CN108230277B - Dual-energy CT image decomposition method based on convolutional neural network - Google Patents
Dual-energy CT image decomposition method based on convolutional neural network Download PDFInfo
- Publication number
- CN108230277B CN108230277B CN201810136595.1A CN201810136595A CN108230277B CN 108230277 B CN108230277 B CN 108230277B CN 201810136595 A CN201810136595 A CN 201810136595A CN 108230277 B CN108230277 B CN 108230277B
- Authority
- CN
- China
- Prior art keywords
- energy
- image
- neural network
- convolutional neural
- dual
- 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.)
- Active
Links
- 238000013527 convolutional neural network Methods 0.000 title claims abstract description 69
- 238000000354 decomposition reaction Methods 0.000 title claims abstract description 45
- 238000000034 method Methods 0.000 title claims abstract description 41
- 238000012549 training Methods 0.000 claims abstract description 14
- 238000003062 neural network model Methods 0.000 claims abstract description 6
- 238000013507 mapping Methods 0.000 claims abstract description 4
- 230000009977 dual effect Effects 0.000 claims abstract description 3
- 238000000605 extraction Methods 0.000 claims description 3
- 238000012937 correction Methods 0.000 claims description 2
- 208000037170 Delayed Emergence from Anesthesia Diseases 0.000 claims 1
- 238000012545 processing Methods 0.000 abstract description 3
- 238000004422 calculation algorithm Methods 0.000 description 18
- 238000001228 spectrum Methods 0.000 description 9
- 210000001519 tissue Anatomy 0.000 description 6
- 210000000988 bone and bone Anatomy 0.000 description 5
- 238000013170 computed tomography imaging Methods 0.000 description 5
- 238000005516 engineering process Methods 0.000 description 4
- 238000012805 post-processing Methods 0.000 description 4
- 238000004088 simulation Methods 0.000 description 4
- 238000013461 design Methods 0.000 description 3
- 238000012360 testing method Methods 0.000 description 3
- 230000009286 beneficial effect Effects 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 230000006870 function Effects 0.000 description 2
- 238000003384 imaging method Methods 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000007781 pre-processing Methods 0.000 description 2
- 230000003595 spectral effect Effects 0.000 description 2
- 238000013528 artificial neural network Methods 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 230000001419 dependent effect Effects 0.000 description 1
- 238000002059 diagnostic imaging Methods 0.000 description 1
- 238000007689 inspection Methods 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 238000009659 non-destructive testing Methods 0.000 description 1
- 238000010606 normalization Methods 0.000 description 1
- 238000011160 research 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
- G06T5/80—Geometric correction
-
- 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]
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- General Health & Medical Sciences (AREA)
- General Engineering & Computer Science (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- Data Mining & Analysis (AREA)
- Evolutionary Computation (AREA)
- Artificial Intelligence (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- Biomedical Technology (AREA)
- Life Sciences & Earth Sciences (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Health & Medical Sciences (AREA)
- Image Processing (AREA)
- Apparatus For Radiation Diagnosis (AREA)
- Image Analysis (AREA)
Abstract
The invention relates to the technical field of medical dual-energy image decomposition and image processing, in particular to a dual-energy CT image decomposition method, and particularly relates to a dual-energy CT image decomposition method based on a convolutional neural network. A dual-energy CT image decomposition method based on a convolutional neural network comprises the following steps: designing a convolutional neural network model as a mapping function D (mu) in a dual energy decomposition modelH,L(ii) a Θ); training the convolutional neural network through a convolutional neural network model and a training data set, and effectively estimating parameters theta of the convolutional neural network; and (3) carrying out base material efficient decomposition on the dual-energy CT image by using the trained convolutional neural network and the convolutional neural network parameter theta obtained in the step (2). According to the method, through the establishment of a double-input and double-output convolution neural network model and cross convolution, reasonable shunting of different base materials in the high-energy CT image and the low-energy CT image is realized, so that the quality of base material decomposition of the double-energy CT image is effectively improved.
Description
Technical Field
The invention relates to the technical field of medical dual-energy image decomposition and image processing, in particular to a dual-energy CT image decomposition method, and particularly relates to a dual-energy CT image decomposition method based on a convolutional neural network.
Background
Dual-energy CT image reconstruction has been increasingly applied to the fields of medical imaging, security inspection, nondestructive testing, and the like, and compared with the conventional single-energy-spectrum CT imaging technology, dual-energy CT can recognize different material materials by using image attenuation information under different energy spectrums. The dual-energy CT technology breaks through the physical limitation of the traditional single-energy spectrum CT and becomes a hotspot and difficult problem of the research in the CT imaging field.
The core theory of the dual-energy CT imaging technology is the dual-energy CT image reconstruction algorithm, in which the decoupling of material and energy cross information is one of the key issues. Compared with the traditional CT image reconstruction theory, the difficulty of the dual-energy CT image reconstruction algorithm lies in the characteristics of nonlinearity, multiple resolvability, inadaptability, high dimensionality and the like of the problem. Due to different application requirements of dual-energy CT imaging, different imaging modes and different acquired data information, corresponding dual-energy CT image reconstruction algorithms are different. At present, the dual-energy CT image reconstruction algorithm can be mainly classified into three categories: an iterative class direct reconstruction algorithm, an image reconstruction algorithm based on projection domain preprocessing, and an image reconstruction algorithm based on image domain post-processing.
The iterative dual-energy CT image reconstruction algorithm is suitable for geometrically consistent or inconsistent high and low energy projection sets, and the reconstructed image has high signal-to-noise ratio. However, these methods usually require huge computation overhead and are slow in computation speed, which seriously reduces the practicability of the algorithm. The image reconstruction algorithm based on projection domain preprocessing fully utilizes high-energy and low-energy spectral information and a multi-color projection generation model, and converts a nonlinear solving problem into a linear solving problem, so that the dual-energy CT image reconstruction is realized by using a conventional CT image reconstruction algorithm, the influence of hardening artifacts can be effectively eliminated theoretically, accurate physical parameter distribution information is obtained, the calculation is simple and convenient, the efficiency is high, and the method is a mainstream reconstruction method of the current dual-energy CT image reconstruction technology. However, such methods are highly dependent on the calibration process and require that the high and low energy projection data sets be geometrically consistent in space, i.e., each pair of high and low energy projection measurements need to follow the same ray path.
The image reconstruction algorithm based on image domain post-processing firstly reconstructs high and low energy CT images from high and low energy projection data by using a traditional CT image reconstruction algorithm, and carries out material decomposition on the high and low energy CT reconstructed images in an image domain to obtain physical parameter distribution images of object faults. The realization of the decomposition of different materials under high and low energy images is the key of image reconstruction algorithms based on image post-processing. The dual-energy CT image reconstruction algorithm based on image domain post-processing has low requirements on the space geometric consistency of high-energy projection data and low-energy projection data, is simple and convenient to calculate, and can inhibit hardening artifacts to a certain extent. In addition, the method can be directly applied to the existing imaging system, no additional hardware equipment is needed, and the cost is saved. Therefore, the method is widely applied to the existing dual-energy CT imaging system. Based on the method, the patent designs a dual-energy CT image decomposition method based on a convolutional neural network.
Disclosure of Invention
Aiming at the problems that a base material image obtained by the existing dual-energy CT image decomposition method contains a large amount of noise and is low in signal-to-noise ratio, the invention provides the dual-energy CT image decomposition method based on the convolutional neural network, and reasonable shunting of different base materials in a high-energy CT image and a low-energy CT image is realized through establishment of a dual-input and dual-output convolutional neural network model and cross convolution, so that the quality of base material decomposition of the dual-energy CT image is effectively improved.
In order to achieve the purpose, the invention adopts the following technical scheme:
a dual-energy CT image decomposition method based on a convolutional neural network comprises the following steps:
step 1: designing a convolutional neural network model as a mapping function D (mu) in a dual energy decomposition modelH,L；Θ)；
Step 2: training the convolutional neural network through a convolutional neural network model and a training data set, and effectively estimating parameters theta of the convolutional neural network;
and step 3: and (3) carrying out base material efficient decomposition on the dual-energy CT image by using the trained convolutional neural network and the convolutional neural network parameter theta obtained in the step (2).
Preferably, the convolutional neural network model is designed into a dual-input and dual-output network structure model so as to realize direct input of the high-energy CT image and the low-energy CT image and direct output of the images of different materials.
Preferably, the network structure model with double input and double output establishes cross convolution to realize reasonable distribution of different base material information in the high-energy CT image and the low-energy CT image.
Preferably, short links are established in the convolutional neural network model.
Preferably, the training data set includes input data and output data of a convolutional neural network model, the output data is a basis material image, the input data is a dual-energy CT image obtained according to the basis material image and corresponding energy information, and the dual-energy CT image includes a high-energy CT image and a low-energy CT image; the output data is taken as tag data.
Compared with the prior art, the invention has the following beneficial effects:
1. the invention decomposes the dual-energy CT image based on the dual-input and dual-output convolution neural network model, and can effectively avoid the information crosstalk of different energies and different material information at the input end and the output end.
2. The cross convolution-based convolutional neural network model can reasonably shunt different material information in the high-energy CT image and the low-energy CT image and reach different output ends along the cross network structure.
3. Based on the idea of network residual error design, the method establishes the short link in the convolutional neural network model, can effectively improve the training efficiency of the convolutional neural network, and is beneficial to the design of a subsequent deeper network.
Drawings
Fig. 1 is a schematic diagram of a basic flow of a dual-energy CT image decomposition method based on a convolutional neural network according to the present invention.
FIG. 2 is a dual-input and dual-output network structure model of the dual-energy CT image decomposition method based on the convolutional neural network.
FIG. 3 is a dual-input and dual-output cross network structure model of the dual-energy CT image decomposition method based on the convolutional neural network.
FIG. 4 is a dual-energy CT image convolution neural network model of the dual-energy CT image decomposition method based on the convolution neural network of the present invention;
FIG. 5 is a digital simulation test phantom filled with bone and tissue materials for a dual-energy CT image decomposition method based on a convolutional neural network according to the present invention.
FIG. 6 is an X-ray high and low energy spectrum information diagram of SpekCalc software simulation of the dual-energy CT image decomposition method based on the convolutional neural network.
Fig. 7 is a high-energy CT image generated using the test phantom of fig. 5 and the energy spectrum information of fig. 6 according to a dual-energy CT image decomposition method based on a convolutional neural network of the present invention.
FIG. 8 is a low-energy CT image generated using the test phantom of FIG. 5 and the energy spectral information of FIG. 6 according to a dual-energy CT image decomposition method based on a convolutional neural network of the present invention.
FIG. 9 is a bone image obtained by using simulation data according to the dual-energy CT image decomposition method based on the convolutional neural network.
FIG. 10 is a tissue image obtained from simulation data by the dual-energy CT image decomposition method based on the convolutional neural network.
FIG. 11 is an actual high-energy CT image of the dual-energy CT image decomposition method based on the convolutional neural network.
FIG. 12 is an actual low-energy CT image of the dual-energy CT image decomposition method based on the convolutional neural network.
FIG. 13 is a bone image obtained from actual data by the dual-energy CT image decomposition method based on the convolutional neural network.
FIG. 14 is a tissue image obtained from actual data by the dual-energy CT image decomposition method based on the convolutional neural network.
Detailed Description
For the sake of understanding, some terms appearing in the detailed description of the invention are explained below:
BP algorithm: and (4) an error back propagation algorithm.
The invention is further illustrated by the following examples in conjunction with the accompanying drawings:
the first embodiment is as follows:
as shown in fig. 1, the dual-energy CT image decomposition method based on the convolutional neural network of the present invention includes the following steps:
step S101: designing a double-input and double-output convolution neural network model as a mapping function D (mu) in a double-energy decomposition modelH,L(ii) a Θ) in which μH,LThe convolutional neural network model is designed into a dual-input and dual-output network structure model for a dual-energy CT image, as shown in FIG. 2, A, B in the graph is an input high-energy CT image and a low-energy CT image, M1 and M2 in the graph are output basis material 1 and basis material 2, "→" represents two different network structures, and cross convolution is established in the dual-input and dual-output network structure model, as shown in FIG. 3, short links are established in the convolutional neural network model, as shown in FIG. 4, wherein the convolutional neural network model comprises twelve convolutional layers, except for the last output convolution (two corresponding to two label Data), each convolution is followed by a Batch Normalization BN and a linear correction unit (RECTID linear unit), each block in FIG. 4 represents a combination of convolution, BN and ReLU, Data A, Data _ 4 _ A, Label B sequentially represents a combination of an input low-energy CT image, an input high-energy CT image, an output image, a 1-dimensional image, a ReLU, and a total convolutional neural network structure model, and Data 635 _ 2 _ 8 _ 9 _ 3 is a convolutional layer, wherein the last convolutional layer is composed of a convolutional layer, which is composed of three layers, and a convolutional layer of a convolutional layer, and a residual feature extraction layer is composed of a convolutional layer, and a convolutional layer, wherein the last convolutional layer is composed of three layers of a convolutional layer, and a convolutional layer is composed of a convolutional layer.
Step S102: training the convolutional neural network through a convolutional neural network model and a training data set, and effectively estimating parameters theta of the convolutional neural network; as an implementable manner, the convolutional neural network is trained by combining the training data set and the designed convolutional neural network model through the BP algorithm, so that the convolutional neural network parameter Θ is effectively estimated to obtain the trained convolutional neural network, and in the process of supervised training, the initial learning rate, the step length and the weight value are respectively set to be 10e-6, 0.95 and 0.0005. The training data set comprises input data and output data of a convolutional neural network model, the output data is a base material image, the input data is a dual-energy CT image obtained according to a decomposed base material image and corresponding energy information, and the dual-energy CT image comprises a high-energy CT image and a low-energy CT image; the output data is taken as tag data. As an implementable mode, a high-energy CT image and a low-energy CT image of a human body are obtained by using a dual-energy CT device, the high-energy CT image and the low-energy CT image of the human body are segmented according to clinical experience of a doctor, a skeleton image (base material 1) and a tissue image (base material 2) under different energies are obtained and serve as output data, namely label data, high-energy spectrum information and low-energy spectrum information of X rays are obtained by using SpekCalc software, a projection is generated according to the high-energy spectrum information and the low-energy spectrum information, the high-energy CT image and the low-energy CT image are reconstructed to obtain the high-energy CT image and the low-energy CT image. As an implementation, 1300 pairs of 512 × 512 pixel bone and tissue images are obtained, and 1300 pairs of 512 × 512 pixel high and low energy CT images are generated. The bone image and the tissue image of which the size is 512 multiplied by 512 pixels is divided by 1300 with the step size of 48, so that 105300 pair of image blocks of which the size is 128 multiplied by 128 pixels is obtained and used as output data of a convolutional neural network model, namely label data; 1300 pairs of high-energy and low-energy CT images with 512 × 512 pixels are segmented with the step size of 48, and 105300 pairs of image blocks with the size of 128 × 128 pixels are obtained as input data of the convolutional neural network model.
Step S103: and performing base material efficient decomposition on the dual-energy CT image by using the trained convolutional neural network and the convolutional neural network parameter theta obtained in the step S102.
As an implementation mode, the high-energy CT image and the low-energy CT image shown in fig. 7 and 8 are generated by simulation using fig. 5 and 6, and the output decomposition image results obtained by processing fig. 7 and 8 using the trained convolutional neural network are shown in fig. 9 and 10.
As an implementation manner, the high-energy CT image and the low-energy CT image shown in fig. 11 and 12 are respectively shown in fig. 13 and 14 by using the network decomposition result obtained by the method of the present embodiment.
The above shows only the preferred embodiments of the present invention, and it should be noted that it is obvious to those skilled in the art that various modifications and improvements can be made without departing from the principle of the present invention, and these modifications and improvements should also be considered as the protection scope of the present invention.
Claims (3)
1. A dual-energy CT image decomposition method based on a convolutional neural network is characterized by comprising the following steps:
step 1: designing a convolutional neural network model as a mapping function D (mu) in a dual energy decomposition modelH,L(ii) a Θ) in which μH,LThe convolution neural network model comprises twelve convolution layers, except for the final output convolution, each convolution layer is followed by a batch and linear correction unit, the whole convolution neural network model is divided into three layers in total, wherein the first layer is a characteristic extraction layer, the second layer is a shunt layer, the third layer is an output layer, the characteristic extraction layer consists of convolutions of × 38764 of 1 × × 7, a 64-dimensional characteristic image is output, the shunt layer consists of 4 cross convolutions and 2 residual blocks, the convolution dimensionalities are 64 × 3893 × 64, the output layer consists of convolutions of 64 × 865 5 ×, and the 64-dimensional characteristic is combined into an output image;
step 2: training the convolutional neural network through a convolutional neural network model and a training data set, and effectively estimating parameters theta of the convolutional neural network;
and step 3: and (3) carrying out base material efficient decomposition on the dual-energy CT image by using the trained convolutional neural network and the convolutional neural network parameter theta obtained in the step (2).
2. The dual-energy CT image decomposition method based on the convolutional neural network as claimed in claim 1, wherein a short link is established in the convolutional neural network model.
3. The dual-energy CT image decomposition method based on the convolutional neural network as claimed in claim 1, wherein the training data set comprises input data and output data of the convolutional neural network model, the output data is a basis material image, the input data is a dual-energy CT image obtained according to the basis material image and corresponding energy information, and the dual-energy CT image comprises a high-energy CT image and a low-energy CT image; the output data is taken as tag data.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810136595.1A CN108230277B (en) | 2018-02-09 | 2018-02-09 | Dual-energy CT image decomposition method based on convolutional neural network |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810136595.1A CN108230277B (en) | 2018-02-09 | 2018-02-09 | Dual-energy CT image decomposition method based on convolutional neural network |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN108230277A CN108230277A (en) | 2018-06-29 |
| CN108230277B true CN108230277B (en) | 2020-09-11 |
Family
ID=62661486
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810136595.1A Active CN108230277B (en) | 2018-02-09 | 2018-02-09 | Dual-energy CT image decomposition method based on convolutional neural network |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108230277B (en) |
Families Citing this family (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20210232894A1 (en) * | 2018-10-10 | 2021-07-29 | Leapmind Inc. | Neural network processing apparatus, neural network processing method, and neural network processing program |
| CN109523024A (en) * | 2018-11-22 | 2019-03-26 | 天津大学 | Energy spectrum correction method towards medical X-ray detector |
| CN109727238A (en) * | 2018-12-27 | 2019-05-07 | 贵阳朗玛信息技术股份有限公司 | The recognition methods of x-ray chest radiograph and device |
| CN109916933B (en) * | 2019-01-04 | 2021-10-01 | 中国人民解放军战略支援部队信息工程大学 | X-ray computed tomography energy spectrum estimation method based on convolutional neural network |
| CN110189389B (en) * | 2019-05-31 | 2022-08-05 | 首都师范大学 | Deep learning-based decomposition method and device for base material of dual-energy-spectrum CT projection domain |
| CN110428395B (en) * | 2019-06-20 | 2021-10-08 | 浙江大学 | Multi-material decomposition method of single-energy spectrum CT image |
| CN110706299B (en) * | 2019-09-16 | 2023-04-07 | 天津大学 | Substance decomposition imaging method for dual-energy CT |
| CN110675467B (en) * | 2019-09-30 | 2023-05-16 | 中国科学院深圳先进技术研究院 | Material decomposition method and device for energy spectrum CT, storage medium and electronic equipment |
| CN111179228A (en) * | 2019-12-16 | 2020-05-19 | 浙江大学 | Single-energy CT energy spectrum imaging method based on deep learning |
| CN111192224B (en) * | 2020-01-13 | 2024-03-19 | 北京联合大学 | Image enhancement method, image enhancement device, electronic equipment and computer readable storage medium |
| US11967015B2 (en) | 2020-02-06 | 2024-04-23 | Apple Inc. | Neural rendering |
| EP4085427A2 (en) * | 2020-02-06 | 2022-11-09 | Apple Inc. | Neural rendering |
Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101647706A (en) * | 2008-08-13 | 2010-02-17 | 清华大学 | Image reconstruction method for high-energy double-energy CT system |
-
2018
- 2018-02-09 CN CN201810136595.1A patent/CN108230277B/en active Active
Patent Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101647706A (en) * | 2008-08-13 | 2010-02-17 | 清华大学 | Image reconstruction method for high-energy double-energy CT system |
Non-Patent Citations (4)
| Title |
|---|
| Material depth reconstruction method of multi-energy X-ray images using neural network;Woo-Jin Lee et al.;《2012 Annual International Conference of the IEEE Engineering in Medicine and Biology Society》;20121231;1514-1516 * |
| Woo-Jin Lee et al..Material depth reconstruction method of multi-energy X-ray images using neural network.《2012 Annual International Conference of the IEEE Engineering in Medicine and Biology Society》.2012,1514-1516. * |
| 卷积神经网络研究综述;李彦冬等;《计算机应用》;20160910;第36卷(第9期);2511 * |
| 并行交叉的深度卷积神经网络模型;汤鹏杰等;《中国图象图形学报》;20160316;第21卷(第3期);342-343 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN108230277A (en) | 2018-06-29 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108230277B (en) | Dual-energy CT image decomposition method based on convolutional neural network | |
| Zhang et al. | Image domain dual material decomposition for dual‐energy CT using butterfly network | |
| Liu et al. | Deep learning with noise‐to‐noise training for denoising in SPECT myocardial perfusion imaging | |
| CN108961237A (en) | A kind of low-dose CT picture breakdown method based on convolutional neural networks | |
| CN110189389B (en) | Deep learning-based decomposition method and device for base material of dual-energy-spectrum CT projection domain | |
| CN108010098A (en) | A kind of dual intensity composes CT sill image iterative reconstruction methods | |
| US20180182129A1 (en) | Spectral ct image reconstructing method and spectral ct imaging system | |
| CN110175957B (en) | Multi-energy CT-based material decomposition method | |
| CN104408758A (en) | Low-dose processing method of energy spectrum CT image | |
| WO2019067524A1 (en) | Monochromatic ct image reconstruction from current-integrating data via machine learning | |
| KR102182048B1 (en) | System and method for scatter correction of x-ray image | |
| CN110728727A (en) | Low-dose energy spectrum CT projection data recovery method | |
| Zhang et al. | Constrained total generalized p-variation minimization for few-view X-ray computed tomography image reconstruction | |
| Shao et al. | SPECTnet: a deep learning neural network for SPECT image reconstruction | |
| Zhao et al. | Robust beam hardening artifacts reduction for computed tomography using spectrum modeling | |
| CN109916933B (en) | X-ray computed tomography energy spectrum estimation method based on convolutional neural network | |
| Clark et al. | Deep learning based spectral extrapolation for dual‐source, dual‐energy x‐ray computed tomography | |
| Kim et al. | Sparsier2Sparse: Self‐supervised convolutional neural network‐based streak artifacts reduction in sparse‐view CT images | |
| Zhou et al. | The synthesis of high-energy CT images from low-energy CT images using an improved cycle generative adversarial network | |
| CN110706299A (en) | Substance decomposition imaging method for dual-energy CT | |
| CN111145281B (en) | Dual-energy CT direct iteration base material decomposition image reconstruction method | |
| Zhang et al. | A locally weighted linear regression look-up table-based iterative reconstruction method for dual spectral CT | |
| Cui et al. | Prior Knowledge-guided Triple-Domain Transformer-GAN for Direct PET Reconstruction from Low-Count Sinograms | |
| CN116630738A (en) | Energy spectrum CT imaging method based on depth convolution sparse representation reconstruction network | |
| Liang et al. | A model-based unsupervised deep learning method for low-dose CT reconstruction |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |