CN109472754A - CT image metal artifact removing method based on image repair - Google Patents
CT image metal artifact removing method based on image repair Download PDFInfo
- Publication number
- CN109472754A CN109472754A CN201811310963.6A CN201811310963A CN109472754A CN 109472754 A CN109472754 A CN 109472754A CN 201811310963 A CN201811310963 A CN 201811310963A CN 109472754 A CN109472754 A CN 109472754A
- Authority
- CN
- China
- Prior art keywords
- picture
- image
- metal artifacts
- denoted
- model
- 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
- 239000002184 metal Substances 0.000 title claims abstract description 58
- 229910052751 metal Inorganic materials 0.000 title claims abstract description 58
- 238000000034 method Methods 0.000 title claims abstract description 26
- 230000008439 repair process Effects 0.000 title claims abstract description 12
- 238000012549 training Methods 0.000 claims abstract description 17
- 238000005070 sampling Methods 0.000 claims description 12
- 238000004364 calculation method Methods 0.000 claims description 6
- 238000013527 convolutional neural network Methods 0.000 claims description 6
- 238000011478 gradient descent method Methods 0.000 claims description 6
- 238000005215 recombination Methods 0.000 claims description 6
- 230000006798 recombination Effects 0.000 claims description 6
- 238000012360 testing method Methods 0.000 claims description 3
- 238000013528 artificial neural network Methods 0.000 abstract description 6
- 230000000694 effects Effects 0.000 abstract description 5
- 238000012545 processing Methods 0.000 abstract description 4
- 230000008030 elimination Effects 0.000 abstract description 3
- 238000003379 elimination reaction Methods 0.000 abstract description 3
- 230000006870 function Effects 0.000 description 4
- XEEYBQQBJWHFJM-UHFFFAOYSA-N Iron Chemical compound [Fe] XEEYBQQBJWHFJM-UHFFFAOYSA-N 0.000 description 2
- 238000003672 processing method Methods 0.000 description 2
- 238000005299 abrasion Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000015556 catabolic process Effects 0.000 description 1
- 238000013170 computed tomography imaging Methods 0.000 description 1
- 238000013135 deep learning Methods 0.000 description 1
- 230000007812 deficiency Effects 0.000 description 1
- 238000006731 degradation reaction Methods 0.000 description 1
- 238000002474 experimental method Methods 0.000 description 1
- 229910052742 iron Inorganic materials 0.000 description 1
- 210000004218 nerve net Anatomy 0.000 description 1
- 230000005855 radiation Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T5/00—Image enhancement or restoration
- G06T5/70—Denoising; Smoothing
-
- 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
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Molecular Biology (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- Artificial Intelligence (AREA)
- Evolutionary Computation (AREA)
- General Health & Medical Sciences (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)
Abstract
The invention belongs to computer vision, field of medical image processing, specially a kind of CT image metal artifact removing method based on image repair.This method comprises: training neural network repairs CT image;Automatically generate the label image of no metal artifacts;Training neural network eliminates metal artifacts.The present invention solves the problems, such as that no artifact label image is difficult to obtain, and deep neural network method has been successfully applied in metal artifacts elimination task, better effects can be obtained.
Description
Technical field
The invention belongs to deep learning, computer vision, field of medical image processing is specially a kind of to be based on image repair
CT image metal artifact removing method.
Background technique
In CT imaging, inevitably encounters and there is metallics in patient body, such as metal pontic, brassiere needle, safety pin, iron content makeup
Instrument abrasion piece when product and operation etc..It is a large amount of that the presence of these metal objects generates the image after rebuilding in metallic perimeter
Black ribbon and bright radiation streak artifact, as metal artifacts.These artifacts make picture quality degradation, give doctor
Judgement bring extreme difficulties.
For common CT, generalling use two kinds of means reduces influence of the metal artifacts to scanning effect.First is that using
Different scanning parameter setting such as reduces artifact effects by the way of reducing screw pitch and thin layer scanning to ladder-like artifact;Second is that
Using image processing method, such as different types of metal artifacts are eliminated using image processing algorithm or are reduced with metal artifacts shadow
It rings.Metal artifacts based on image processing method, which eliminate means, substantially three classes: projection interpolation method, iterative method and mixing method.This
A little method prevalent effects are bad, and the expection of people is not achieved.Deep neural network method has huge in image processing tasks
Advantage, but metal artifacts elimination in application it is few, this is primarily due to be difficult to get and metal artifacts image
Correspondingly without artifact label image.It is proposed that a kind of CT image metal artifact removing method based on image repair, energy
It is enough to automatically generate the label image of no metal artifacts under the auxiliary of a small amount of markup information (only mark out artifact region), into
And neural network is trained to eliminate the metal artifacts in CT image.
Summary of the invention
For above-mentioned there are problem or deficiency, the present invention provides a kind of, and the CT image metal artifact based on image repair disappears
Except method.
The technical solution adopted by the present invention is that:
(1) prepare the CT image set A without containing metal artifacts and the CT image set B containing metal artifacts.
(2) on CT image set A, convolutional neural networks, training CT inpainting model are based on.
(3) with trained CT inpainting model in (2), the tally set of CT image set B is generated.
(4) on CT image set B and tally set, convolutional neural networks are based on, artifact model is removed in training.
CT inpainting model training process in the step (2) specifically includes:
(21) defining CT inpainting model is φ, basic structure encoder-decoder, and input picture passes through
Convolutional layer and down-sampling layer in encoder extract high dimensional information, gradually restore to scheme using the up-sampling layer in decoder
The spatial resolution of piece, final output dimension of picture are consistent with input picture.Define the pre-training VGG on ImageNet data set
Model is
(22) picture is taken out from CT image set A as label, is denoted as Igt.It is randomly generated and contains only 0 and 1 binary map
Piece M, 0 to represent current pixel be the hole for needing to repair.Input picture is denoted as Iin, export picture and be denoted as Iout, Iin=Igt
⊙ M, Iout=φ (Iin)。
(23) it is trained model φ using stochastic gradient descent method, using recombination losses function, LholeIt is the L at hole1
Loss, LvalidIt is the L at non-hole1Loss, LperceptualIt is the L of higher dimensional space1Loss, calculation formula are as follows:
Lhole=| | (1-M) ⊙ (Iout-Igt)||1
Lvalid=| | M ⊙ (Iout-Igt)||1
Ltotal=Lvalid+5Lhole+0.1Lperceptual
The process that metal artifacts picture in the step (3) generates label specifically includes:
(31) a picture I is taken out from CT image set B, I is marked manually, obtains containing only 0 and 1 binary map
Piece M, 0 to represent current pixel region be metal artifacts, and 1 to represent current pixel region be not metal artifacts.
(32) input picture obtains output picture by trained model φ in (21).Input picture is denoted as Iin, output
Picture is denoted as Iout, Iin=I ⊙ M, Iout=φ (Iin)。
(33) label for generating the picture I containing metal artifacts, is denoted as Igt, Igt=I ⊙ M+Iout⊙(1-M)。
CT image in the step (4) goes artifact model to specifically include:
(41) defining CT image and removing artifact model is κ, basic structure encoder-decoder, and input picture passes through
Convolutional layer and down-sampling layer in encoder extract high dimensional information, gradually restore picture using the up-sampling layer in decoder
Spatial resolution, final output dimension of picture with input picture it is consistent.Define the pre-training VGG mould on ImageNet data set
Type is
(42) the input picture by the metal artifacts picture I in (33) as network κ, output picture are denoted as Iout, Iout=κ
(I), the picture I that will be generated in (33)gtAs label picture.
(43) it is trained model κ using stochastic gradient descent method, using recombination losses function, LmarIt is to eliminate metal puppet
The L of shadow1Loss, LperceptualIt is the L of higher dimensional space1Loss, calculation formula are as follows:
Lmar=| | (Iout-Igt)||1
Ltotal=Lmar+0.1Lperceptual
(44) the test pictures of metal artifacts will be contained as input, removes artifact model by trained in (43), i.e.,
It can obtain eliminating the picture of metal artifacts.
The beneficial effects of the present invention are:
The present invention proposes a kind of CT image metal artifact removing method based on image repair, can be in a small amount of markup information
Under the auxiliary of (need to only mark out artifact region), the label image of no metal artifacts is automatically generated, and then training neural network is come
Eliminate the metal artifacts in CT image.The present invention solves the problems, such as that no artifact label image is difficult to obtain, by depth nerve net
Network method has been successfully applied in metal artifacts elimination task, achieves preferable effect.
Detailed description of the invention
Fig. 1 is the CT image containing metal artifacts
Fig. 2 is the image after artifact model treatment
Specific embodiment
Below with reference to attached drawing, the present invention will be described in detail.
The invention discloses a kind of CT image metal artifact removing method based on image repair, specific implementation step packet
It includes:
(1) prepare the CT image set A without containing metal artifacts and the CT image set B containing metal artifacts.
(2) on CT image set A, convolutional neural networks, training CT inpainting model are based on.
(3) with trained CT inpainting model in (2), the tally set of CT image set B is generated.
(4) on CT image set B and tally set, convolutional neural networks are based on, artifact model is removed in training.
CT inpainting model training process in the step (2) specifically includes:
(21) defining CT inpainting model is φ, basic structure encoder-decoder, and input picture passes through
Convolutional layer and down-sampling layer in encoder extract high dimensional information, gradually restore to scheme using the up-sampling layer in decoder
The spatial resolution of piece, final output dimension of picture are consistent with input picture.Define the pre-training VGG on ImageNet data set
Model is
(22) picture is taken out from CT image set A as label, is denoted as Igt.It is randomly generated and contains only 0 and 1 binary map
Piece M, 0 to represent current pixel be the hole for needing to repair.Input picture is denoted as Iin, export picture and be denoted as Iout, Iin=Igt
⊙ M, Iout=φ (Iin)。
(23) it is trained model φ using stochastic gradient descent method, using recombination losses function, LholeIt is the L at hole1
Loss, LvalidIt is the L at non-hole1Loss, LperceptualIt is the L of higher dimensional space1Loss, calculation formula are as follows:
Lhole=| | (1-M) ⊙ (Iout-Igt)||1
Lvalid=| | M ⊙ (Iout-Igt)||1
Ltotal=Lvalid+5Lhole+0.1Lperceptual
The process that metal artifacts picture in the step (3) generates label specifically includes:
(31) a picture I is taken out from CT image set B, I is marked manually, obtains containing only 0 and 1 binary map
Piece M, 0 to represent current pixel region be metal artifacts, and 1 to represent current pixel region be not metal artifacts.
(32) input picture obtains output picture by trained model φ in (21).Input picture is denoted as Iin, output
Picture is denoted as Iout, Iin=I ⊙ M, Iout=φ (Iin)。
(33) label for generating the picture I containing metal artifacts, is denoted as Igt, Igt=I ⊙ M+Iout⊙(1-M)。
CT image in the step (4) goes artifact model to specifically include:
(41) defining CT image and removing artifact model is κ, basic structure encoder-decoder, and input picture passes through
Convolutional layer and down-sampling layer in encoder extract high dimensional information, gradually restore picture using the up-sampling layer in decoder
Spatial resolution, final output dimension of picture with input picture it is consistent.Define the pre-training VGG mould on ImageNet data set
Type is
(42) the input picture by the metal artifacts picture I in (33) as network κ, output picture are denoted as Iout, Iout=κ
(I), the picture I that will be generated in (33)gtAs label picture.
(43) it is trained model κ using stochastic gradient descent method, using recombination losses function, LmarIt is to eliminate metal puppet
The L of shadow1Loss, LperceptualIt is the L of higher dimensional space1Loss, calculation formula are as follows:
Lmar=| | (Iout-Igt)||1
Ltotal=Lmar+0.1Lperceptual
(44) the test pictures of metal artifacts will be contained as input, removes artifact model by trained in (43), i.e.,
It can obtain eliminating the picture of metal artifacts.
CT image containing metal artifacts is as shown in Figure 1, by going the image after artifact model treatment as shown in Figure 2.Experiment
The result shows that the present invention can effectively eliminate the metal artifacts in CT image.
Claims (4)
1. a kind of CT image metal artifact removing method based on image repair, which is characterized in that the described method includes:
(1) prepare the CT image set A without containing metal artifacts and the CT image set B containing metal artifacts;
(2) on CT image set A, convolutional neural networks, training CT inpainting model are based on;
(3) with trained CT inpainting model in (2), the tally set of CT image set B is generated;
(4) on CT image set B and tally set, convolutional neural networks are based on, artifact model is removed in training.
2. the method according to claim 1, wherein being specifically included in the step (2):
(21) defining CT inpainting model is φ, basic structure encoder-decoder, and input picture passes through
Convolutional layer and down-sampling layer in encoder extract high dimensional information, gradually restore picture using the up-sampling layer in decoder
Spatial resolution, final output dimension of picture with input picture it is consistent, define ImageNet data set on pre-training VGG mould
Type is
(22) picture is taken out from CT image set A as label, is denoted as Igt, it is randomly generated and contains only 0 and 1 two-value picture M,
0, which represents current pixel, is one and needs the hole repaired;Input picture is denoted as Iin, export picture and be denoted as Iout, Iin=Igt⊙ M,
Iout=φ (Iin);
(23) it is trained model φ using stochastic gradient descent method, using recombination losses function, LholeIt is the L at hole1Damage
It loses, LvalidIt is the L at non-hole1Loss, LperceptualIt is the L of higher dimensional space1Loss, calculation formula are as follows:
3. the method according to claim 1, wherein being specifically included in the step (3):
(31) a picture I is taken out from CT image set B, I is marked manually, obtains containing only 0 and 1 two-value picture M,
0 to represent current pixel region be metal artifacts, and 1 to represent current pixel region be not metal artifacts;
(32) input picture obtains output picture by trained model φ in (21), and input picture is denoted as Iin, export picture
It is denoted as Iout, Iin=I ⊙ M, Iout=φ (Iin);
(33) label for generating the picture I containing metal artifacts, is denoted as Igt, Igt=I ⊙ M+Iout⊙(1-M)。
4. the method according to claim 1, wherein being specifically included in the step (4):
(41) defining CT image and removing artifact model is κ, basic structure encoder-decoder, and input picture passes through
Convolutional layer and down-sampling layer in encoder extract high dimensional information, gradually restore picture using the up-sampling layer in decoder
Spatial resolution, final output dimension of picture with input picture it is consistent, define ImageNet data set on pre-training VGG mould
Type is
(42) the input picture by the metal artifacts picture I in (33) as network κ, output picture are denoted as Iout, Iout=κ (I),
The picture I that will be generated in (33)gtAs label picture;
(43) it is trained model κ using stochastic gradient descent method, using recombination losses function, LmarIt is to eliminate metal artifacts
L1 loss, LperceptualIt is the L1 loss of higher dimensional space, calculation formula is as follows:
Lmar=| | (Iout-Igt)||1
Ltotal=Lmar+0.1Lperceptual
(44) using the test picture for containing metal artifacts as input, artifact model is removed by trained in (43), can be obtained
To the picture for eliminating metal artifacts.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201811310963.6A CN109472754A (en) | 2018-11-06 | 2018-11-06 | CT image metal artifact removing method based on image repair |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201811310963.6A CN109472754A (en) | 2018-11-06 | 2018-11-06 | CT image metal artifact removing method based on image repair |
Publications (1)
Publication Number | Publication Date |
---|---|
CN109472754A true CN109472754A (en) | 2019-03-15 |
Family
ID=65666965
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201811310963.6A Pending CN109472754A (en) | 2018-11-06 | 2018-11-06 | CT image metal artifact removing method based on image repair |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN109472754A (en) |
Cited By (8)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110517194A (en) * | 2019-07-19 | 2019-11-29 | 深圳安科高技术股份有限公司 | A kind of training method of substance decomposition model, substance decomposition method and relevant device |
CN110555834A (en) * | 2019-09-03 | 2019-12-10 | 明峰医疗***股份有限公司 | CT bad channel real-time detection and reconstruction method based on deep learning network |
CN111862258A (en) * | 2020-07-23 | 2020-10-30 | 深圳高性能医疗器械国家研究院有限公司 | Image metal artifact suppression method |
CN112037146A (en) * | 2020-09-02 | 2020-12-04 | 广州海兆印丰信息科技有限公司 | Medical image artifact automatic correction method and device and computer equipment |
CN112017131B (en) * | 2020-10-15 | 2021-02-09 | 南京安科医疗科技有限公司 | CT image metal artifact removing method and device and computer readable storage medium |
KR20210069389A (en) * | 2019-12-03 | 2021-06-11 | 서울대학교산학협력단 | Apparatus and method for removing metal artifact of computer tomography image based on artificail intelligence |
CN114241070A (en) * | 2021-12-01 | 2022-03-25 | 北京长木谷医疗科技有限公司 | Method and device for removing metal artifacts from CT image and training model |
CN114494498A (en) * | 2022-01-28 | 2022-05-13 | 复旦大学 | Metal artifact removing method based on double-domain Fourier neural network |
Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN105046675A (en) * | 2015-08-24 | 2015-11-11 | 重庆邮电大学 | Method for eliminating metal artifacts in KVCT (Kilovolt Computed Tomography) image on the basis of MVCT (Megavolt Computed Tomography) image |
DE102014217966A1 (en) * | 2014-09-09 | 2016-03-10 | Siemens Aktiengesellschaft | Adaptive metal artifact reduction for computed tomography |
CN105701778A (en) * | 2016-01-11 | 2016-06-22 | 赛诺威盛科技(北京)有限公司 | Method of removing metal artifact from CT image |
CN107507148A (en) * | 2017-08-30 | 2017-12-22 | 南方医科大学 | The method that the down-sampled artifact of MRI is removed based on convolutional neural networks |
CN108492269A (en) * | 2018-03-23 | 2018-09-04 | 西安电子科技大学 | Low-dose CT image de-noising method based on gradient canonical convolutional neural networks |
CN108596861A (en) * | 2018-05-14 | 2018-09-28 | 南方医科大学 | A kind of CT image metal artifact minimizing technologies based on the study of depth residual error |
-
2018
- 2018-11-06 CN CN201811310963.6A patent/CN109472754A/en active Pending
Patent Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
DE102014217966A1 (en) * | 2014-09-09 | 2016-03-10 | Siemens Aktiengesellschaft | Adaptive metal artifact reduction for computed tomography |
CN105046675A (en) * | 2015-08-24 | 2015-11-11 | 重庆邮电大学 | Method for eliminating metal artifacts in KVCT (Kilovolt Computed Tomography) image on the basis of MVCT (Megavolt Computed Tomography) image |
CN105701778A (en) * | 2016-01-11 | 2016-06-22 | 赛诺威盛科技(北京)有限公司 | Method of removing metal artifact from CT image |
CN107507148A (en) * | 2017-08-30 | 2017-12-22 | 南方医科大学 | The method that the down-sampled artifact of MRI is removed based on convolutional neural networks |
CN108492269A (en) * | 2018-03-23 | 2018-09-04 | 西安电子科技大学 | Low-dose CT image de-noising method based on gradient canonical convolutional neural networks |
CN108596861A (en) * | 2018-05-14 | 2018-09-28 | 南方医科大学 | A kind of CT image metal artifact minimizing technologies based on the study of depth residual error |
Non-Patent Citations (2)
Title |
---|
A. ŠERIFOVIĆ-TRBALIĆ 等: "Image-based metal artifact reduction in CT images", 《2016 39TH INTERNATIONAL CONVENTION ON INFORMATION AND COMMUNICATION TECHNOLOGY, ELECTRONICS AND MICROELECTRONICS (MIPRO)》 * |
高净植 等: "改进深度残差卷积神经网络的LDCT图像估计", 《计算机工程与应用》 * |
Cited By (13)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110517194A (en) * | 2019-07-19 | 2019-11-29 | 深圳安科高技术股份有限公司 | A kind of training method of substance decomposition model, substance decomposition method and relevant device |
CN110555834A (en) * | 2019-09-03 | 2019-12-10 | 明峰医疗***股份有限公司 | CT bad channel real-time detection and reconstruction method based on deep learning network |
CN110555834B (en) * | 2019-09-03 | 2020-09-22 | 明峰医疗***股份有限公司 | CT bad channel real-time detection and reconstruction method based on deep learning network |
KR20210069389A (en) * | 2019-12-03 | 2021-06-11 | 서울대학교산학협력단 | Apparatus and method for removing metal artifact of computer tomography image based on artificail intelligence |
KR102342954B1 (en) | 2019-12-03 | 2021-12-24 | 서울대학교산학협력단 | Apparatus and method for removing metal artifact of computer tomography image based on artificial intelligence |
CN111862258A (en) * | 2020-07-23 | 2020-10-30 | 深圳高性能医疗器械国家研究院有限公司 | Image metal artifact suppression method |
CN112037146A (en) * | 2020-09-02 | 2020-12-04 | 广州海兆印丰信息科技有限公司 | Medical image artifact automatic correction method and device and computer equipment |
CN112037146B (en) * | 2020-09-02 | 2023-12-22 | 广州海兆印丰信息科技有限公司 | Automatic correction method and device for medical image artifacts and computer equipment |
CN112017131B (en) * | 2020-10-15 | 2021-02-09 | 南京安科医疗科技有限公司 | CT image metal artifact removing method and device and computer readable storage medium |
CN114241070A (en) * | 2021-12-01 | 2022-03-25 | 北京长木谷医疗科技有限公司 | Method and device for removing metal artifacts from CT image and training model |
CN114241070B (en) * | 2021-12-01 | 2022-09-16 | 北京长木谷医疗科技有限公司 | Method and device for removing metal artifacts from CT image and training model |
CN114494498A (en) * | 2022-01-28 | 2022-05-13 | 复旦大学 | Metal artifact removing method based on double-domain Fourier neural network |
CN114494498B (en) * | 2022-01-28 | 2023-04-18 | 复旦大学 | Metal artifact removing method based on double-domain Fourier neural network |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN109472754A (en) | CT image metal artifact removing method based on image repair | |
Hu et al. | Artifact correction in low‐dose dental CT imaging using Wasserstein generative adversarial networks | |
Kim et al. | Jsi-gan: Gan-based joint super-resolution and inverse tone-mapping with pixel-wise task-specific filters for uhd hdr video | |
Jiang et al. | Unsupervised decomposition and correction network for low-light image enhancement | |
CN110021031B (en) | X-ray image enhancement method based on image pyramid | |
CN107481297B (en) | CT image reconstruction method based on convolutional neural network | |
Gooch et al. | Human facial illustrations: Creation and psychophysical evaluation | |
Zheng et al. | Edge heuristic GAN for non-uniform blind deblurring | |
CN104700365B (en) | A kind of method for enhancing picture contrast | |
CN104463777A (en) | Human-face-based real-time depth of field method | |
WO2023202265A1 (en) | Image processing method and apparatus for artifact removal, and device, product and medium | |
CN112819736A (en) | Workpiece character image local detail enhancement fusion method based on multiple exposures | |
CN115769249A (en) | Method for denoising dental images by domain adaptation | |
US20220076459A1 (en) | Image optimization method, apparatus, device and storage medium | |
CN103971349B (en) | Computed tomography images method for reconstructing and ct apparatus | |
Huo et al. | High‐dynamic range image generation from single low‐dynamic range image | |
Yang et al. | Low‐light image enhancement based on Retinex decomposition and adaptive gamma correction | |
TWI559255B (en) | Illusion analysis device, visual illusion adding image generating device, visual illusion analysis method, visual illusion adding image generation method, and program | |
Venkatesh et al. | Restoration of marker occluded hematoxylin and eosin stained whole slide histology images using generative adversarial networks | |
Sun et al. | Structure-guided virtual restoration for defective silk cultural relics | |
CN107424121A (en) | A kind of blurred picture ultra-resolution ratio reconstructing method based on fractional order differential | |
CN110415188A (en) | A kind of HDR image tone mapping method based on Multiscale Morphological | |
Zhang | A selection of image processing techniques: from fundamentals to research front | |
Huang et al. | Quaternion screened Poisson equation for low-light image enhancement | |
CN117197270A (en) | Metal artifact correction method and device for nonlinear projection decomposition and imaging equipment |
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 | ||
WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20190315 |
|
WD01 | Invention patent application deemed withdrawn after publication |