CT imaging method, device, storage medium and electronic equipment
Through continuous sampling and sparse simulation technology, combined with neural network post-processing, the problem of high-dose radiation in CT scans is solved, and the reconstruction of high-quality images at low doses is achieved, reducing the radiation risk of patients.
Patent Information
- Application Number
- CN202210562765.9
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2022-05-23
- Publication Date
- 2025-08-12
- Estimated Expiration
- 2042-05-23
AI Technical Summary
In existing CT scanning techniques, patients are exposed to high X-ray doses during multiple scans, resulting in an increased risk of radiation-induced cancer and gene mutations.
Using continuous sampling and sparse simulation technology, the target sparseness is determined by sparse simulation of continuous sampling images, and the image is scanned and reconstructed according to the sparseness, and the neural network is used for post-processing to optimize image quality.
Obtaining high-quality CT images at low doses reduces the risk of radiation exposure in patients and improves image reconstruction effect.
Smart Images
Figure CN115131454B_ABST
Abstract
Description
Technical Field
[0001] The present application relates to the field of CT scanning technology, and in particular to a CT imaging method, device, storage medium and electronic equipment. Background Art
[0002] Computed tomography (CT) is an important method for obtaining information about a patient's internal structures. During a CT scan, a tube emits high doses of X-rays that pass through the body. Detectors then receive these X-rays, generating scan data. Reconstruction algorithms reconstruct cross-sectional images from these raw data. However, due to the radiation effects of X-rays, the risk of radiation-induced cancer and genetic mutations increases with increasing radiation dose.
[0003] Existing perfusion scans and other similar scanning methods first inject a contrast agent into the patient's body, then repeatedly scan a specific area of the patient to observe the flow of the contrast agent in the patient's body and determine whether there are abnormal blood flow tissues in the patient's body. Because patients need to undergo multiple and long scans, they are at risk of being exposed to high X-ray doses. Summary of the Invention
[0004] In view of this, the present application provides a CT imaging method, apparatus, medium and equipment, which can obtain high-quality images at low doses, solving the problem of excessively high X-ray doses in the prior art.
[0005] According to one aspect of the present application, a CT imaging method is provided, wherein the method is used for CT imaging of repeated scanning of a region, comprising:
[0006] Continuously sampling the target object to obtain continuous sampling images, performing sparse simulation on the continuous sampling images, and determining the target sparsity according to the simulation results;
[0007] Scanning is completed and imaging is created according to the target sparsity.
[0008] Optionally, performing sparse simulation on the continuously sampled images specifically includes:
[0009] determining a plurality of simulated sparsities according to a preset spacing;
[0010] The continuous sampling images are processed respectively according to each of the simulated sparsity to obtain a sparse sampling image corresponding to each of the simulated sparsity.
[0011] Optionally, determining the target sparsity according to the simulation result specifically includes:
[0012] Inputting each of the sparsely sampled images and the continuously sampled images into a preset quality assessment model, and calculating the quality value of the sparsely sampled images using the quality assessment model;
[0013] The target sparsity is determined from among the plurality of sparsities according to the quality value.
[0014] Optionally, determining the target sparsity from the multiple sparsities according to the quality value specifically includes:
[0015] Determine the sparsely sampled image whose quality value is greater than or equal to a preset quality threshold as a candidate image;
[0016] The sparsity corresponding to the candidate image with the smallest quality value is determined as the target sparsity.
[0017] Optionally, completing scanning and imaging according to target sparsity specifically includes:
[0018] Acquire the number of sparse sampling sequences, and determine a target scanning angle for each sparse sampling sequence according to the number of sparse sampling sequences and the target sparsity;
[0019] The target object is sampled based on the target scanning angle of each sparse sampling sequence to obtain a sparse sampling image corresponding to each sparse sampling sequence.
[0020] Optionally, determining a target scanning angle for each sparse sampling sequence according to the number of the sparse sampling sequences and the target sparsity specifically includes:
[0021] Determining an interval scanning angle between target scanning angles of each of the sparse sampling sequences according to the target sparsity;
[0022] An initial scanning angle of each of the sparse sampling sequences is determined according to the interval scanning angle and the number of the sparse sampling sequences, and the target scanning angle is determined according to the initial scanning angle and the interval scanning angle.
[0023] Optionally, determining the initial scanning angle of each of the sparse sampling sequences according to the interval scanning angle and the number of the sparse sampling sequences specifically includes:
[0024] Get the initial scan angle of the first sparse sampling sequence
[0025] The initial scanning angle of the i-th sparse sampling sequence is in, is the interval scanning angle, n is the number of sparse sampling sequences, i∈[2,n].
[0026] Optionally, after obtaining the sparse sampling image corresponding to each of the sparse sampling sequences, the method further includes:
[0027] Post-processing is performed on the sparsely sampled image to obtain a target image.
[0028] According to another aspect of the present application, a CT imaging device is provided, comprising:
[0029] A calculation module is used to continuously sample the target object to obtain a continuous sampling image, perform sparse simulation on the continuous sampling image, and determine the target sparsity according to the simulation result;
[0030] The scanning module is used to complete scanning and create an image according to the target sparsity.
[0031] Optionally, the operation module is specifically configured to:
[0032] determining a plurality of simulated sparsities according to a preset spacing;
[0033] The continuous sampling images are processed respectively according to each of the simulated sparsity to obtain a sparse sampling image corresponding to each of the simulated sparsity.
[0034] Optionally, the operation module is further configured to:
[0035] Inputting each of the sparsely sampled images and the continuously sampled images into a preset quality assessment model, and calculating the quality value of the sparsely sampled images using the quality assessment model;
[0036] The target sparsity is determined from among the plurality of sparsities according to the quality value.
[0037] Optionally, the operation module is further configured to:
[0038] Determine the sparsely sampled image whose quality value is greater than or equal to a preset quality threshold as a candidate image;
[0039] The sparsity corresponding to the candidate image with the smallest quality value is determined as the target sparsity.
[0040] Optionally, the scanning module is further configured to:
[0041] Acquire the number of sparse sampling sequences, and determine a target scanning angle for each sparse sampling sequence according to the number of sparse sampling sequences and the target sparsity;
[0042] The target object is sampled based on the target scanning angle of each sparse sampling sequence to obtain a sparse sampling image corresponding to each sparse sampling sequence.
[0043] Optionally, the scanning module is specifically configured to:
[0044] Determining an interval scanning angle between target scanning angles of each of the sparse sampling sequences according to the target sparsity;
[0045] An initial scanning angle of each of the sparse sampling sequences is determined according to the interval scanning angle and the number of the sparse sampling sequences, and the target scanning angle is determined according to the initial scanning angle and the interval scanning angle.
[0046] Optionally, the scanning module is further configured to:
[0047] Get the initial scan angle of the first sparse sampling sequence
[0048] The initial scanning angle of the i-th sparse sampling sequence is in, is the interval scanning angle, n is the number of sparse sampling sequences, i∈[2,n].
[0049] Optionally, the device further includes a post-processing module, specifically configured to:
[0050] Post-processing is performed on the sparsely sampled image to obtain a target image.
[0051] According to another aspect of the present application, a storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, the above-mentioned CT imaging method is implemented.
[0052] According to another aspect of the present application, an electronic device is provided, comprising a storage medium, a processor, and a computer program stored on the storage medium and executable on the processor, wherein the processor implements the above-mentioned CT imaging method when executing the computer program.
[0053] By means of the above technical solution, the present application performs sparse simulation and quality evaluation on the images obtained by continuous sampling, and determines the most appropriate sparsity based on the evaluation results. Subsequently, the sparsity is used to perform complementary scanning of the angles between sequences, that is, there is a deviation in the target sampling angles of different sparse sampling sequences. Since different sparse sampling sequences are all scanning the same position, this complementary scanning method can make the information between each sparse sampling sequence complement each other. Finally, fbp imaging is used to obtain images of all sequences. Since the data acquired at different times have a strong correlation, the imaging of the same position at different times, that is, different sampling sequences, is used as input, and the input is post-processed using methods such as neural networks to output high-quality reconstructed images of the corresponding sequence. Higher-quality images can be obtained under low-dose conditions, which solves the problem of excessive X-ray dose in the prior art and reduces the risk of patients being exposed to high X-ray doses.
[0054] The above description is only an overview of the technical solution of the present application. In order to more clearly understand the technical means of the present application, it can be implemented in accordance with the contents of the specification. In order to make the above and other purposes, features and advantages of the present application more obvious and easy to understand, the specific implementation methods of the present application are listed below. BRIEF DESCRIPTION OF THE DRAWINGS
[0055] The drawings described herein are used to provide a further understanding of the present application and constitute a part of the present application. The illustrative embodiments of the present application and their descriptions are used to explain the present application and do not constitute an improper limitation on the present application. In the drawings:
[0056] Figure 1 A schematic diagram of a CT imaging method according to an embodiment of the present application is shown;
[0057] Figure 2 A schematic diagram of a process of another CT imaging method provided in an embodiment of the present application is shown;
[0058] Figure 3 A schematic diagram of a process of another CT imaging method provided in an embodiment of the present application is shown;
[0059] Figure 4 A schematic diagram of a process of another CT imaging method provided in an embodiment of the present application is shown;
[0060] Figure 5 A schematic diagram of a process of another CT imaging method provided in an embodiment of the present application is shown;
[0061] Figure 6 A schematic diagram of a process of another CT imaging method provided in an embodiment of the present application is shown;
[0062] Figure 7 A schematic diagram of a process of another CT imaging method provided in an embodiment of the present application is shown;
[0063] Figure 8 A schematic diagram of an initial scanning angle of another CT imaging method provided in an embodiment of the present application is shown;
[0064] Figure 9 A schematic diagram of a process of another CT imaging method provided in an embodiment of the present application is shown;
[0065] Figure 10 A neural network diagram of another CT imaging method provided in an embodiment of the present application is shown;
[0066] Figure 11 A structural block diagram of a CT imaging device provided in an embodiment of the present application is shown. DETAILED DESCRIPTION
[0067] In this embodiment, a CT imaging method is provided for CT imaging of repeated scanning of an area, such as Figure 1 As shown, the method includes:
[0068] Step 101: Continuously sample the target object to obtain continuous sampling images, perform sparse simulation on the continuous sampling images, and determine the target sparsity based on the simulation results.
[0069] The CT imaging method provided in the embodiment of the present application is aimed at perfusion or other imaging methods that repeatedly scan the same area. The imaging method first injects a contrast agent into the patient's body, and then performs multiple scans on the position to be scanned, thereby analyzing the flow of the contrast agent in the patient's body based on the results of the multiple scans to obtain the final scanning result.
[0070] Based on this, this application first adopts a continuous sampling method and performs FPB (Filtered Back Projection) imaging based on the sampling results to obtain a continuous sampling image. Then, software is used to perform sparse simulation under different sparsities, and then a target sparsity is determined based on the simulation results. Among them, common software in the industry can be used and is not limited here.
[0071] Step 102: Complete scanning and create an image according to the target sparsity.
[0072] In this embodiment, X-rays are emitted according to the target sparsity to obtain scanned projections, and then an image is created using the scanned projections.
[0073] By applying the technical solution of this embodiment, sparse simulation is performed on continuously sampled images and their quality is evaluated. Based on the evaluation results, the most appropriate sparsity, also known as the target sparsity, is determined. Scanning is then performed using this target sparsity, and the resulting projections are used to construct an image. This embodiment produces high-quality images at low doses, resolving the issue of excessive X-ray doses in existing technologies and reducing the risk of high X-ray dose exposure for patients.
[0074] One embodiment of the present application includes Figure 2Specifically, the steps shown are as follows: a continuous sampling method is adopted in the first sequence, and then FPB imaging is performed. Subsequently, a sparse simulation is performed in the software background, and the imaging quality under different sparsities is evaluated. An appropriate sparsity is automatically selected based on the imaging quality. Subsequent sparse sampling sequences are scanned according to this sparsity, wherein there are multiple sparse sampling sequences, and the target scanning angles of different sparse sampling sequences are different to achieve complementarity of the target scanning angles. After the sampling results of all sparse sampling sequences are respectively subjected to FBP imaging, the continuous sampling images obtained by continuous sampling and the sparse sampling images obtained by sparse angle sampling are input into the neural network, and each sparse sampling image is post-processed and optimized using a deep learning method to improve image quality.
[0075] Furthermore, as a refinement and extension of the specific implementation of the above embodiment, in order to fully illustrate the specific implementation process of this embodiment, another CT imaging method is provided, such as Figure 3 As shown, sparse simulation is performed on the continuous sampling image, specifically including:
[0076] Step 201, determining multiple simulation sparsities according to a preset spacing;
[0077] Step 202 : Process the continuous sampling images according to each simulated sparsity to obtain a sparse sampling image corresponding to each simulated sparsity.
[0078] In this embodiment, different simulated sparsities are determined based on a preset spacing. Simulations of different sparsities are performed on continuously sampled data at equal intervals to obtain simulation results corresponding to each sparsity. FBP imaging is then performed on each simulation result to obtain a sparsely sampled image. The preset spacing can be set according to actual needs.
[0079] Specifically, the range of the simulated sparsity is [0, 1], and the spacing between multiple simulated sparsities is the same, which is the preset spacing. For example, if the preset spacing is 0.2, then the following simulated sparsities can be determined: 0, 0.2, 0.4, 0.6, 0.8, 1.
[0080] Furthermore, if Figure 4 As shown, in another CT imaging method, the target sparsity is determined according to the simulation results, specifically including:
[0081] Step 301: Input each sparsely sampled image and the continuously sampled image into a preset quality assessment model, and calculate the quality value of the sparsely sampled image using the quality assessment model;
[0082] Step 302: Determine a target sparsity from among multiple sparsities according to the quality value.
[0083] In this embodiment, sparsely sampled images and continuously sampled images are input into a preset quality assessment model, where the preset quality assessment model is used to evaluate image quality. The continuously sampled images are used as a benchmark for evaluation, and the sparsely sampled images are compared with this benchmark to obtain a quality value for each sparsely sampled image. A target sparsity level is then selected from multiple sparsities based on the quality value.
[0084] The SSIM (structural similarity index) value can be used to characterize image quality. From the perspective of image composition, the SSIM defines structural information as a property that reflects the structure of objects in a scene, independent of brightness and contrast. It also models distortion as a combination of three different factors: brightness, contrast, and structure. The mean is used as an estimate of brightness, the standard deviation as an estimate of contrast, and the covariance as a measure of structural similarity.
[0085] Specifically, the sparsely sampled image I p With the continuous sampling image I 1.0 The SSIM value between the two is used as the sparse image I p The calculation formula is as follows:
[0086]
[0087] in, and u p represent the mean of the continuously sampled image and the sparsely sampled image respectively; and σ p represent the standard deviations of continuously sampled images and sparsely sampled images respectively; and represent the variance of the continuously sampled image and the sparsely sampled image respectively; Represents the covariance of the two images. C1 and C2 are constants to avoid the denominator being zero and maintain stability.
[0088] The mean, variance and covariance mentioned above may be calculated by using methods such as Gaussian function. Of course, other methods such as traversing pixel points may also be used, which is not limited here.
[0089] In addition, although the specification uses the preset quality assessment model established based on SSIM as an example, other models that can evaluate image quality can also be adopted, for example, a model established based on PSNR (Peak Signal to Noise Ration), which is also an embodiment of the present invention.
[0090] Furthermore, if Figure 5As shown, in another CT imaging method, determining a quality value determines a target sparsity among multiple sparsities, specifically including:
[0091] Step 401: Determine a sparsely sampled image with a quality value greater than or equal to a preset quality threshold as a candidate image;
[0092] Step 402: Determine the sparsity corresponding to the candidate image with the smallest quality value as the target sparsity.
[0093] In this embodiment, it is understood that the better the image quality, the more it reflects the patient's scanned position. Therefore, a preset quality threshold can be set in advance, and sparsely sampled images with quality values greater than the preset quality threshold are selected as candidate images, thereby eliminating sparsely sampled images of low quality. The preset quality threshold can be obtained based on historical scan records or the operator's experience.
[0094] In addition, the higher the image quality, the higher the scanning dose. Excessive scanning dose may cause harm to the patient. Based on this, the one with the smallest quality value is screened out from the alternative images, and the sparsity corresponding to this alternative image is used as the target sparsity.
[0095] In this embodiment, the solution of determining the target sparsity based on the quality value comprehensively considers the accuracy of the scan result and the scan dose received by the patient, and minimizes the radiation received by the patient while the accuracy meets the requirements.
[0096] Furthermore, if Figure 6 As shown, in another CT imaging method, scanning and imaging are completed according to the target sparsity, specifically including:
[0097] Step 501: Obtain the number of sparse sampling sequences, and determine the target scanning angle of each sparse sampling sequence according to the number of sparse sampling sequences and the target sparsity;
[0098] Step 502 : performing sampling processing on the target object based on the target scanning angle of each sparse sampling sequence to obtain a sparse sampling image corresponding to each sparse sampling sequence.
[0099] In this embodiment, multiple sparse sampling operations are performed after continuous sampling, each time executing a sparse sampling sequence. The scanning angles of two adjacent sparse sampling sequences are different. For example, the target scanning angles of the first sparse sampling sequence are 0 degrees, 5 degrees, 10 degrees, and 15 degrees; while the target scanning angles of the second sparse sampling sequence are 1 degree, 6 degrees, 11 degrees, and 16 degrees. By scanning the same location at different angles, the information between the sparse sampling sequences is mutually complementary.
[0100] Therefore, the number of sparse sampling sequences is obtained, and the target scanning angle of the sampling sequence is determined based on the number of sparse sampling sequences and the target sparsity. It can be understood that the smaller the number of sparse sampling sequences, the greater the deviation in scanning angle between two adjacent sparse sampling sequences; and the greater the target sparsity, the greater the deviation in scanning angle between two adjacent sparse sampling sequences.
[0101] Finally, X-rays are emitted according to the target scanning angle of each sparse sampling sequence to obtain a projection based on the target scanning angle, and then a sparse sampling image corresponding to the sampling sequence is obtained.
[0102] Among them, the number of sparse sampling sequences can be set by the doctor according to the patient's condition.
[0103] This embodiment uses target sparsity to perform inter-sequence angle complementary scanning, that is, there is a deviation in the target sampling angles of different sparse sampling sequences. Since different sparse sampling sequences scan the same position, this complementary scanning method can enable the information between the sparse sampling sequences to complement each other.
[0104] Furthermore, if Figure 7 As shown, in another CT imaging method, the target scanning angle of each sparse sampling sequence is determined according to the number of sparse sampling sequences and the target sparsity, specifically including:
[0105] Step 601, determining the interval scanning angle between target scanning angles of each sparse sampling sequence according to the target sparsity;
[0106] Step 602: Determine an initial scanning angle for each sparse sampling sequence according to the interval scanning angle and the number of sparse sampling sequences, and determine a target scanning angle according to the initial scanning angle and the interval scanning angle.
[0107] In this embodiment, each sparse sampling sequence includes multiple target scanning angles, which form an arithmetic progression. It will be appreciated that the sparser the sampling sequence, the fewer target scanning angles each sparse sampling sequence contains, and thus the larger the interval between scanning angles. Based on this, the interval between scanning angles can be determined based on the target sparsity.
[0108] Furthermore, after the first sparse sampling sequence is scanned based on the corresponding first target scanning angle, i.e., the initial scanning angle, the first sparse sampling sequence is scanned based on the corresponding second target scanning angle. After the first sparse sampling sequence is completely scanned, the second sparse sampling sequence is scanned based on its corresponding scanning angle. This operation is performed sequentially for each sparse sampling sequence until all sequences are scanned.
[0109] Therefore, after the interval scanning angle is determined, the initial scanning angle of each sparse sampling sequence can be determined according to the interval scanning angle and the number of sparse sampling sequences, wherein the initial scanning angle of the first sparse sampling sequence is preset.
[0110] Finally, multiple target scanning angles for each sparse sampling sequence are determined based on the initial scanning angle and the interval scanning angle of each sparse sampling sequence. For example, if the initial scanning angle of a sparse sampling sequence is determined to be 0 degrees and the interval scanning angle is determined to be 30 degrees, the target scanning angles can be determined to be 0 degrees, 30 degrees, 60 degrees, 90 degrees, ..., 360 degrees.
[0111] This embodiment uses the above method to form an arithmetic progression between all target scanning angles, so that the scanning of the scanned position is more uniform.
[0112] Furthermore, in another CT imaging method, determining the initial scanning angle of each sparse sampling sequence according to the interval scanning angle and the number of sparse sampling sequences specifically includes:
[0113] Step 701: Obtain the initial scanning angle of the first sparse sampling sequence
[0114] Step 702: The initial scanning angle of the i-th sparse sampling sequence is in, is the interval scanning angle, n is the number of sparse sampling sequences, i∈[2,n].
[0115] In this embodiment, the initial scanning angle may be preset, and the initial scanning angle of each sparse sampling sequence may be calculated in sequence according to the above formula.
[0116] For example, the initial scan angle of the first sparse sampling sequence is is 0 degrees, the interval scanning angle is 30 degrees, and the number of sparse sampling sequences is 5, then the initial scanning angle of the second sparse sampling sequence can be calculated according to the above formula Similarly, the initial scanning angles of the third to fifth sparse sampling sequences are 12 degrees, 18 degrees, and 24 degrees, respectively. At this time, the second target scanning angle of the first sparse sampling sequence is 30 degrees, and an arithmetic progression is formed between all target scanning angles.
[0117] Figure 8A schematic diagram of three sparse angular scans is shown, including three sparse sampling sequences, each with its scanning angles marked with a different line type. As can be seen from the figure, the target scanning angles of the three sparse sampling sequences are arranged alternately, and the differences between all angles are the same, thus achieving complementary angular scanning. In this embodiment, the interval between two adjacent target scanning angles in each sparse sampling sequence is the same, that is, the target scanning angles in each sparse sampling sequence are equally spaced.
[0118] In other exemplary implementations, the interval angles between two adjacent target scanning angles of each sparse sampling sequence may be different, and the interval angle may be determined according to the attenuation or eccentricity of the region of interest, wherein the eccentricity may be the degree to which the center of the region of interest deviates from the scanning center.
[0119] This embodiment has simple calculation steps for the target scanning angle and realizes complementary scanning of angles to obtain scanning results at different angles.
[0120] Furthermore, in another CT imaging method, after obtaining the sparse sampling image corresponding to each sparse sampling sequence, the method further includes:
[0121] The sparsely sampled image is post-processed to obtain the target image.
[0122] In this embodiment, since the data acquired at different times have a strong correlation, the imaging of the same position at different times, i.e., different sampling sequences, is used as input, and the input is post-processed using methods such as neural networks to output high-quality reconstructed images of the corresponding sequence.
[0123] Specifically, a neural network can be used to perform post-processing operations on the sparsely sampled image to improve the quality of the sparsely sampled image and obtain the target image. Of course, sparsely sampled images can also be processed by iteration or other methods, which are not limited here.
[0124] The neural network may be a convolutional neural network (CNN) or other neural networks.
[0125] Furthermore, the above operation can be performed for each position to be scanned of the patient respectively to obtain a high-quality target image corresponding to the position, until all positions to be scanned are processed.
[0126] Furthermore, if Figure 9 As shown in Figure 2, if a neural network is used to post-process the sparsely sampled image, the following steps must be included before post-processing:
[0127] Step 801, constructing a neural network and setting a loss function;
[0128] Step 802: continuously sampling the samples to obtain a first image, and performing sparse simulation on the first image to obtain a second image;
[0129] Step 803: input the first image and the second image into a neural network, and calculate a function value of a loss function according to an output of the neural network;
[0130] Step 804: Adjust the parameters of the neural network according to the function value until the function value satisfies a second preset condition.
[0131] In this embodiment, before using a neural network to process sparsely sampled images, the neural network can be trained to optimize the neural network and produce better results. Specifically, a large number of samples are scanned using a continuous angle sampling method to acquire high-dose scan data, resulting in multiple first images. A sparse angle simulation tool is then used to perform a sparse simulation on each first image, generating multiple second images corresponding to the first image. A data set is then established based on the first and second images.
[0132] Then the first image and the corresponding second image are input into the neural network respectively. Since the first image obtained by continuous sampling is the best image that can be obtained by sparse sampling under ideal conditions, the first image can be used as the label of the neural network and the second image as the input of the neural network.
[0133] Finally, the loss function value is calculated based on the output, and then the parameters of the neural network are adjusted to reduce the loss function value. That is, the network parameters are adjusted to make the output result closer to the first image. In this way, the neural network is trained to make the image output by the neural network higher in quality.
[0134] Furthermore, the loss function is as follows:
[0135]
[0136] Among them, T, W and H are the number, width and height of the input images respectively, x t,w,h and y t,w,h are the pixel values of the output image and the label image (ie the first image) at (t, w, h) respectively.
[0137] Furthermore, the loss function may also be a loss function that can perform image reconstruction, such as an MSE (Mean Squared Error) loss function, a perceptual loss function, an adversarial loss function, and the like.
[0138] Figure 10A schematic diagram of a neural network of an embodiment of the present application is shown. As shown in the figure, 3Dconv is a convolutional layer, which is used for feature extraction and image reduction; among them, LReLU, also known as Leaky-ReLU, is an activation function for deep learning. Including an excitation function in the convolutional layer can help express complex features; Max pooling is in the pooling layer, which is used to implement maximum pooling. After feature extraction in the convolutional layer, the output feature map will be passed to the pooling layer for feature selection and information filtering, and the result of a single point in the feature map will be replaced by the feature map statistics of its adjacent area; Concat is a connection layer, which is used to splice multiple feature maps; up-sampling is upsampling, which can expand the size of the feature map.
[0139] Furthermore, as a specific implementation of the above-mentioned CT imaging method, the embodiment of the present application provides a CT imaging device, such as Figure 11 As shown, the CT imaging device includes: a calculation module and a scanning module.
[0140] A calculation module is used to continuously sample the target object to obtain a continuous sampling image, perform sparse simulation on the continuous sampling image, and determine the target sparsity according to the simulation result;
[0141] The scanning module is used to complete scanning and build images according to the target sparsity.
[0142] Optionally, the operation module is specifically configured to:
[0143] determining a plurality of simulated sparsities according to a preset spacing;
[0144] The continuous sampling images are processed according to each simulated sparsity to obtain a sparse sampling image corresponding to each simulated sparsity.
[0145] Optionally, the computing module is further configured to:
[0146] Input each sparsely sampled image and the continuously sampled image into a preset quality assessment model, and use the quality assessment model to calculate the quality value of the sparsely sampled image;
[0147] A target sparsity is determined among multiple sparsities according to a quality value.
[0148] Optionally, the computing module is further configured to:
[0149] Determine the sparsely sampled images whose quality values are greater than or equal to a preset quality threshold as candidate images;
[0150] The sparsity corresponding to the candidate image with the smallest quality value is determined as the target sparsity.
[0151] Optionally, the scanning module is further configured to:
[0152] Obtaining the number of sparse sampling sequences, and determining a target scanning angle for each sparse sampling sequence according to the number of sparse sampling sequences and target sparsity;
[0153] Based on the target scanning angle of each sparse sampling sequence, the target object is sampled and processed to obtain a sparse sampling image corresponding to each sparse sampling sequence.
[0154] Optionally, the scanning module is specifically configured to:
[0155] Determining the interval scanning angle between the target scanning angles of each sparse sampling sequence according to the target sparsity;
[0156] An initial scanning angle of each sparse sampling sequence is determined according to the interval scanning angle and the number of sparse sampling sequences, and a target scanning angle is determined according to the initial scanning angle and the interval scanning angle.
[0157] Optionally, the scanning module is further configured to:
[0158] Get the initial scan angle of the first sparse sampling sequence
[0159] The initial scanning angle of the i-th sparse sampling sequence is in, is the interval scanning angle, n is the number of sparse sampling sequences, i∈[2,n].
[0160] Optionally, the device further includes a post-processing module, specifically configured to:
[0161] The sparsely sampled image is post-processed to obtain the target image.
[0162] It should be noted that for other corresponding descriptions of the functional modules involved in the CT imaging device provided in the embodiment of the present application, please refer to Figures 1 to 10 The corresponding description in will not be repeated here.
[0163] Based on the above Figures 1 to 10 The method shown in FIG. 1 is a method for performing the above-mentioned operation. Accordingly, the embodiment of the present application further provides a storage medium on which a computer program is stored. When the program is executed by a processor, the above-mentioned operation is performed. Figures 1 to 10 CT imaging method shown.
[0164] Based on this understanding, the technical solution of the present application can be embodied in the form of a software product, which can be stored in a non-volatile storage medium (which can be a CD-ROM, USB flash drive, mobile hard disk, etc.), including a number of instructions for enabling an electronic device (which can be a personal computer, server, or network device, etc.) to execute the methods described in each implementation scenario of the present application.
[0165] Based on the above Figures 1 to 10 The method shown, and Figure 11 In order to achieve the above-mentioned purpose, the embodiment of the CT imaging device shown in the figure further provides an electronic device, which can be a personal computer, a server, a network device, etc., and the electronic device includes a storage medium and a processor; the storage medium is used to store a computer program; the processor is used to execute the computer program to achieve the above-mentioned Figures 1 to 10 CT imaging method shown.
[0166] Optionally, the electronic device may further include a user interface, a network interface, a camera, a radio frequency (RF) circuit, a sensor, an audio circuit, a Wi-Fi module, etc. The user interface may include a display, an input unit such as a keyboard, etc., and the optional user interface may also include a USB interface, a card reader interface, etc. The network interface may optionally include a standard wired interface, a wireless interface (such as a Bluetooth interface, a Wi-Fi interface), etc.
[0167] Those skilled in the art will understand that the electronic device structure provided in this embodiment does not limit the electronic device, and may include more or fewer components, or combine certain components, or arrange the components differently.
[0168] The storage medium may also include an operating device and a network communication module. The operating device is a program that manages and stores the hardware and software resources of the electronic device, supporting the operation of information processing programs and other software and / or programs. The network communication module is used to enable communication between the various controls within the storage medium and with other hardware and software in the physical device.
[0169] Through the description of the above implementation methods, those skilled in the art can clearly understand that the present application can be implemented by means of software plus necessary general hardware platforms.
[0170] Those skilled in the art will understand that the accompanying drawings are only schematic diagrams of a preferred implementation scenario, and the units or processes in the accompanying drawings are not necessarily required to implement the present application. Those skilled in the art will understand that the units in the devices in the implementation scenario can be distributed in the devices of the implementation scenario according to the implementation scenario description, or can be changed accordingly and located in one or more devices different from the implementation scenario. The units of the above-mentioned implementation scenario can be combined into one unit, or can be further split into multiple sub-units.
[0171] The serial numbers of the above application are for descriptive purposes only and do not represent the advantages or disadvantages of the implementation scenarios. The above disclosure only discloses several specific implementation scenarios of the present application, but the present application is not limited thereto. Any changes that can be conceived by those skilled in the art should fall within the scope of protection of the present application.
Claims
1. A CT imaging method, characterized in that: The method is used for CT imaging of repeated scanning of an area, and the method comprises: Continuously sampling the target object to obtain continuous sampling images, performing sparse simulation on the continuous sampling images, and determining the target sparsity according to the simulation results; Complete scanning and image creation according to the target sparsity, specifically including: obtaining the number of sparse sampling sequences; Determining an interval scanning angle between target scanning angles of each of the sparse sampling sequences according to the target sparsity; determining an initial scanning angle for each of the sparse sampling sequences according to the interval scanning angle and the number of the sparse sampling sequences, and determining a target scanning angle according to the initial scanning angle and the interval scanning angle, wherein the target scanning angles of two adjacent sparse sampling sequences are different; The target object is sampled based on the target scanning angle of each sparse sampling sequence to obtain a sparse sampling image corresponding to each sparse sampling sequence.
2. The method according to claim 1, characterized in that The performing sparse simulation on the continuous sampling images specifically includes: determining a plurality of simulated sparsities according to a preset spacing; The continuous sampling images are processed respectively according to each of the simulated sparsity to obtain a sparse sampling image corresponding to each of the simulated sparsity.
3. The method according to claim 2, characterized in that Determining the target sparsity according to the simulation results specifically includes: Inputting each of the sparsely sampled images and the continuously sampled images into a preset quality assessment model, and calculating the quality value of the sparsely sampled images using the quality assessment model; The target sparsity is determined from among the plurality of sparsities according to the quality value.
4. The method according to claim 3, characterized in that Determining the target sparsity from the multiple sparsities according to the quality value specifically includes: Determine the sparsely sampled image whose quality value is greater than or equal to a preset quality threshold as a candidate image; The sparsity corresponding to the candidate image with the smallest quality value is determined as the target sparsity.
5. The method according to claim 1, wherein The determining the initial scanning angle of each of the sparse sampling sequences according to the interval scanning angle and the number of the sparse sampling sequences specifically includes: Get the initial scan angle of the first sparse sampling sequence The initial scanning angle of the i-th sparse sampling sequence is in, is the interval scanning angle, n is the number of sparse sampling sequences, i∈[2,n].
6. A CT imaging device, characterized in that: The device is used for CT imaging of repeated scanning of an area, and the device comprises: A calculation module is used to continuously sample the target object to obtain a continuous sampling image, perform sparse simulation on the continuous sampling image, and determine the target sparsity according to the simulation result; The scanning module is used to complete scanning and image creation according to the target sparsity, specifically including: obtaining the number of sparse sampling sequences; Determining an interval scanning angle between target scanning angles of each of the sparse sampling sequences according to the target sparsity; determining an initial scanning angle for each of the sparse sampling sequences according to the interval scanning angle and the number of the sparse sampling sequences, and determining a target scanning angle according to the initial scanning angle and the interval scanning angle, wherein the target scanning angles of two adjacent sparse sampling sequences are different; The target object is sampled based on the target scanning angle of each sparse sampling sequence to obtain a sparse sampling image corresponding to each sparse sampling sequence.
7. A storage medium having a program or instruction stored thereon, characterized in that: When the program or instruction is executed by a processor, the steps of the method according to any one of claims 1 to 5 are implemented.
8. An electronic device, characterized in that: The method comprises a storage medium, a processor, and a computer program stored on the storage medium and executable on the processor, wherein the processor implements the method according to any one of claims 1 to 5 when executing the program.