanned.
[0008] In an optional embodiment, scanning the same sample using the second preset number of channels of the field of view and the preset scanning parameters to obtain second reconstructed data includes: manually reducing the number of channels to ensure that the scanning field of view using the number of channels covers part of the area to be scanned; scanning the same sample using the same preset scanning parameters to obtain second reconstructed data.
[0009] In an optional embodiment, manually reducing the number of channels includes: simultaneously reducing the number of channels of a preset number at both ends of the detector.
[0010] In an optional embodiment, the preset number is 2-10.
[0011] In an optional embodiment, scanning the same sample using the second preset number of channels of the field of view and the same preset scanning parameters to obtain second reconstruction data also includes: scanning the same sample using multiple different numbers of channels of the second preset field of view and the same preset scanning parameters to obtain multiple second reconstruction data.
[0012] In an optional embodiment, the scanning of multiple different samples and the corresponding preset scanning parameters to obtain multiple sets of the first reconstruction data and the second reconstruction data includes: each second reconstruction data is matched with the corresponding first reconstruction data, to obtain multiple sets of the first reconstruction data and the second reconstruction data.
[0013] In an optional embodiment, the scanning of multiple different samples and the corresponding preset scanning parameters to obtain multiple sets of the first reconstruction data and the second reconstruction data includes: one second reconstruction data matches a corresponding first reconstruction data to obtain multiple sets of the first reconstruction data and the second reconstruction data.
[0014] According to a specific embodiment of the present disclosure, on the other hand, the present disclosure provides a unit for obtaining reconstruction data, which is used for CT scanning and is configured to obtain reconstruction data by executing a method as described in any one of the above technical solutions.
[0015] According to a specific implementation of the present disclosure, on another aspect, the present disclosure provides a CT machine, the CT machine comprising: a unit for obtaining reconstruction data as described in the above technical solution.
[0016] According to a specific implementation of the present disclosure, on the other hand, the present disclosure provides a computer-readable storage medium having a computer program stored thereon, wherein the program, when executed by a processor, implements a method as described in any one of the above technical solutions.
[0017] According to a specific implementation of the present disclosure, on another aspect, the present disclosure provides an electronic device, comprising: one or more processors; a storage device for storing one or more programs, and when the one or more programs are executed by the one or more processors, the one or more processors implement a method as described in any one of the above technical solutions.
[0018] Compared with the prior art, the above solution of the embodiment of the present disclosure has at least the following beneficial effects:
[0019] The present invention obtains multiple sets of first reconstruction data and second reconstruction data by artificially reducing the detector channels; then constructs a pre-trained convolutional neural network model with the multiple sets of first reconstruction data and second reconstruction data to obtain the relationship between reconstruction data with and without truncation under different sample information and corresponding scanning parameters. During the actual scanning, if there is truncation, the reconstruction data with truncation is input into the pre-trained convolutional neural network model to obtain standard reconstruction data. The method disclosed in the present invention can achieve automatic data completion, reconstruct an image without truncation artifacts, and improve the resolution of the extended reconstruction beyond the scanning field of view. BRIEF DESCRIPTION OF THE DRAWINGS
[0020] Figure 1 A flow chart of a method for obtaining reconstruction data according to an embodiment of the present disclosure is shown.
[0021] Figure 2 A schematic diagram of scanning a sample by adopting a method for obtaining reconstructed data according to an embodiment of the present disclosure is shown, wherein the circle surrounded by dotted lines is the first preset field of view, and the ellipse surrounded by solid lines is the sample.
[0022] Figure 3 A schematic diagram of scanning a sample by adopting a method for obtaining reconstructed data according to another embodiment of the present disclosure is shown, wherein the circle surrounded by dotted lines is the second preset field of view, and the ellipse surrounded by solid lines is the sample.
[0023] Figure 4 A flowchart of a pre-trained convolutional neural network model according to an embodiment of the present disclosure is shown.
[0024] Figure 5 A schematic diagram of a connection structure of an electronic device according to an embodiment of the present disclosure is shown.
[0025] Reference numerals:
[0026] 100: tube; 200: first preset field of view; 300: second preset field of view; 400: sample; 500: detector;
[0027] 301: processing system; 302: ROM; 303: RAM; 304: bus; 305: I / O interface; 306: input system; 307: output system; 308: storage system; 309: communication system. DETAILED DESCRIPTION
[0028] In order to make the purpose, technical solutions and advantages of the present disclosure clearer, the present disclosure will be further described in detail below in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present disclosure, rather than all of the embodiments. Based on the embodiments in the present disclosure, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present disclosure.
[0029] The terms used in the embodiments of the present disclosure are only for the purpose of describing specific embodiments, and are not intended to limit the present disclosure. The singular forms "a", "said" and "the" used in the embodiments of the present disclosure and the appended claims are also intended to include plural forms, unless the context clearly indicates other meanings, and "multiple" generally includes at least two.
[0030] It should be understood that the term "and / or" used in this article is only a description of the association relationship of associated objects, indicating that there can be three relationships. For example, A and / or B can represent: A exists alone, A and B exist at the same time, and B exists alone. In addition, the character " / " in this article generally indicates that the associated objects before and after are in an "or" relationship.
[0031] It should be understood that although the terms first, second, third, etc. may be used to describe structures in the disclosed embodiments, these structures should not be limited to these terms. These terms are only used to distinguish different structures. For example, without departing from the scope of the disclosed embodiments, a first component may also be referred to as a second component, and similarly, a second component may also be referred to as a first component.
[0032] As used herein, the words "if" and "if" may be interpreted as "at the time of" or "when" or "in response to determining" or "in response to detecting", depending on the context. Similarly, the phrases "if it is determined" or "if (stated condition or event) is detected" may be interpreted as "when it is determined" or "in response to determining" or "when detecting (stated condition or event)" or "in response to detecting (stated condition or event)", depending on the context.
[0033] It should also be noted that the term "includes", "comprising" or any other variation thereof is intended to cover non-exclusive inclusion, so that a commodity or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such commodity or device. In the absence of more restrictions, the elements defined by the sentence "comprising a ..." do not exclude the existence of other identical elements in the commodity or device including the elements.
[0034] In the related technology, the human body is positioned as far as possible within the scanning field of view to avoid truncation. However, in some cases, such as when the patient is too large to be within the scanning field of view, or when the scanning field of view of the CT machine itself is small (to save costs or for head CT, the scanning field of view is designed to be very small), it is inevitable that the object will be outside the scanning field of view. For the second case mentioned above, CT manufacturers use pure mathematical methods, that is, to expand the channel data outside the scanning field of view through interpolation. These channels do not exist in themselves and are generated through interpolation. In this way, non-truncated data is obtained, and the reconstructed image does not have artifacts. There is currently no good method to solve the problem of low resolution of the reconstructed image generated by interpolation beyond the scanning field of view.
[0035] The data interpolation processing method used to correct truncation artifacts is to interpolate the data without truncation and beyond the scanning field of view before the convolution back-projection algorithm in the reconstruction process, so that the reconstruction field of the final generated image is different. For example, the scanning field of view of a CT machine is 500mm. When reconstructing the field of view of 500mm, the data is truncated. Through truncation correction, the bright arc-shaped artifacts at the edge of the scanning field of view disappear. At this time, the 700mm image can also be reconstructed. Not only does the bright arc-shaped artifact disappear at 500mm, but there are also images outside the scanning field of view. The above method is obtained through mathematical interpolation. The image resolution will be reduced at the edge of the image, and accurate diagnosis cannot be made.
[0036] In order to solve at least one of the technical problems mentioned above, the present disclosure provides a method for obtaining reconstruction data, an electronic device and a CT machine; the method for obtaining reconstruction data is used for CT scanning, and the method for obtaining reconstruction data includes: obtaining actual reconstruction data; judging whether the actual reconstruction data is truncated; if the actual reconstruction data is truncated, inputting the reconstruction data with truncation into a pre-trained convolutional neural network model to obtain standard reconstruction data; wherein the pre-trained convolutional neural network model includes: scanning the sample with a first preset field of view 200 of channels and preset scanning parameters to obtain first reconstruction data; the first preset field of view 200 is a scanning field of view covering the entire area to be scanned; scanning the same sample with a second preset field of view 300 of channels and the same preset scanning parameters to obtain second reconstruction data; the second preset field of view 300 is a scanning field of view covering part of the area to be scanned; scanning multiple different samples and corresponding preset scanning parameters to obtain multiple groups of the first reconstruction data and the second reconstruction data; constructing a pre-trained convolutional neural network model based on the multiple groups of the first reconstruction data and the second reconstruction data. The present invention obtains multiple sets of first reconstruction data and second reconstruction data by artificially reducing the detector channels; then constructs a pre-trained convolutional neural network model with the multiple sets of first reconstruction data and second reconstruction data to obtain the relationship between reconstruction data with and without truncation under different sample 400 information and corresponding scanning parameters. During the actual scanning, if there is truncation, the reconstruction data with truncation is input into the pre-trained convolutional neural network model to obtain standard reconstruction data. The method disclosed in the present invention can achieve automatic data completion, reconstruct an image without truncation artifacts, and improve the resolution of the extended reconstruction beyond the scanning field of view.
[0037] The optional embodiments of the present disclosure are described in detail below with reference to the accompanying drawings.
[0038] Figure 1 A flow chart of a method for obtaining reconstruction data according to an embodiment of the present disclosure is shown. Figure 2A schematic diagram of scanning a sample 400 by adopting a method for obtaining reconstructed data according to an embodiment of the present disclosure is shown, wherein the circle surrounded by dotted lines is the first preset field of view 200 , and the ellipse surrounded by solid lines is the sample 400 . Figure 3 FIG. 4 is a schematic diagram showing a method for obtaining reconstruction data according to another embodiment of the present disclosure for scanning a sample 400, wherein the circle surrounded by dotted lines is the second preset field of view 300, and the ellipse surrounded by solid lines is the sample 400. Figure 1 , Figure 2 and Figure 3 As shown, according to a specific embodiment of the present disclosure, on the one hand, a method for obtaining reconstruction data is provided, the method for obtaining reconstruction data is used for CT scanning, and the method for obtaining reconstruction data at least comprises the following steps:
[0039] S100: Acquire actual reconstruction data.
[0040] S200: Determine whether the actual reconstructed data is truncated.
[0041] S300: If the actual reconstructed data is truncated, the reconstructed data with truncation is input into a pre-trained convolutional neural network model to obtain standard reconstructed data.
[0042] Among them, in step S100 and step S200, when the CT machine is actually used for scanning, it is determined whether the actual reconstructed data is truncate after scanning. If truncation is present, the reconstructed data with truncation is input into the pre-trained convolutional neural network model to obtain standard reconstructed data.
[0043] Among them, in step S300, through the method disclosed in the present invention, the reconstructed data with truncation can be input into the pre-trained convolutional neural network model to obtain standard reconstructed data. This correction method is used to remove truncation artifacts on the one hand, and on the other hand, the resolution of the image obtained by the expanded reconstruction is not reduced.
[0044] Figure 4 FIG. 4 shows a flow chart of a pre-trained convolutional neural network model according to an embodiment of the present disclosure. Figure 4 As shown, the pre-trained convolutional neural network model includes:
[0045] S400, scanning the sample using the number of channels of a first preset field of view 200 and preset scanning parameters to obtain first reconstructed data; the first preset field of view 200 is a scanning field of view covering the entire area to be scanned.
[0046] S500, scanning the same sample using the number of channels of a second preset field of view 300 and the same preset scanning parameters to obtain second reconstructed data; the second preset field of view 300 is a scanning field of view covering part of the area to be scanned.
[0047] S600, scanning a plurality of different samples and the corresponding preset scanning parameters to obtain a plurality of sets of the first reconstruction data and the second reconstruction data.
[0048] S700: construct a pre-trained convolutional neural network model based on multiple groups of the first reconstruction data and the second reconstruction data.
[0049] Among them, the first preset field of view 200 is a scanning field of view covering the entire area to be scanned, and the second preset field of view 300 is a scanning field of view covering part of the area to be scanned. The present disclosure obtains multiple sets of first reconstruction data and second reconstruction data by artificially reducing the detector channels; then constructs a pre-trained convolutional neural network model with the multiple sets of first reconstruction data and second reconstruction data to obtain the relationship between reconstruction data with and without truncation under different sample 400 information and corresponding scanning parameters. During the actual scanning, if there is truncation, the reconstruction data with truncation is input into the pre-trained convolutional neural network model to obtain standard reconstruction data. The method disclosed in the present disclosure can achieve automatic data completion, reconstruct an image without truncation artifacts, and improve the resolution of the extended reconstruction beyond the scanning field of view.
[0050] In step S400 and step S500, the target is scanned to obtain first reconstruction data under the premise that the scanning field of view covers the entire area to be scanned. Then the scanning field of view is adjusted to cover part of the area to be scanned, and the same target is scanned using the same scanning parameters to obtain second reconstruction data. At this time, the difference between the first reconstruction data and the second reconstruction data is whether there is truncation - the sizes of the two are different.
[0051] Wherein, in step S600, different samples 400 are scanned in the manner of the above-mentioned steps S400 and S500 to obtain multiple sets of matching first reconstruction data and second reconstruction data. In actual use, the first reconstruction data and the second reconstruction data obtained by scanning the same sample 400 twice are a matching set. It should be noted that the preset scanning parameters are adjusted according to the shape and size of the sample 400 so that a clear image can be obtained in the scanning field of view.
[0052] Wherein, in step S700, a pre-trained convolutional neural network model is constructed and trained using multiple groups of the first reconstruction data and the second reconstruction data. The pre-trained convolutional neural network model can be constructed based on structures such as UNet and Resnet, and may include multiple convolutional layers, multiple pooling layers, etc., and may perform operations such as feature extraction, downsampling, upsampling, and batch normalization. The trained pre-trained convolutional neural network model can replace the interpolation method to complete the data, but it does not lose image resolution like the interpolation method. After the pre-trained convolutional neural network model is trained, it can be applied to CT machines with smaller scanning fields of view with the same geometric parameters in the future for truncation correction processing. For example, it can be used on machines with a scanning field of view of 400mm, or even smaller, which also brings benefits to reducing the cost of CT.
[0053] The step S400 includes the following steps:
[0054] S410: Determine the size of the first preset field of view 200 based on the shape and size of the sample 400
[0055] S420: Determine the number of channels based on the size of the first preset field of view 200 to ensure that the scanning field of view using the number of channels covers the entire area to be scanned.
[0056] The step S500 includes the following steps:
[0057] S510, manually reducing the number of channels to ensure that the scanning field of view using the number of channels covers part of the area to be scanned;
[0058] S520, scanning the same sample 400 using the same preset scanning parameters to obtain second reconstructed data.
[0059] The step S500 further includes the following steps:
[0060] S510`, scanning the same sample 400 using a plurality of different numbers of channels of the second preset field of view 300 and the same preset scanning parameters to obtain a plurality of second reconstruction data.
[0061] S520` Each second reconstruction data is matched with the corresponding first reconstruction data to obtain multiple groups of the first reconstruction data and the second reconstruction data;
[0062] The step S500 further includes the following steps:
[0063] S510``, one second reconstruction data is matched with a corresponding one first reconstruction data, to obtain multiple groups of the first reconstruction data and the second reconstruction data.
[0064] Specifically, the pre-trained convolutional neural network model is:
[0065] (1) Data collection: Use CT machines with different scanning fields of view (e.g., 700 mm, 400 mm) to scan phantoms and clinical human data. The scans include eccentric scans and non-eccentric scans.
[0066] (2) Data preprocessing: When the scanning field of view is 700 mm, the original 960-channel data is divided into 800 and 160-channel data; the 800-channel data is normalized; the 160-channel data is used as a label; the data of other different scanning field sizes are processed in the same way;
[0067] (3) Data enhancement: Apply transformations such as cropping and scaling to increase the diversity of data so that the model can learn image features at different scales.
[0068] (4) Data segmentation: Divide the dataset into training set, validation set and test set, with the ratios of 70%, 15% and 15% respectively.
[0069] The U-Net architecture is used to achieve in-depth extraction and accurate reconstruction of image features through its unique encoder-decoder structure. In addition, the model has particularly enhanced its adaptability to multi-scale fields of view, so that CT images of both large and small fields of view can be effectively corrected and reconstructed through the network, thereby improving the applicability and economy of CT scanning equipment without sacrificing image quality.
[0070] Input layer: accepts CT image slices of any size, denoted as I, which is the second reconstructed data.
[0071] Convolutional layer: The network consists of multiple convolutional blocks in the encoder part, each of which contains two convolutional layers (Conv2D), followed by a ReLU activation function and a pooling layer (MaxPooling). Specifically, for the lth convolutional block, its operation can be expressed as:
[0072] F l =ReLU(Conv2D(ReLU(Conv2D(I l-1 ))))
[0073] I l =MaxPooling(F l )
[0074] Among them, I 0 =I is the input image, that is, the second reconstruction data, I l is the feature map after the lth convolutional block, F lis the intermediate feature map of the convolutional block. Through this stacking method, the network effectively extracts key texture and structural information from the input CT image, and reduces the spatial dimension of the feature through the pooling layer, providing a higher level of abstract feature representation for subsequent network layers, thereby maintaining image details while achieving generalization capabilities for different scanning field sizes.
[0075] Decoder: The decoder part restores the details and resolution of the image through upsampling and skip connections of the network. Each upsampling layer is followed by two convolutional layers to refine the features and reduce the aliasing effect caused by upsampling. The skip connection concatenates the feature map of the corresponding layer of the encoder with the feature map of the decoder after upsampling to preserve spatial information. Specifically, for the lth decoding block, its operation can be expressed as:
[0076] U l =Conv2D(ReLU(Conv2D(Concatenate(I L-1 ,U l+1 ))))
[0077] Here, L is the number of convolutional blocks in the encoder, and U l is the feature map after the l-th decoding block, I L-1 is the corresponding encoder feature map. In this way, the decoder gradually restores the high-resolution details of the image, ensuring the accuracy and quality of the reconstructed image. Finally, the output of the decoder is a reconstructed image of the same size as the original image, denoted as That is the first reconstruction data.
[0078] The training process is:
[0079] Initialization: Before starting training, the network weights are randomly initialized to ensure that the network can learn from scratch.
[0080] Forward propagation: For each training sample, pass the input image I through the network to get the predicted image This process involves encoder feature extraction and decoder feature reconstruction.
[0081] Loss function design: In order to ensure the quality of the reconstructed image, we use a combination of structural similarity index (SSIM) and mean square error (MSE) as the loss function. This combination can balance the image structure and pixel-level error. The formula is as follows:
[0082]
[0083] Here, α is the balance coefficient, which is usually set around 0.5 to ensure that the two losses are equally important.
[0084] Backpropagation: Calculate the gradient of the loss function Τ with respect to the network parameters and update the weights of the network using a gradient descent algorithm (such as the Adam optimizer).
[0085] Iterative training: Repeat steps 2 to 4 until the model loss on the training set no longer decreases significantly, or the predetermined number of iterations is reached.
[0086] Verification and tuning: Evaluate the performance of the model on the validation set, and adjust hyperparameters such as the learning rate and the balance coefficient α in the loss function based on the verification results.
[0087] Save and test: When the model achieves satisfactory performance on the validation set, save the model parameters. Finally, evaluate the generalization ability of the model on an independent test set to ensure that the model can handle unseen data.
[0088] Through the above training process, we are able to train a robust U-Net model, which can effectively process CT images with different scanning field sizes and achieve high-quality image reconstruction.
[0089] In some embodiments, the method for obtaining reconstruction data further includes: scanning a patient to obtain actual reconstruction data; determining whether the actual reconstruction data has truncation; and if the actual reconstruction data has truncation, obtaining the reconstruction data with truncation.
[0090] In some embodiments, the sample 400 is scanned using the number of channels of the first preset field of view 200 and preset scanning parameters to obtain first reconstructed data, including: determining the size of the first preset field of view 200 based on the shape and size of the sample 400; determining the number of channels based on the size of the first preset field of view 200 to ensure that the scanning field of view using this number of channels covers the entire area to be scanned.
[0091] In some embodiments, the method of scanning the same sample 400 using the number of channels of the second preset field of view 300 and the preset scanning parameters to obtain second reconstructed data includes: manually reducing the number of channels to ensure that the scanning field of view using the number of channels covers part of the area to be scanned; and scanning the same sample 400 using the same preset scanning parameters to obtain second reconstructed data.
[0092] In some embodiments, manually reducing the number of channels includes: simultaneously reducing the number of channels of a preset number at both ends of the detector 500. In an optional embodiment, the preset number is 2-10. Figure 2 and Figure 3 During scanning, penetrating rays are emitted through the tube 100 and are received by the detector 500 after passing through the scanning area.
[0093] In some embodiments, scanning the same sample 400 with the same number of channels of the second preset field of view 300 and the same preset scanning parameters to obtain second reconstruction data also includes: scanning the same sample 400 with multiple different numbers of channels of the second preset field of view 300 and the same preset scanning parameters to obtain multiple second reconstruction data.
[0094] In some embodiments, the scanning of multiple different samples 400 and the corresponding preset scanning parameters to obtain multiple sets of the first reconstruction data and the second reconstruction data includes: each second reconstruction data is matched with the corresponding first reconstruction data to obtain multiple sets of the first reconstruction data and the second reconstruction data.
[0095] In some embodiments, the scanning of multiple different samples 400 and the corresponding preset scanning parameters to obtain multiple sets of the first reconstruction data and the second reconstruction data includes: one second reconstruction data matches a corresponding first reconstruction data to obtain multiple sets of the first reconstruction data and the second reconstruction data.
[0096] According to a specific embodiment of the present disclosure, on the other hand, a unit for obtaining reconstruction data is provided, the unit for obtaining reconstruction data is used for CT scanning, and the unit for obtaining reconstruction data is configured to obtain reconstruction data by executing a method as described in any one of the above embodiments.
[0097] According to a specific embodiment of the present disclosure, in another aspect, a CT machine is provided. The CT machine is used for CT scanning, and the CT machine includes: a unit for obtaining reconstruction data as described in the above embodiment.
[0098] According to a specific embodiment of the present disclosure, on another aspect, an electronic device is provided, the device is used for obtaining a method for reconstructing data, the electronic device comprises: at least one processor; and a memory connected to the at least one processor in communication; wherein,
[0099] The memory stores instructions that can be executed by the one processor, and the instructions are executed by the at least one processor so that the at least one processor can: obtain actual reconstruction data; determine whether the actual reconstruction data is truncated; if the actual reconstruction data is truncated, input the reconstruction data with truncation into the pre-trained convolutional neural network model to obtain standard reconstruction data; wherein the pre-trained convolutional neural network model includes: scanning the sample with a first preset field of view 200 of channels and preset scanning parameters to obtain first reconstruction data; the first preset field of view 200 is a scanning field of view covering the entire area to be scanned; scanning the same sample with a second preset field of view 300 of channels and the same preset scanning parameters to obtain second reconstruction data; the second preset field of view 300 is a scanning field of view covering part of the area to be scanned; scanning multiple different samples and corresponding preset scanning parameters to obtain multiple groups of the first reconstruction data and the second reconstruction data; constructing a pre-trained convolutional neural network model based on multiple groups of the first reconstruction data and the second reconstruction data.
[0100] An embodiment of the present disclosure provides a non-volatile computer storage medium, wherein the computer storage medium stores computer executable instructions, and the computer executable instructions can execute the method for obtaining reconstruction data in any of the above method embodiments.
[0101] Reference below Figure 5 , which shows a schematic diagram of the structure of an electronic device suitable for implementing the embodiments of the present disclosure. The terminal device in the embodiments of the present disclosure may include but is not limited to mobile terminals such as mobile phones, laptop computers, digital broadcast receivers, PDAs (personal digital assistants), PADs (tablet computers), PMPs (portable multimedia players), vehicle-mounted terminals (such as vehicle-mounted navigation terminals), etc., and fixed terminals such as digital TVs, desktop computers, etc. Figure 5 The electronic device shown is only an example and should not bring any limitation to the functions and scope of use of the embodiments of the present disclosure.
[0102] like Figure 5 As shown, the electronic device may include a processing system (e.g., a central processing unit, a graphics processor, etc.) 301, which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) 302 or a program loaded from a storage system 308 into a random access memory (RAM) 303. In RAM 303, various programs and data required for the operation of the electronic device are also stored. The processing system 301, ROM 302, and RAM 303 are connected to each other via a bus 304. An input / output (I / O) interface 305 is also connected to the bus 304.
[0103] Typically, the following systems may be connected to the I / O interface 305: an input system 306 including, for example, a touch screen, a touch pad, a keyboard, a mouse, a camera, a microphone, an accelerometer, a gyroscope, etc.; an output system 307 including, for example, a liquid crystal display (LCD), a speaker, a vibrator, etc.; a storage system 308 including, for example, a magnetic tape, a hard disk, etc.; and a communication system 309. The communication system 309 may allow the electronic device to communicate with other devices wirelessly or by wire to exchange data. Although Figure 5 An electronic device with various systems is shown, but it should be understood that it is not required to implement or have all of the systems shown. More or fewer systems may be implemented or have instead.
[0104] In particular, according to an embodiment of the present disclosure, the process described above with reference to the flowchart can be implemented as a computer software program. For example, an embodiment of the present disclosure includes a computer program product, which includes a computer program carried on a computer-readable medium, and the computer program includes a program code for executing the method shown in the flowchart. In such an embodiment, the computer program can be downloaded and installed from the network through the communication system 309, or installed from the storage system 308, or installed from the ROM 302. When the computer program is executed by the processing system 301, the above-mentioned functions defined in the method of the embodiment of the present disclosure are executed.
[0105] It should be noted that the computer-readable medium mentioned above in the present disclosure may be a computer-readable signal medium or a computer-readable storage medium or any combination of the above two. The computer-readable storage medium may be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, system or device, or any combination of the above. More specific examples of computer-readable storage media may include, but are not limited to: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In the present disclosure, a computer-readable storage medium may be any tangible medium containing or storing a program that can be used by or in combination with an instruction execution system, system or device. In the present disclosure, a computer-readable signal medium may include a data signal propagated in a baseband or as part of a carrier wave, in which a computer-readable program code is carried. This propagated data signal may take a variety of forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the above. The computer readable signal medium may also be any computer readable medium other than a computer readable storage medium, which may send, propagate or transmit a program for use by or in conjunction with an instruction execution system, system or device. The program code contained on the computer readable medium may be transmitted using any suitable medium, including but not limited to: wires, optical cables, RF (radio frequency), etc., or any suitable combination of the above.
[0106] The computer-readable medium may be included in the electronic device, or may exist independently without being incorporated into the electronic device.
[0107] The computer-readable medium carries one or more programs. When the one or more programs are executed by the electronic device, the electronic device: The present disclosure obtains multiple sets of first reconstruction data and second reconstruction data by artificially reducing the detector channels; then constructs a pre-trained convolutional neural network model with the multiple sets of first reconstruction data and second reconstruction data to obtain the relationship between reconstruction data with and without truncation under different sample 400 information and corresponding scanning parameters. During the actual scanning, if there is truncation, the reconstruction data with truncation is input into the pre-trained convolutional neural network model to obtain standard reconstruction data. The method disclosed in the present disclosure can achieve automatic data completion, reconstruct an image without truncation artifacts, and improve the resolution of the extended reconstruction beyond the scanning field of view.
[0108] Alternatively, the computer-readable medium carries one or more programs, and when the one or more programs are executed by the electronic device, the electronic device: The present disclosure obtains multiple sets of first reconstruction data and second reconstruction data by artificially reducing the detector channels; then the multiple sets of first reconstruction data and second reconstruction data are used to construct a pre-trained convolutional neural network model to obtain the relationship between reconstruction data with and without truncation under different sample 400 information and corresponding scanning parameters. During the actual scanning, if there is truncation, the reconstruction data with truncation is input into the pre-trained convolutional neural network model to obtain standard reconstruction data. The method disclosed in the present disclosure can achieve automatic data completion, reconstruct an image without truncation artifacts, and improve the resolution of the extended reconstruction beyond the scanning field of view.
[0109] Computer program code for performing the operations of the present disclosure may be written in one or more programming languages, or a combination thereof, including object-oriented programming languages, such as Java, Smalltalk, C++, and conventional procedural programming languages, such as "C" or similar programming languages. The program code may be executed entirely on the user's computer, partially on the user's computer, as a separate software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving a remote computer, the remote computer may be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computer (e.g., through the Internet using an Internet service provider).
[0110] The flow chart and block diagram in the accompanying drawings illustrate the possible architecture, function and operation of the system, method and computer program product according to various embodiments of the present disclosure. In this regard, each square box in the flow chart or block diagram can represent a module, a program segment or a part of a code, and the module, the program segment or a part of the code contains one or more executable instructions for realizing the specified logical function. It should also be noted that in some implementations as replacements, the functions marked in the square box can also occur in a sequence different from that marked in the accompanying drawings. For example, two square boxes represented in succession can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each square box in the block diagram and / or flow chart, and the combination of the square boxes in the block diagram and / or flow chart can be implemented with a dedicated hardware-based system that performs a specified function or operation, or can be implemented with a combination of dedicated hardware and computer instructions.
[0111] According to a specific embodiment of the present disclosure, on another aspect, a CT machine is provided, and the CT machine may include: a computer-readable storage medium as described in any one of the above embodiments.
[0112] According to a specific embodiment of the present disclosure, on another aspect, a CT machine is provided, and the CT machine may include: the electronic device as described in any one of the above embodiments.
[0113] The present disclosure aims to protect a method, electronic device and CT machine for obtaining reconstruction data; obtaining actual reconstruction data; determining whether the actual reconstruction data is truncated; if the actual reconstruction data is truncated, inputting the reconstruction data with truncation into a pre-trained convolutional neural network model to obtain standard reconstruction data; wherein the pre-trained convolutional neural network model includes: scanning a sample with a first preset field of view 200 of channels and preset scanning parameters to obtain first reconstruction data; the first preset field of view 200 is a scanning field of view covering the entire area to be scanned; scanning the same sample with a second preset field of view 300 of channels and the same preset scanning parameters to obtain second reconstruction data; the second preset field of view 300 is a scanning field of view covering part of the area to be scanned; scanning a plurality of different samples and the corresponding preset scanning parameters to obtain a plurality of groups of the first reconstruction data and the second reconstruction data; constructing a pre-trained convolutional neural network model based on the plurality of groups of the first reconstruction data and the second reconstruction data. The present invention obtains multiple sets of first reconstruction data and second reconstruction data by artificially reducing the detector channels; then constructs a pre-trained convolutional neural network model with the multiple sets of first reconstruction data and second reconstruction data to obtain the relationship between reconstruction data with and without truncation under different sample 400 information and corresponding scanning parameters. During the actual scanning, if there is truncation, the reconstruction data with truncation is input into the pre-trained convolutional neural network model to obtain standard reconstruction data. The method disclosed in the present invention can achieve automatic data completion, reconstruct an image without truncation artifacts, and improve the resolution of the extended reconstruction beyond the scanning field of view.
[0114] Finally, it should be noted that the various embodiments in this specification are described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the various embodiments can be referred to each other. For the system or device disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and the relevant parts can be referred to the method part description.
[0115] The above embodiments are only used to illustrate the technical solutions of the present disclosure, rather than to limit them. Although the present disclosure has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present disclosure.
Claims
1. A method for obtaining reconstruction data for CT scanning, characterized in that: include: Obtaining actual reconstruction data; Determining whether the actual reconstruction data is truncated; If the actual reconstructed data is truncated, the reconstructed data with truncation is input into a pre-trained convolutional neural network model to obtain standard reconstructed data; Wherein, the pre-trained convolutional neural network model includes: Scanning the sample using the number of channels and preset scanning parameters of a first preset field of view to obtain first reconstructed data; the first preset field of view is a scanning field of view covering the entire area to be scanned; Scanning the same sample using the number of channels of a second preset field of view and the same preset scanning parameters to obtain second reconstructed data; the second preset field of view is a scanning field of view covering part of the area to be scanned; Scanning a plurality of different samples and the corresponding preset scanning parameters to obtain a plurality of sets of the first reconstruction data and the second reconstruction data; A pre-trained convolutional neural network model is constructed based on multiple sets of the first reconstruction data and the second reconstruction data.
2. The method for obtaining reconstruction data according to claim 1, characterized in that: Scanning the sample using the number of channels of the first preset field of view and the preset scanning parameters to obtain the first reconstruction data includes: Determining the size of the first preset field of view based on the shape and size of the sample; The number of channels is determined based on the size of the first preset field of view to ensure that the scanning field of view using the number of channels covers the entire area to be scanned.
3. The method for obtaining reconstruction data according to claim 2, characterized in that: Scanning the same sample using the number of channels of the second preset field of view and the preset scanning parameters to obtain second reconstruction data includes: Manually reduce the number of channels to ensure that the scanning field of view using this number of channels covers part of the area to be scanned; The same sample is scanned using the same preset scanning parameters to obtain second reconstructed data.
4. The method for obtaining reconstruction data according to claim 3, characterized in that: The manual reduction of the number of channels includes: At the same time, the number of channels at both ends of the detector is reduced to a preset number; The preset number is 2-10.
5. The method for obtaining reconstruction data according to claim 1, characterized in that: Scanning the same sample using the second preset number of channels of the field of view and the same preset scanning parameters to obtain second reconstruction data further comprises: The same sample is scanned using a plurality of different numbers of channels of the second preset field of view and the same preset scanning parameters to obtain a plurality of second reconstruction data.
6. The method for obtaining reconstruction data according to claim 5, characterized in that: Scanning a plurality of different samples and the corresponding preset scanning parameters to obtain a plurality of sets of the first reconstruction data and the second reconstruction data includes: Each second reconstruction data is matched with the corresponding first reconstruction data to obtain multiple groups of the first reconstruction data and the second reconstruction data.
7. A method for obtaining a reconstruction data unit for CT scanning, characterized in that: The method is configured to obtain reconstruction data by executing the method according to any one of claims 1 to 6.
8. A CT machine, characterized in that: include: Obtaining a reconstruction data unit as claimed in claim 7.
9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the program is executed by a processor, the method according to any one of claims 1 to 6 is implemented.
10. An electronic device, characterized in that: include: one or more processors; A storage device for storing one or more programs, which, when executed by the one or more processors, enables the one or more processors to implement the method according to any one of claims 1 to 6.