CN111882503A - Image noise reduction method and application thereof - Google Patents
Image noise reduction method and application thereof Download PDFInfo
- Publication number
- CN111882503A CN111882503A CN202010771759.5A CN202010771759A CN111882503A CN 111882503 A CN111882503 A CN 111882503A CN 202010771759 A CN202010771759 A CN 202010771759A CN 111882503 A CN111882503 A CN 111882503A
- Authority
- CN
- China
- Prior art keywords
- image
- noise
- noise reduction
- network
- self
- 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 40
- 230000009467 reduction Effects 0.000 title claims abstract description 29
- 238000013527 convolutional neural network Methods 0.000 claims abstract description 14
- 238000013507 mapping Methods 0.000 claims abstract description 8
- 238000013528 artificial neural network Methods 0.000 claims abstract description 7
- 238000012549 training Methods 0.000 claims abstract description 6
- 238000002595 magnetic resonance imaging Methods 0.000 claims description 22
- 238000005457 optimization Methods 0.000 claims description 8
- 238000010606 normalization Methods 0.000 claims description 7
- 238000005070 sampling Methods 0.000 claims description 6
- 238000012545 processing Methods 0.000 claims description 4
- ORILYTVJVMAKLC-UHFFFAOYSA-N Adamantane Natural products C1C(C2)CC3CC1CC2C3 ORILYTVJVMAKLC-UHFFFAOYSA-N 0.000 claims description 3
- 238000002600 positron emission tomography Methods 0.000 claims description 3
- 238000002603 single-photon emission computed tomography Methods 0.000 claims description 3
- 238000003745 diagnosis Methods 0.000 abstract description 7
- 238000003384 imaging method Methods 0.000 abstract description 5
- 230000006870 function Effects 0.000 description 9
- 238000010586 diagram Methods 0.000 description 4
- 230000009471 action Effects 0.000 description 2
- 210000004556 brain Anatomy 0.000 description 2
- 230000008859 change Effects 0.000 description 2
- 238000001727 in vivo Methods 0.000 description 2
- 230000008569 process Effects 0.000 description 2
- 230000005855 radiation Effects 0.000 description 2
- 238000011084 recovery Methods 0.000 description 2
- 238000010521 absorption reaction Methods 0.000 description 1
- 230000004913 activation Effects 0.000 description 1
- 238000004458 analytical method Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 230000003727 cerebral blood flow Effects 0.000 description 1
- 238000012937 correction Methods 0.000 description 1
- 238000013135 deep learning Methods 0.000 description 1
- 238000002059 diagnostic imaging Methods 0.000 description 1
- 201000010099 disease Diseases 0.000 description 1
- 208000037265 diseases, disorders, signs and symptoms Diseases 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 230000001037 epileptic effect Effects 0.000 description 1
- 230000003993 interaction Effects 0.000 description 1
- 208000015122 neurodegenerative disease Diseases 0.000 description 1
- 210000000056 organ Anatomy 0.000 description 1
- 238000011176 pooling Methods 0.000 description 1
- 238000004321 preservation Methods 0.000 description 1
- 239000000941 radioactive substance Substances 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 238000012216 screening Methods 0.000 description 1
- 230000011218 segmentation Effects 0.000 description 1
- 239000000126 substance Substances 0.000 description 1
- 230000001629 suppression Effects 0.000 description 1
- 230000007704 transition Effects 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
- G06T5/70—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
- G06T2207/10—Image acquisition modality
- G06T2207/10072—Tomographic images
- G06T2207/10088—Magnetic resonance imaging [MRI]
-
- 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]
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)
- Magnetic Resonance Imaging Apparatus (AREA)
- Image Processing (AREA)
Abstract
The application belongs to the technical field of image scanning, and particularly relates to an image noise reduction method and application thereof. Under the existing rapid imaging condition, the reconstructed MRI image has overlarge noise, which affects the diagnosis condition of doctors. The application provides an image noise reduction method, which comprises the following steps: constructing a self-correcting convolutional neural network; taking the L1 norm as a loss function; optimizing the self-correcting convolution neural network; taking the image with noise as the input of a network, taking the image with noise and a corresponding noise-free image as network labels, and training the network to obtain a mapping relation from the image with noise to the noise-free image; and denoising the image to be denoised through a trained network to obtain a denoised image. The problem of unclear images is solved.
Description
Technical Field
The application belongs to the technical field of image scanning, and particularly relates to an image noise reduction method and application thereof.
Background
Magnetic Resonance Imaging (MRI) is a new examination technique which is adopted according to the principle that the nucleus with magnetic distance can generate transition between energy levels under the action of a magnetic field, and the MRI is helpful for examining the energy state and cerebral blood flow condition of the brain of an epileptic and has great value for diagnosing degenerative diseases. MRI is realized by radiating energy to generate signals from substances in vivo to the surrounding environment through the action of an external high-frequency magnetic field, the imaging process is similar to image reconstruction and CT, only MRI does not depend on radiation, absorption and reflection of the outside or gamma radiation of radioactive substances in vivo, but images by utilizing the interaction of an external magnetic field and an object, and the high-energy magnetic field is harmless to human bodies.
Magnetic Resonance Imaging (MRI) scanning is one of the main medical imaging methods for screening and diagnosis of diseases of human organs, and is an effective tool for medical diagnosis. However, noise generated during image acquisition or transmission can impair MRI quality and severely degrade the accuracy of these diagnoses. Noisy low-quality MRI images can affect the accuracy of automated computer analysis, such as classification, segmentation, and registration. Therefore, the research on MRI denoising is of great significance for obtaining high-quality MRI output, and meanwhile has great scientific significance and application prospect in the field of medical diagnosis.
Under the existing rapid imaging condition, the reconstructed MRI image has overlarge noise, which affects the diagnosis condition of doctors.
Disclosure of Invention
1. Technical problem to be solved
Based on the problem that the condition of a doctor is affected due to overlarge noise of a reconstructed MRI image under the specific condition and the existing rapid imaging condition, the application provides an image noise reduction method and application thereof.
2. Technical scheme
In order to achieve the above object, the present application provides an image noise reduction method, comprising the steps of:
step 1: constructing a self-correcting U-net convolutional neural network based on a U-net network;
step 2: mixing L with1Norm as a loss function;
and step 3: optimizing parameters of the self-correcting U-net convolutional neural network to make a loss function converge;
and 4, step 4: taking the image with noise as the input of a network, taking the image with noise and a corresponding noise-free image as network labels, and training the network to obtain a mapping relation from the image with noise to the noise-free image;
and 5: and denoising the image to be denoised through a trained network to obtain a denoised image.
Another embodiment provided by the present application is: the self-correcting convolutional neural network in the step 1 is a coding and decoding network similar to a U-net network.
Another embodiment provided by the present application is: the self-correcting convolutional neural network in the step 1 comprises a self-correcting convolutional layer, and the self-correcting convolutional layer comprises a first channel and a second channel.
Another embodiment provided by the present application is: and performing convolution and batch normalization processing on the data of the first channel.
Another embodiment provided by the present application is: downsampling data of the second channel.
Another embodiment provided by the present application is: the sampling rate is 4.
Another embodiment provided by the present application is: l in said step 21Norm is:
wherein x isiIs the pixel value, y, of the image before noise reductioniThe pixel values of a real image without noise are obtained, and n is the total number of the pixel values.
Another embodiment provided by the present application is: and optimizing the self-correcting convolution neural network by adopting an Adam optimization algorithm during optimization in the step 3.
The application also provides an application of the image noise reduction method, which applies the image noise reduction method of any one of claims 1 to 9 to single photon emission computed tomography, magnetic resonance imaging, low-dose CT image or low-count positron emission tomography.
3. Advantageous effects
Compared with the prior art, the image denoising method provided by the application has the beneficial effects that:
the application provides an image noise reduction method, which is an under-sampling MRI image noise reduction method based on self-correcting convolution.
The image noise reduction method provided by the application is based on noise reduction of the MR image of the human brain. Considering the fact that the MRI technology inherently adds Rician noise to the output image, the MR image is denoised before image capture is further processed, and structure detail information clearer than that of the existing denoising method is obtained, so that the noise suppression and structure preservation capability of the MRI image is demonstrated.
The image noise reduction method provided by the application aims at solving the problem that the reconstructed image is relatively high in noise ratio under certain rapid MRI imaging conditions, and the image is unclear.
According to the image noise reduction method, the self-correcting convolution is used for replacing an original convolution layer, so that the receptive field of the image can be increased under the condition that network parameters are not increased, and the characteristic of the image can be more accurately represented by the characteristic diagram. The receptive field is defined as the area where the convolutional neural network features can see the input image, in other words, the feature output is affected by the pixels in the receptive field area. Therefore, the larger the field of view, the more representative the feature map representing the image features.
The application provides an image noise reduction method, which uses L in a loss function1Norm replacing original L2The norm enables the image to better keep edge details so as to meet the diagnosis requirement of a doctor.
The image denoising method provided by the application improves the image quality.
According to the image denoising method, a self-correcting convolution is used for reducing the noise of the image in the deep learning background, so that the receptive field is increased while the network parameters are not increased, and the image quality of denoising is further improved.
Drawings
FIG. 1 is a schematic diagram of a generator network for the image denoising method of the present application;
FIG. 2 is a schematic diagram of the self-correcting convolution process of the present application;
fig. 3 is a schematic diagram of a self-correcting block of the present application.
Detailed Description
Hereinafter, specific embodiments of the present application will be described in detail with reference to the accompanying drawings, and it will be apparent to those skilled in the art from this detailed description that the present application can be practiced. Features from different embodiments may be combined to yield new embodiments, or certain features may be substituted for certain embodiments to yield yet further preferred embodiments, without departing from the principles of the present application.
The L1 norm is the sum of the absolute values of the individual parameters, follows a Laplace distribution, is not completely differentiable, and is manifested by many corners appearing on the image.
The L2 norm is the sum of the squares of the individual parameters, follows a Gaussian distribution, and is completely differentiable. The corners on the image are much rounded compared to L1.
Referring to fig. 1 to 3, the present application provides an image denoising method, including the steps of:
step 1: constructing a self-correcting U-net convolutional neural network based on a U-net network;
step 2: mixing L with1Norm as a loss function;
and step 3: optimizing parameters of the self-correcting U-net convolutional neural network to make a loss function converge;
and 4, step 4: taking the image with noise as the input of a network, taking the image with noise and a corresponding noise-free image as network labels, and training the network to obtain a mapping relation from the image with noise to the noise-free image;
and 5: and denoising the image to be denoised through a trained network to obtain a denoised image.
And 3, performing step 3 on the parameters in the network constructed in the step 1, namely training the parameters, and continuously performing iterative optimization to change the parameters to achieve a good mapping relation.
Further, the self-correcting convolutional neural network in the step 1 is a coding and decoding network similar to a U-net network.
The self-correcting convolutional neural network is a coding and decoding network similar to a U-net network: the image is subjected to a layer of convolution, a layer of self-correcting convolution, pooling, self-correcting convolution, bilinear interpolation upsampling, self-correcting convolution and convolution. The number of convolution kernels is respectively: 1, 16, 16, 32, 32, 64, 64, 128, 128, 128, 64, 64, 32, 32, 16, 16, 1. the size of the image is: 128 × 128, 64 × 64, 32 × 32, 16 × 16, 32 × 32, 64 × 64, 128 × 128.
And taking the MRI image with noise as the input of the network and the corresponding noise-free MRI image as a network label to train the network.
And training the network to obtain a mapping relation G from the MRI image with noise to the MRI image without noise.
And denoising the MR image with the noise through a trained network to obtain a denoised image which is in line with the diagnosis of a doctor.
Further, the self-correcting convolutional neural network in step 1 comprises a self-correcting convolutional layer, and the self-correcting convolutional layer comprises a first channel and a second channel.
Further, the data of the first channel is subjected to volume and batch normalization processing.
Further, data of the second channel is downsampled.
Further, the sampling rate is 4.
The self-correcting convolution structure of fig. 1 is shown in fig. 2. The number of channels is equally divided into an upper channel, i.e., a second channel, and a lower channel, i.e., a first channel, with respect to input data. And processing the data in the upper layer by a self-correction module, combining the data in the lower channel with the data in the upper layer after convolution and batch normalization, and obtaining the data with the same size as the input data as input.
The self-correcting block network structure in fig. 2 is shown in fig. 3, and the dimension of input data is C/2 × H × W. The data above is down sampled with a sampling rate r, set to 4 in this application. The dimensionality of the sampled data is as follows: c/2 XH/rxW/r, then performing convolution and batch normalization operation, then performing dimension recovery of data by upsampling with the sampling rate r, then adding the dimension recovery data with the input data, and performing Sigmoid activation function on the obtained result.
The result of the convolution and batch normalization operation on the following data is multiplied by the previous result, and the obtained result is subjected to the convolution and batch normalization operation to obtain the final self-correcting convolution result.
Further, L in the step 21Norm is:
wherein x isiIs the pixel value, y, of the image before noise reductioniThe pixel values of a real image without noise are obtained, and n is the total number of the pixel values. In the self-correcting convolution U-net network framework, in order to make image details sharper, detail edges of the denoised image are reserved.
Further, during the optimization in the step 3, an Adam optimization algorithm is adopted to optimize the self-correcting convolutional neural network.
The application also provides an application of the image noise reduction method, which applies the image noise reduction method of any one of claims 1 to 9 to single photon emission computed tomography, magnetic resonance imaging, low-dose CT image or low-count positron emission tomography.
Examples
The method comprises the following steps: building a self-correcting convolution neural network framework, and converting an original convolution module into the self-correcting convolution shown in the figure 2 on the basis of a U-net network;
step two: taking the L1 norm between the noise reduction result obtained by the low-quality magnetic resonance image in the paired data after passing through the network designed in the step one and the high quality as a loss function;
step three: optimizing the loss function in the second step by using an optimizer, and designing parameters in the network in the iterative optimization step I to finally make the loss function in the second step converge;
step four: the method comprises the steps that noisy and noiseless data are given to a network, L1 norms between the data with noise and the noiseless data are obtained through the network by the noisy data, the data with noise is used as a loss function, an optimizer is used for optimizing the data to change parameters in the network designed in the step one, and finally mapping of network parameters in the step one is obtained;
step five: and C, performing the network parameters of the mapping relation trained in the step four on the tested image to finally obtain the noise-reduced image.
Although the present application has been described above with reference to specific embodiments, those skilled in the art will recognize that many changes may be made in the configuration and details of the present application within the principles and scope of the present application. The scope of protection of the application is determined by the appended claims, and all changes that come within the meaning and range of equivalency of the technical features are intended to be embraced therein.
Claims (9)
1. An image noise reduction method, characterized by: the method comprises the following steps:
step 1: constructing a self-correcting convolutional neural network;
step 2: mixing L with1Norm as a loss function;
and step 3: optimizing the self-correcting convolution neural network to make the loss function converge;
and 4, step 4: taking the image with noise as the input of a network, taking the image with noise and a corresponding noise-free image as network labels, and training the network to obtain a mapping relation from the image with noise to the noise-free image;
and 5: and denoising the image to be denoised through a trained network to obtain a denoised image.
2. An image noise reduction method according to claim 1, characterized by: the self-correcting convolution neural network in the step 1 is a self-correcting U-net convolution neural network based on a U-net network.
3. An image noise reduction method according to claim 1, characterized by: the self-correcting convolutional neural network in the step 1 comprises a self-correcting convolutional layer, and the self-correcting convolutional layer comprises a first channel and a second channel.
4. An image noise reduction method according to claim 3, characterized in that: and performing convolution and batch normalization processing on the data of the first channel.
5. An image noise reduction method according to claim 3, characterized in that: downsampling data of the second channel.
6. An image noise reduction method according to claim 5, wherein: the sampling rate is 4.
7. An image noise reduction method according to claim 1, characterized by: l in said step 21Norm is:
wherein x isiIs the pixel value, y, of the image before noise reductioniThe pixel values of a real image without noise are obtained, and n is the total number of the pixel values.
8. An image noise reduction method according to claim 1, characterized by: and optimizing the self-correcting convolution neural network by adopting an Adam optimization algorithm during optimization in the step 3.
9. An application of an image noise reduction method, characterized by: the image noise reduction method of any one of claims 1 to 8 is applied to single photon emission computed tomography, magnetic resonance imaging, low dose CT images, or low count positron emission tomography.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010771759.5A CN111882503A (en) | 2020-08-04 | 2020-08-04 | Image noise reduction method and application thereof |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010771759.5A CN111882503A (en) | 2020-08-04 | 2020-08-04 | Image noise reduction method and application thereof |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN111882503A true CN111882503A (en) | 2020-11-03 |
Family
ID=73210208
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010771759.5A Pending CN111882503A (en) | 2020-08-04 | 2020-08-04 | Image noise reduction method and application thereof |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111882503A (en) |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112509093A (en) * | 2020-12-17 | 2021-03-16 | 深圳高性能医疗器械国家研究院有限公司 | Image attenuation correction method and application thereof |
| CN112598597A (en) * | 2020-12-25 | 2021-04-02 | 华为技术有限公司 | Training method of noise reduction model and related device |
| CN114219820A (en) * | 2021-12-08 | 2022-03-22 | 苏州工业园区智在天下科技有限公司 | Neural network generation method, denoising method and device |
| CN114757847A (en) * | 2022-04-24 | 2022-07-15 | 汕头市超声仪器研究所股份有限公司 | Multi-information extraction extended U-Net and application method thereof in low-dose X-ray imaging |
| CN115409743A (en) * | 2022-11-03 | 2022-11-29 | 长春理工大学 | Model construction method for brain magnetic resonance image processing based on deep learning |
| US11540798B2 (en) | 2019-08-30 | 2023-01-03 | The Research Foundation For The State University Of New York | Dilated convolutional neural network system and method for positron emission tomography (PET) image denoising |
-
2020
- 2020-08-04 CN CN202010771759.5A patent/CN111882503A/en active Pending
Cited By (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US11540798B2 (en) | 2019-08-30 | 2023-01-03 | The Research Foundation For The State University Of New York | Dilated convolutional neural network system and method for positron emission tomography (PET) image denoising |
| CN112509093A (en) * | 2020-12-17 | 2021-03-16 | 深圳高性能医疗器械国家研究院有限公司 | Image attenuation correction method and application thereof |
| CN112598597A (en) * | 2020-12-25 | 2021-04-02 | 华为技术有限公司 | Training method of noise reduction model and related device |
| WO2022134971A1 (en) * | 2020-12-25 | 2022-06-30 | 华为技术有限公司 | Noise reduction model training method and related apparatus |
| CN114219820A (en) * | 2021-12-08 | 2022-03-22 | 苏州工业园区智在天下科技有限公司 | Neural network generation method, denoising method and device |
| CN114757847A (en) * | 2022-04-24 | 2022-07-15 | 汕头市超声仪器研究所股份有限公司 | Multi-information extraction extended U-Net and application method thereof in low-dose X-ray imaging |
| CN115409743A (en) * | 2022-11-03 | 2022-11-29 | 长春理工大学 | Model construction method for brain magnetic resonance image processing based on deep learning |
| CN115409743B (en) * | 2022-11-03 | 2023-03-24 | 长春理工大学 | Model construction method for brain magnetic resonance image processing based on deep learning |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111882503A (en) | Image noise reduction method and application thereof | |
| Park et al. | Unpaired image denoising using a generative adversarial network in X-ray CT | |
| Hu et al. | Artifact correction in low‐dose dental CT imaging using Wasserstein generative adversarial networks | |
| Wang et al. | Advances in data preprocessing for biomedical data fusion: An overview of the methods, challenges, and prospects | |
| 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 | |
| Luo et al. | Adaptive rectification based adversarial network with spectrum constraint for high-quality PET image synthesis | |
| US9002134B2 (en) | Multi-scale image normalization and enhancement | |
| CN112258415B (en) | Chest X-ray film super-resolution and denoising method based on generation countermeasure network | |
| CN111311704A (en) | Image reconstruction method and device, computer equipment and storage medium | |
| Li et al. | A multiscale double-branch residual attention network for anatomical–functional medical image fusion | |
| CN106530236A (en) | Medical image processing method and system | |
| WO2022027216A1 (en) | Image denoising method and application thereof | |
| Chen et al. | Robust restoration of low-dose cerebral perfusion CT images using NCS-Unet | |
| CN115018728A (en) | Image fusion method and system based on multi-scale transformation and convolution sparse representation | |
| Xia et al. | Deep residual neural network based image enhancement algorithm for low dose CT images | |
| Zhu et al. | STEDNet: Swin transformer‐based encoder–decoder network for noise reduction in low‐dose CT | |
| Li et al. | Deep attention super-resolution of brain magnetic resonance images acquired under clinical protocols | |
| Liu et al. | BTMF-GAN: A multi-modal MRI fusion generative adversarial network for brain tumors | |
| Zhang et al. | Medical image fusion based on low-level features | |
| CN111860836A (en) | Self-supervision learning method and application | |
| CN113436118B (en) | Low-dose CT image restoration method based on multi-scale convolutional coding network | |
| CN116245969A (en) | Low-dose PET image reconstruction method based on deep neural network | |
| Al-Ameen et al. | Prevalent degradations and processing challenges of computed tomography medical images: A compendious analysis | |
| Mahmoud et al. | Variant Wasserstein Generative Adversarial Network Applied on Low Dose CT Image Denoising. |
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 | ||
| TA01 | Transfer of patent application right | ||
| TA01 | Transfer of patent application right |
Effective date of registration: 20230630 Address after: 381, West Side, Phase 1, Huiheng Building, No. 138 Gaoxin South Seventh Road, Gaoxin District Community, Yuehai Street, Shenzhen, Guangdong Province, 518000 Applicant after: Guochuang Yucheng Medical Device Development (Shenzhen) Co.,Ltd. Address before: Room A101, building 1, Yinxing Zhijie phase II, No. 1301-76, sightseeing Road, Xinlan community, Guanlan street, Longhua District, Shenzhen, Guangdong 518000 Applicant before: Shenzhen National Research Institute of high performance Medical Devices Co.,Ltd. |