CN111260583A - Multi-discriminant-based multi-analysis network missing CT projection data estimation method - Google Patents
Multi-discriminant-based multi-analysis network missing CT projection data estimation method Download PDFInfo
- Publication number
- CN111260583A CN111260583A CN202010056809.1A CN202010056809A CN111260583A CN 111260583 A CN111260583 A CN 111260583A CN 202010056809 A CN202010056809 A CN 202010056809A CN 111260583 A CN111260583 A CN 111260583A
- Authority
- CN
- China
- Prior art keywords
- projection data
- missing
- image
- sparse
- region
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 35
- 238000004458 analytical method Methods 0.000 title claims abstract description 25
- 238000001914 filtration Methods 0.000 claims abstract description 4
- 238000012549 training Methods 0.000 claims description 35
- 230000003042 antagnostic effect Effects 0.000 claims description 12
- 238000012217 deletion Methods 0.000 claims description 10
- 230000037430 deletion Effects 0.000 claims description 10
- 230000005484 gravity Effects 0.000 claims description 9
- 230000008485 antagonism Effects 0.000 claims description 8
- 238000012545 processing Methods 0.000 claims description 5
- ORILYTVJVMAKLC-UHFFFAOYSA-N Adamantane Natural products C1C(C2)CC3CC1CC2C3 ORILYTVJVMAKLC-UHFFFAOYSA-N 0.000 claims description 3
- 238000005457 optimization Methods 0.000 claims description 3
- 230000006870 function Effects 0.000 description 6
- 230000005855 radiation Effects 0.000 description 5
- 238000004422 calculation algorithm Methods 0.000 description 3
- 238000002591 computed tomography Methods 0.000 description 3
- 230000007812 deficiency Effects 0.000 description 3
- 238000013459 approach Methods 0.000 description 2
- 238000003759 clinical diagnosis Methods 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 230000009467 reduction Effects 0.000 description 2
- 206010028980 Neoplasm Diseases 0.000 description 1
- 238000013528 artificial neural network Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 201000011510 cancer Diseases 0.000 description 1
- 230000015556 catabolic process Effects 0.000 description 1
- 238000013170 computed tomography imaging Methods 0.000 description 1
- 238000013527 convolutional neural network Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000006731 degradation reaction Methods 0.000 description 1
- 238000003745 diagnosis Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 230000001815 facial effect Effects 0.000 description 1
- 230000005865 ionizing radiation Effects 0.000 description 1
- 230000008407 joint function Effects 0.000 description 1
- 208000032839 leukemia Diseases 0.000 description 1
- 230000006371 metabolic abnormality Effects 0.000 description 1
- 210000000056 organ Anatomy 0.000 description 1
- 238000007781 pre-processing Methods 0.000 description 1
- 230000008569 process Effects 0.000 description 1
- 238000007637 random forest analysis Methods 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 238000005070 sampling Methods 0.000 description 1
- 238000012360 testing method Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T5/00—Image enhancement or restoration
- G06T5/77—Retouching; Inpainting; Scratch removal
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10072—Tomographic images
- G06T2207/10081—Computed x-ray tomography [CT]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20081—Training; Learning
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20084—Artificial neural networks [ANN]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30004—Biomedical image processing
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- Data Mining & Analysis (AREA)
- Evolutionary Computation (AREA)
- Artificial Intelligence (AREA)
- Biomedical Technology (AREA)
- Computing Systems (AREA)
- General Engineering & Computer Science (AREA)
- Life Sciences & Earth Sciences (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Health & Medical Sciences (AREA)
- Apparatus For Radiation Diagnosis (AREA)
- Image Processing (AREA)
Abstract
The invention discloses a multi-resolution network missing CT projection data estimation method based on multiple discriminators, which comprises the following steps: acquiring an incomplete CT projection data image; inputting the incomplete CT projection data image into a multiple analysis network model to obtain a complete CT projection data image; and reconstructing the complete CT projection data image by a convolution filtering back projection method to obtain a CT image. The method can obviously improve the definition of the predicted CT projection data image and the continuity of the boundary of the missing region, thereby improving the quality of the CT image, and the peak signal-to-noise ratio value and the structural similarity value both represent the superiority of the method.
Description
Technical Field
The invention relates to the technical field of image reconstruction, in particular to a multi-resolution network missing CT projection data estimation method based on multiple discriminators.
Background
As a currently common and effective clinical medical diagnostic tool, computed tomography provides clinicians with a wealth of information about the organs and tissues of the body for their diagnosis. But related studies have shown that: a complete CT scan is usually accompanied by a high level of ionizing radiation, which can induce metabolic abnormalities in the human body as well as cancer, leukemia, etc.
One of the important clinical methods for reducing the radiation dose of patients is to reduce the CT scanning range, i.e. limit the rotation angle range of the detector within a certain interval smaller than the standard, so as to greatly reduce the X-ray radiation dose of the patients as a whole. Although the limitation of the scanning range of the CT apparatus can reduce the amount of X-ray radiation to the patient, it can cause the obtained CT projection data to be partially missing, i.e. incomplete projection data is obtained, so that the quality of the reconstructed CT image is significantly reduced, and the requirement of clinical diagnosis cannot be met. Also in multi-row CT imaging, a reduction in the amount of X-ray radiation can cause a significant degradation in the quality of the reconstructed image. With the reduction of the scanning range, although the radiation dose to the patient is greatly reduced, a large amount of star-stripe artifacts and noise appear in the reconstructed image, and the resolution of the feature points is seriously influenced. Therefore, how to reconstruct a high-quality CT image meeting the clinical diagnosis requirement under the condition of reducing the scanning range, i.e., incomplete projection data, has important scientific significance and clinical practical value, and has attracted more and more attention of scholars at home and abroad.
The university of north carolina church mountain school IDEA research team in usa in 2014 utilizes algorithms such as random forests, convolutional neural networks and the like, and combines an automatic context model to estimate a normal dose PET image or a CT image from MRI and low dose PET images. Boublil et al propose a theoretical framework for using artificial neural networks to improve the performance of common CT image reconstruction algorithms and successfully apply them to low-dose medical image reconstruction. Dosovitskiy et al demonstrate that it is possible to reconstruct an image of an object by inverting the features of a deep convolutional network through a decoder network. Kingma et al propose a Variational Automatic Encoder (VAE) that normalizes the encoder by imposing an a priori on the potential units so that an image can be generated by sampling or inserting the potential units from the potential units. However, VAE generated images are often blurred due to their training targets based on pixel-wise gaussian likelihood. Wright et al accomplish an image as a task to recover a sparse signal from an input, and by solving a sparse linear system, an image can be recovered from some corrupted inputs. However, this algorithm requires that the image be highly structured (i.e., assuming the data points are located in a low-dimensional subspace), for example: well aligned facial images. Pathak et al propose to reconstruct images using a context encoder model, but still suffer from the problem of non-uniformity of pixel values that generate missing boundaries of the image.
Disclosure of Invention
Aiming at the defects of the prior art, the invention aims to provide a multi-resolution network missing CT projection data estimation method based on multiple discriminators, so as to solve the problem of insufficient definition of CT projection data in the prior art.
In order to solve the technical problems proposed by the background, the invention is realized by the following technical scheme:
a multi-discriminant-based multi-resolution network-missing CT projection data estimation method, the method comprising:
acquiring a sparse missing CT projection data image;
inputting the sparse missing CT projection data image into a multiple analysis network model to generate a complete CT projection data estimation image;
reconstructing the complete CT projection data estimation image by a convolution filtering back projection method to obtain a CT image;
the multiple analysis network model comprises a generation network, a plurality of missing area discrimination networks and a global discrimination network.
Further, the method for establishing the multiple analysis network model comprises the following steps:
taking the complete CT projection data image set as a training data image set;
performing sparse deletion processing on the training data image set to obtain a training set;
training a generated network through the training set to obtain a prediction result of a sparse missing part of the image;
cutting and extracting the prediction result of the sparse missing part to obtain an estimation image of a sparse missing region;
integrating the real projection data image of the non-sparse missing part and the estimation image of the sparse missing region to obtain an integrated sparse missing projection data estimation image;
training a region discrimination network through a real projection data image corresponding to the sparse missing region and an estimation image of the sparse missing region;
and training a global discriminator through the complete CT projection data image and the integrated sparse missing projection data estimation image to obtain a multiple analysis network model.
Further, the reconstruction loss of the multiple analysis network model is as follows:
wherein x is the true projection data image value;
is a binary mask: 1 indicates that the pixel value is input, and 0 indicates that the pixel value is discarded, i.e., the pixel value is missing; f (x) is the output of the network encoder section.
Further, the adversarial loss of the multi-resolution network model is as follows:
where x is the true projection data image value, D (x) is the output value obtained by inputting the data x into the discriminator D, and g (z) is a parameter function.
Further, the joint loss function of the multiple analysis network model is as follows:
J(x)＝Lrec+λgLag+λ1La1+λ2La2+...λiLai(3)
wherein L isrecI.e. the reconstruction loss L described earlierrec(x)，λgIs the specific weight ratio of the antagonistic losses of the global arbiter in the joint losses, LagIs the antagonism loss of the global arbiter; lambda [ alpha ]1Is the specific gravity ratio, λ, of the antagonistic loss of the region discriminator 1 in the joint loss2Is the specific gravity ratio, λ, of the antagonistic loss of the region discriminator 2 in the joint lossiIs the specific gravity ratio of the antagonistic loss of the region discriminator i in the combined loss, wherein i is the number of the region discriminators; l isaiIs the antagonism loss of the region discriminator i.
Further, the training optimization function of the multi-resolution network model is an Adam optimizer.
Compared with the prior art, the invention has the following beneficial effects:
the invention adopts the combination training of the global discriminator and the plurality of area discriminators to extract the image information characteristics, the plurality of area discriminators can well predict the projection data of each sparse missing area and help to generate the detail characteristics with clearer boundaries, the global discriminator can ensure the integral authenticity of the generated image, the image definition of the predicted CT projection data and the continuity of the boundaries of the missing areas can be obviously improved, the quality of the CT image is further improved, and the peak signal-to-noise ratio value and the structural similarity value both represent the superiority of the method.
Drawings
FIG. 1 is a general flow diagram of the process of the present invention;
FIG. 2 is a multi-resolution network model of a multi-arbiter.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention more apparent, the present invention is further described in detail with reference to the following embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.
A multi-discriminant-based multi-resolution network-missing CT projection data estimation method, the method comprising:
acquiring an incomplete CT projection data image;
inputting the incomplete CT projection data image into a multiple analysis network model to obtain a prediction result of the complete CT projection data image;
and cutting and extracting a plurality of sparse missing regions of the prediction result, and integrating the plurality of sparse missing regions with real projection data of the non-missing part of the obtained incomplete CT projection data image to obtain an integrated sparse missing projection data estimation image.
And reconstructing the integrated sparse missing projection data estimation image by a convolution filtering back projection method to obtain a CT image.
The multiple analysis network model comprises a generation network, a plurality of missing area judgment networks and a global judgment network.
The method for establishing the multi-analysis network model comprises the following steps:
taking the complete CT projection data image set as a training data image set;
performing sparse deletion processing on the training data image set to obtain a training set;
training a generated network through the training set to obtain a prediction result of a sparse missing part of the image;
cutting and extracting the prediction result of the sparse missing part to obtain an estimation image of a sparse missing region;
integrating the real projection data image of the non-sparse missing part and the estimation image of the sparse missing region to obtain an integrated sparse missing projection data estimation image;
training a region discrimination network through a real projection data image corresponding to the sparse missing region and an estimation image of the sparse missing region;
and training a global discriminator through the complete CT projection data image and the integrated sparse missing projection data estimation image to obtain a multiple analysis network model.
The training method of the multiple analytic network model comprises the following steps:
(1) taking the complete CT projection data image set as a training data image set, preprocessing a certain number of complete CT projection data images, namely performing sparse deletion processing on a scanning angle or a detector direction to obtain the training set, wherein the corresponding projection data images are respectively deleted in the vertical direction and the horizontal direction, and the pixel value of the projection data image in the deletion range is set to be 0.
(2) A multi-resolution network model of a multi-discriminator is constructed, as shown in fig. 1 and fig. 2, and the model includes a generation network, a plurality of missing area discrimination networks and a global discrimination network. The innovation of the invention is that a plurality of area discrimination networks are added on the basis of generating a antagonism network to form a multi-analysis network.
(3) Training a generation network by taking CT projection data of a training set as an input image, and acquiring a prediction result of a sparse missing part of the image; cutting and extracting the prediction result of the sparse missing part to obtain estimation graphs of a plurality of sparse missing areas;
(4) and training the estimation graphs of the sparse regions and the real values corresponding to the estimation graphs in the complete CT projection data together to a region discrimination network corresponding to the missing region, so that the estimation result of the missing region continuously approaches to the real data value.
(5) And integrating the real projection data image of the non-sparse missing part and the estimation image of the sparse missing region to obtain an integrated sparse missing projection data estimation image. And training the integrated sparse missing projection data estimation graph and the real value of the complete CT projection data together to a global discriminator so that the estimation result of the integral CT projection data approaches to the real value of the complete CT projection data.
(6) The setting of the region discriminator depends on the number of regions with sparse missing CT projection data, and the number of convolution layers, step length and the like of a specific discrimination network are related to the size of the sparse missing region.
(7) A CT image is reconstructed from the generated Projection data image using a convolution Filtered Back-Projection (FBP) method.
The invention aims to provide a multi-discriminant multi-analysis network model, which is mainly characterized in that a certain innovation is carried out on a discriminant network, a plurality of discriminants are used for processing sparse missing projection data images, a one-to-one corresponding relation is kept, and the reconstruction work of various image missing sizes can be met. In medical image reconstruction, for CT projection data images with deficiency in limited scanning angles and deficiency in projection angles in the detector direction, the CT image with higher definition and higher consistency of deficiency boundaries can be reconstructed by the model.
Examples
The embodiment is a missing CT projection data estimation method of a multi-resolution network based on a multi-discriminator, and in practical application, the method comprises the following steps:
(1) the training data is 124 complete CT projection data, and the image size is 720 multiplied by 1024 pixels;
and (3) simulating a CT projection data image under the condition of sparse deletion at the CT limited scanning angle in the medical science, namely, carrying out shielding treatment on the image in the vertical direction, and setting the pixel value of the region to be 0. In this embodiment, 4 occlusion areas are set in the scanning angle, and the appearance on the projection data image is as follows: the initial pixel points of the shielding are respectively [0, 0], [0, 256], [0, 512], [0, 768], the width of the shielding is 128 pixel points, and the height is 720 pixel points. The processed missing CT projection data image is a sparse missing projection data image with 4 missing regions.
(2) The multi-discriminant multi-analysis network model trains 124 images at a time, trains and iterates 250 times, and defines the reconstruction loss as the L2 distance between the projection data estimation result output by the generator and the actual value of the complete projection data:
where x is the true value of the projection data image,
is a binary value: 1 indicates that the pixel value is input, and 0 indicates that the pixel value is discarded, i.e., the pixel value is missing; f (x) is the output of the network encoder section.
Will lose the resistance
Is defined as:
where x is the true projection data image value, D (x) is the output value obtained by inputting the data x into the discriminator D, and G is a parameter function that maps the pixel value from the noise distribution (generated pixel value) z to the data distribution (original projection data image pixel value) χ.
The joint loss function j (x) is a joint function of the reconstruction loss in (1) and the multiple countermeasures losses in (2), and there are 4 countermeasures losses in this example, i is 4:
J(x)＝Lrec+λgLag+λ1La1+λ2La2+...+λiLai(3)
wherein L isrecI.e. the reconstruction loss L described earlierrec(x)，λgIs the specific weight ratio of the antagonistic losses of the global arbiter in the joint losses, LagIs the antagonism loss of the global arbiter. Lambda [ alpha ]1Is the specific gravity ratio, λ, of the antagonistic loss of the region discriminator 1 in the joint loss2Is the specific gravity ratio, λ, of the antagonistic loss of the region discriminator 2 in the joint lossiIs the specific gravity ratio of the antagonistic loss of the region discriminator i in the joint loss, and i is the number of the region discriminators. L isaiIs the antagonism loss of the region discriminator i. The Adam optimizer is used for training an optimization function, and the learning rate is 0.0002;
(3) the output of the generation network is an estimated image of projection data with a size of 720 × 1024 pixels, the estimated parts of 4 missing regions corresponding to the estimated image are cut out as F1, F2, F3 and F4, and the corresponding 4 regions of the full CT projection data true value image are also cut out as R1, R2, R3 and R4.
(4) A plurality of region discriminators are constructed, in this example, 4 region discriminators are D1, D2, D3 and D4, respectively. The structure of the region discriminator needs to be set according to the size of the corresponding missing region block, and the input and output of the 4 region discriminators in this embodiment all use 128 × 720 pixels consistent with the size of the missing region.
(5) Taking the first block missing region projection data estimated image F1 and the first block missing region projection data true value R1 in (3) as the input of the region discriminator D1 in (4), F2 and R2 as the input of the region discriminator D2, F3 and R3 as the input of the region discriminator D3, and F4 and R4 as the input of the region discriminator D4, respectively train the discrimination networks of the corresponding regions.
(6) And (3) taking the complete CT projection data real value image in the step (1) and the CT projection data estimation image containing 4 blocks of area sparse deletions as the input of a global discriminator together to train a global discrimination network.
(7) This iterative training 250 is performed to train the multiple resolution network model for the multiple discriminators.
(8) The test data is 31 sparse missing CT projection data, and the image size is 720 multiplied by 1024 pixels; and generating a sparse missing CT projection data estimation image by using a trained multi-analysis network model of a plurality of discriminators.
(9) Reconstructing a CT image from the image generated in (8) using a convolution Filtered Back-Projection (FBP) method.
The invention is a network model based on multiple discriminators, the number of discriminators depends on the number of blocks of a missing region, and each missing part is trained independently to generate prediction projection data of the missing part; the method can not only meet the requirements of sparse view deletion with equal width or equal height, but also can well estimate the sparse deletion with different area widths and heights in the same projection data deletion region.
The above description is only a preferred embodiment of the present invention, and it should be noted that, for those skilled in the art, several modifications and variations can be made without departing from the technical principle of the present invention, and these modifications and variations should also be regarded as the protection scope of the present invention.
Claims (6)
1. A multi-discriminant-based multi-analysis network missing CT projection data estimation method is characterized by comprising the following steps:
acquiring a sparse missing CT projection data image;
inputting the sparse missing CT projection data image into a multiple analysis network model to generate a complete CT projection data estimation image;
reconstructing the complete CT projection data estimation image by a convolution filtering back projection method to obtain a CT image;
the multiple analysis network model comprises a generation network, a plurality of missing area discrimination networks and a global discrimination network.
2. The method as claimed in claim 1, wherein the method for estimating the multiple analysis network missing CT projection data based on multiple discriminators comprises:
taking the complete CT projection data image set as a training data image set;
performing sparse deletion processing on the training data image set to obtain a training set;
training a generated network through the training set to obtain a prediction result of a sparse missing part of the image;
cutting and extracting the prediction result of the sparse missing part to obtain an estimation image of a sparse missing region;
integrating the real projection data image of the non-sparse missing part and the estimation image of the sparse missing region to obtain an integrated sparse missing projection data estimation image;
training a region discrimination network through a real projection data image corresponding to the sparse missing region and an estimation image of the sparse missing region;
and training a global discriminator through the complete CT projection data image and the integrated sparse missing projection data estimation image to obtain a multiple analysis network model.
3. The method of claim 1, wherein the multi-discriminant-based multi-resolution network-missing CT projection data estimation method,
the reconstruction loss of the multi-resolution network model is as follows:
wherein x is the true projection data image value;
is a binary mask: 1 indicates that the pixel value is input, and 0 indicates that the pixel value is discarded, i.e., the pixel value is missing; f (x) is the output of the network encoder section.
4. The method as claimed in claim 1, wherein the multi-discriminant-based multi-resolution network missing CT projection data estimation method is characterized in that the antagonism loss of the multi-resolution network model is:
where x is the true projection data image value, D (x) is the output value obtained by inputting the data x into the discriminator D, and g (z) is a parameter function.
5. The method of claim 1, wherein the joint loss function of the multi-analysis network model is as follows:
J(x)＝Lrec+λgLag+λ1La1+λ2La2+...λiLai(3)
wherein L isrecI.e. the reconstruction loss L described earlierrec(x)，λgIs the specific weight ratio of the antagonistic losses of the global arbiter in the joint losses, LagIs the antagonism loss of the global arbiter; lambda [ alpha ]1Is the specific gravity ratio, λ, of the antagonistic loss of the region discriminator 1 in the joint loss2Is the specific gravity ratio, λ, of the antagonistic loss of the region discriminator 2 in the joint lossiIs the specific gravity ratio of the antagonistic loss of the region discriminator i in the combined loss, wherein i is the number of the region discriminators; l isaiIs the antagonism loss of the region discriminator i.
6. The multi-discriminant-based multi-analysis network-missing CT projection data estimation method of claim 1, wherein a training optimization function of the multi-analysis network model is an Adam optimizer.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010056809.1A CN111260583A (en) | 2020-01-17 | 2020-01-17 | Multi-discriminant-based multi-analysis network missing CT projection data estimation method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010056809.1A CN111260583A (en) | 2020-01-17 | 2020-01-17 | Multi-discriminant-based multi-analysis network missing CT projection data estimation method |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN111260583A true CN111260583A (en) | 2020-06-09 |
Family
ID=70950751
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010056809.1A Pending CN111260583A (en) | 2020-01-17 | 2020-01-17 | Multi-discriminant-based multi-analysis network missing CT projection data estimation method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111260583A (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112270651A (en) * | 2020-10-15 | 2021-01-26 | 西安工程大学 | Image restoration method for generating countermeasure network based on multi-scale discrimination |
| CN113808106A (en) * | 2021-09-17 | 2021-12-17 | 浙江大学 | Ultra-low dose PET image reconstruction system and method based on deep learning |
| JP2023078074A (en) * | 2021-11-25 | 2023-06-06 | 之江実験室 | Sparse Projection Reconstruction Method Based on Graph Learning |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109146988A (en) * | 2018-06-27 | 2019-01-04 | 南京邮电大学 | Non-fully projection CT image rebuilding method based on VAEGAN |
| CN109903356A (en) * | 2019-05-13 | 2019-06-18 | 南京邮电大学 | Missing CT data for projection estimation method based on the multiple parsing network of depth |
| CN110288671A (en) * | 2019-06-25 | 2019-09-27 | 南京邮电大学 | The low dosage CBCT image rebuilding method of network is generated based on three-dimensional antagonism |
-
2020
- 2020-01-17 CN CN202010056809.1A patent/CN111260583A/en active Pending
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109146988A (en) * | 2018-06-27 | 2019-01-04 | 南京邮电大学 | Non-fully projection CT image rebuilding method based on VAEGAN |
| CN109903356A (en) * | 2019-05-13 | 2019-06-18 | 南京邮电大学 | Missing CT data for projection estimation method based on the multiple parsing network of depth |
| CN110288671A (en) * | 2019-06-25 | 2019-09-27 | 南京邮电大学 | The low dosage CBCT image rebuilding method of network is generated based on three-dimensional antagonism |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112270651A (en) * | 2020-10-15 | 2021-01-26 | 西安工程大学 | Image restoration method for generating countermeasure network based on multi-scale discrimination |
| CN112270651B (en) * | 2020-10-15 | 2023-12-15 | 西安工程大学 | Image restoration method for generating countermeasure network based on multi-scale discrimination |
| CN113808106A (en) * | 2021-09-17 | 2021-12-17 | 浙江大学 | Ultra-low dose PET image reconstruction system and method based on deep learning |
| JP2023078074A (en) * | 2021-11-25 | 2023-06-06 | 之江実験室 | Sparse Projection Reconstruction Method Based on Graph Learning |
| JP7430232B2 (en) | 2021-11-25 | 2024-02-09 | 之江実験室 | Sparse projection reconstruction method based on graph learning |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109146988B (en) | Incomplete projection CT image reconstruction method based on VAEGAN | |
| CN108898642B (en) | Sparse angle CT imaging method based on convolutional neural network | |
| CN111325686B (en) | Low-dose PET three-dimensional reconstruction method based on deep learning | |
| Kuanar et al. | Low dose abdominal CT image reconstruction: an unsupervised learning based approach | |
| CN107481297B (en) | CT image reconstruction method based on convolutional neural network | |
| CN111627082B (en) | PET image reconstruction method based on filtering back projection algorithm and neural network | |
| CN102314698B (en) | Total variation minimization dosage CT (computed tomography) reconstruction method based on Alpha divergence constraint | |
| DE602004011620T2 (en) | METALITY FACTOR CORRECTION IN COMPUTER TOMOGRAPHY | |
| CN111260583A (en) | Multi-discriminant-based multi-analysis network missing CT projection data estimation method | |
| CN112258415B (en) | Chest X-ray film super-resolution and denoising method based on generation countermeasure network | |
| Xie et al. | Deep efficient end-to-end reconstruction (DEER) network for few-view breast CT image reconstruction | |
| CN106909947A (en) | CT image metal artifacts removing method and elimination system based on Mean Shift algorithms | |
| CN101980302A (en) | Projection data recovery-guided nonlocal mean low-dose CT reconstruction method | |
| CN102184559B (en) | Particle filtering-based method of reconstructing static PET (Positron Emission Tomograph) images | |
| CN109741254B (en) | Dictionary training and image super-resolution reconstruction method, system, equipment and storage medium | |
| Peng et al. | A cross-domain metal trace restoring network for reducing X-ray CT metal artifacts | |
| Zhu et al. | Metal artifact reduction for X-ray computed tomography using U-net in image domain | |
| CN103034989A (en) | Low-dosage CBCT (Cone Beam Computed Tomography) image denoising method based on high-quality priori image | |
| Lee et al. | Interior tomography using 1D generalized total variation. Part II: Multiscale implementation | |
| Liang et al. | A self-supervised deep learning network for low-dose CT reconstruction | |
| Maeda et al. | Maximum a posteriori X-ray computed tomography using graph cuts | |
| CN116664710A (en) | CT image metal artifact unsupervised correction method based on transducer | |
| Chang et al. | Deep learning image transformation under radon transform | |
| Kim et al. | CNN-based CT denoising with an accurate image domain noise insertion technique | |
| CN109559283A (en) | Medicine PET image denoising method based on the domain DNST bivariate shrinkage and bilateral non-local mean filtering |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |