CN112489153B - Image reconstruction method and application - Google Patents
Image reconstruction method and application Download PDFInfo
- Publication number
- CN112489153B CN112489153B CN202011347074.4A CN202011347074A CN112489153B CN 112489153 B CN112489153 B CN 112489153B CN 202011347074 A CN202011347074 A CN 202011347074A CN 112489153 B CN112489153 B CN 112489153B
- Authority
- CN
- China
- Prior art keywords
- image
- reconstructed
- data
- pixel
- reconstructing
- 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
- 238000000034 method Methods 0.000 title claims abstract description 57
- 238000001914 filtration Methods 0.000 claims abstract description 10
- 238000004422 calculation algorithm Methods 0.000 claims description 17
- 239000011159 matrix material Substances 0.000 claims description 16
- 208000029343 Schaaf-Yang syndrome Diseases 0.000 claims description 13
- 230000004927 fusion Effects 0.000 claims description 6
- 230000009467 reduction Effects 0.000 claims description 3
- 102000008297 Nuclear Matrix-Associated Proteins Human genes 0.000 claims description 2
- 108010035916 Nuclear Matrix-Associated Proteins Proteins 0.000 claims description 2
- 210000000299 nuclear matrix Anatomy 0.000 claims description 2
- 238000002591 computed tomography Methods 0.000 abstract description 10
- 238000013459 approach Methods 0.000 abstract description 2
- 230000006870 function Effects 0.000 description 15
- 238000011084 recovery Methods 0.000 description 5
- 230000009286 beneficial effect Effects 0.000 description 3
- 230000008901 benefit Effects 0.000 description 2
- 238000013170 computed tomography imaging Methods 0.000 description 2
- 238000003745 diagnosis Methods 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 238000005259 measurement Methods 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 230000005855 radiation Effects 0.000 description 2
- 238000004088 simulation Methods 0.000 description 2
- 206010028980 Neoplasm Diseases 0.000 description 1
- 238000007251 Prelog reaction Methods 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 201000011510 cancer Diseases 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000009499 grossing Methods 0.000 description 1
- 238000010801 machine learning Methods 0.000 description 1
- 238000012544 monitoring process Methods 0.000 description 1
- 238000007781 pre-processing Methods 0.000 description 1
- 230000008569 process Effects 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- 238000011002 quantification Methods 0.000 description 1
- 230000003595 spectral effect Effects 0.000 description 1
- 238000013179 statistical model Methods 0.000 description 1
- 230000009466 transformation Effects 0.000 description 1
Abstract
The application belongs to the technical field of computed tomography, and particularly relates to an image reconstruction method and application. In conventional filtered back-projected reconstructed images, this approach may lead to excessive noise and streak artifacts, which in turn lead to further diagnostic and quantitative errors. The application provides an image reconstruction method, which comprises the following steps: step 1: reconstructing the corrected projection data; obtaining reconstruction data; step 2: regularizing the reconstructed data to obtain a regularized image; step 3: reconstructing the regularized image to obtain a reconstructed image; step 4: and taking the reconstructed image as prior information, and conducting guided nuclear filtering under the guidance of the reconstructed image to obtain a restored image. The reconstructed image has good structural information.
Description
Technical Field
The application belongs to the technical field of computed tomography, and particularly relates to an image reconstruction method and application.
Background
The computer X-ray tomographic scanning (X-CT or CT) is a device which uses X-ray to perform tomographic scanning on human body, then the analog signal r received by the detector is changed into digital signal, the attenuation coefficient of each pixel is calculated by the electronic computer, and then the image is reconstructed, so that the tomographic structure of each part of human body can be displayed. Image Fusion (Image Fusion) refers to that Image data about the same target acquired by a multi-source channel is subjected to Image processing, computer technology and the like, beneficial information in each channel is extracted to the greatest extent, and finally, the beneficial information is synthesized into a high-quality Image, so that the utilization rate of the Image information is improved, the interpretation precision and reliability of a computer are improved, the spatial resolution and the spectral resolution of an original Image are improved, and the monitoring is facilitated. The images to be fused are registered well and the pixel bit widths are consistent.
Radiation risk in Computed Tomography (CT) and the cancer risk associated therewith have been major problems of clinical concern. By reducing the scanning scheme of milliampere seconds, the projection number of a human body per rotation is reduced, and low-dose CT imaging can be realized.
However, in conventional filtered back-projected reconstructed images, this approach may lead to excessive noise and streak artifacts, which further lead to diagnostic and quantitative errors.
Disclosure of Invention
1. Technical problem to be solved
Based on the problem that in the traditional filtered back projection reconstructed image, excessive noise and streak artifact can be caused by the method, and further caused by diagnosis and quantification errors, the application provides an image reconstruction method and application.
2. Technical proposal
In order to achieve the above object, the present application provides an image reconstruction method, which includes the steps of: step 1: reconstructing the corrected projection data; obtaining reconstruction data; step 2: regularizing the reconstructed data to obtain a regularized image; step 3: reconstructing the regularized image to obtain a reconstructed image; step 4: and taking the reconstructed image as prior information, and performing Guided nuclear filtering (Guided KERNEL FILTER) under the guidance of the reconstructed image to obtain a restored image.
Another embodiment provided by the application is: and in the step 1, carrying out punishment weighted least square algorithm reconstruction on the corrected projection data to obtain an ideal projection vector to be estimated.
Another embodiment provided by the application is: the step 1 uses a sinogram recovery method based on a penalty weighted least squares algorithm to recover sinogram data using a combination of low milliamp-seconds and sparse view protocols.
Another embodiment provided by the application is: and step 2, using a full variance method based on penalty weighted least squares to use the restored sinogram data for reconstructing an image, and then carrying out neighborhood patch to replace regularization of a single pixel on a non-quadratic penalty function of the full variance method of penalty weighted least squares.
Another embodiment provided by the application is: and 3, reconstructing the regularized image by using a regularization algorithm.
Another embodiment provided by the application is: step 4, constructing a sparse matrix, and normalizing the matrix to obtain a normalized nuclear matrix; and then, guided nuclear filtering is carried out under the guidance of the reconstructed image, so as to obtain a restored image.
Another embodiment provided by the application is: the matrix is constructed by adopting a k-nearest neighbor algorithm.
Another embodiment provided by the application is: in calculating the image roughness between adjacent pixels in the penalty weighted least squares full-variance method, a patch associated with each pixel is used.
Another embodiment provided by the application is: the roughness function of the patch is expressed as:
Where U (μ) is a patch-based roughness function, f j(μ),fk (μ) is a feature vector consisting of intensity values of all pixels at the patch center at pixel j and pixel k, respectively, h is a positive weighting factor, equal to the normalized inverse spatial distance between pixel j and pixel k, and ψ (||f j(μ)-fk(μ)||h) is a non-quadratic penalty function.
The application also provides application of the image reconstruction method, which is applied to image noise reduction and artifact removal or image fusion.
3. Advantageous effects
Compared with the prior art, the image reconstruction method provided by the application has the beneficial effects that:
the image reconstruction method provided by the application is an image reconstruction method for sparse viewing angle and image fusion.
The image reconstruction method provided by the application solves the problem of improving the quality of the low milliampere second or sparse reconstructed image.
The image reconstruction method provided by the application has the advantage that iterative image reconstruction is a promising choice for obtaining high-quality images under low-dose scanning. Therefore, the new low-dose CT reconstruction method is researched and developed, so that the CT imaging quality can be ensured, the harmful radiation dose can be reduced, and the method has important scientific significance and application prospect in the field of medical diagnosis.
According to the image reconstruction method provided by the application, the finally obtained restored image not only suppresses noise, but also retains image details.
The image reconstruction method provided by the application comprises the steps of complete reconstruction and recovery of CT images.
The image reconstruction method provided by the application has the advantage that the reconstructed image of PWLS-TV has good structural information.
The image reconstruction method provided by the application is based on the image update regularized by the non-quadratic penalty function of the patch. The method utilizes spatial regularization to penalize image intensity differences between adjacent pixels, thereby improving image quality. The non-quadratic penalty function may preserve edges.
Drawings
FIG. 1 is a schematic diagram of the overall framework of the image reconstruction method of the present application;
fig. 2 is a schematic diagram of the image reconstruction result of the present application.
Detailed Description
First, the study simulates low dose CT images (ray intensity I0=1 x 10≡5) for different view numbers. For simulation studies of low mas projection data, after calculation of noiseless line integral y i in a model-based direct projection operation, noise measurements at each i are generated from a statistical model of the pre-log projection data
Where I0 is the intensity of the incident X-ray, y i is the sinogram data,The background electronic noise variance was set to 0.1. The noise measurement y i is calculated with a logarithmic transformation of b i. For simulation of projection data, the original 360 views are undersampled to 4 levels, namely: 60. 90, 120 and 180 views.
Hereinafter, specific embodiments of the present application will be described in detail with reference to the accompanying drawings, and according to these detailed descriptions, those skilled in the art can clearly understand the present application and can practice the present application. Features from various embodiments may be combined to obtain new implementations, or substituted for certain features from certain embodiments to obtain further preferred implementations, without departing from the principles of the application.
The existing method has the potential harm to human body due to radiation exposure caused by large dosage in the CT scanning process; the image reconstruction speed is slow due to the large data volume; (3) Artifacts caused by patient movement occur due to the long scan time.
Regularization (regularization) means that in linear algebraic theory, the uncertainty problem is typically defined by a set of linear algebraic equations, and that this set of equations usually originates from an uncertainty inverse problem with a large condition number. Large condition numbers mean that rounding or other errors can seriously affect the outcome of the problem.
Referring to fig. 1-2, the present application provides an image reconstruction method, comprising the steps of:
step 1: reconstructing the corrected projection data; obtaining reconstruction data; step 2: regularizing the reconstructed data to obtain a regularized image; step 3: reconstructing the regularized image to obtain a reconstructed image; step 4: and taking the reconstructed image as prior information, and performing Guided nuclear filtering (Guided KERNEL FILTER) under the guidance of the reconstructed image to obtain a restored image.
Further, in the step 1, the corrected projection data is subjected to a Penalty Weighted Least Squares (PWLS) algorithm to reconstruct to obtain an ideal projection vector to be estimated.
Further, the step 1 uses a sinogram recovery method based on a penalty weighted least squares algorithm to recover sinogram data using a combination of low milliamp-seconds and sparse view protocols.
The corrected projection data is reconstructed by the following PWLS algorithm to obtain P. Performing PWLS-TV reconstruction gives μ, which, although the result of this reconstruction step contains considerable noise, also contains most of the structural information of the CT image.
① PWLS method updates image from sinogram y
② PWLS-TV method image update from P
Y represents the resulting sinogram data (system calibrated and logarithmically transformed projections), i.e., y= (y 1,y2,...,yM)T; P represents the ideal projection vector to be estimated; β is a super-parameter used to balance fidelity, where H represents the projection matrix system; n is the total number of pixels in the image; R (P) is the penalty for image roughness.
Further, the step 2 uses the full variance (PWLS-TV) method based on the penalty weighted least squares to use the restored sinogram data to reconstruct an image, and then performs neighborhood patching on the non-quadratic penalty function of the full variance method of the penalty weighted least squares instead of regularization of a single pixel (PWLS-PR).
Patch-based regularization is performed on non-quadratic penalty functions of reconstructed data that passes through PWLS-TV.
Omega jk(μn) is calculated as follows
Wherein,For a patch regularized based updated function, # jk(μn) is a weighting factor related to the distance between neighborhood N j pixel j and pixel k, μ n is the current objective function,/>Pixels j and k of the current objective function.
Further, the step 3 uses a regularization algorithm to reconstruct the regularized image.
Further, constructing a sparse matrix, and normalizing the matrix to obtain a normalized kernel matrix; and then, guided nuclear filtering is carried out under the guidance of the reconstructed image, so as to obtain a restored image. Although smoothing can effectively remove noise, some detail is lost, resulting in loss of detail in the reconstructed image. Therefore, the lost structure information is recovered by adopting a guided kernel filtering method, so that the details of the image are well reserved in the denoised image, and the structure is clearer.
Further, the matrix is constructed by adopting a k-nearest neighbor algorithm.
Further, in calculating the image roughness between adjacent pixels in the penalty weighted least squares total variation method, a patch associated with each pixel is used.
The sparse matrix is constructed using the k-nearest neighbor algorithm widely used in machine learning. The kNN square finds K similar neighbors for each pixel and normalizes the matrix to obtain a normalized kernel matrix K. Then GK filtering (guided kernel filtering) is carried out under the guidance of the reconstructed image mu TV to obtain a final image x
x＝K·μTV
Where K is the kernel matrix and mu TV is PWLS-TV method image update from P
Further, the roughness function of the patch is expressed as:
Where U (μ) is a patch-based roughness function, f j(μ),fk (μ) is a feature vector consisting of intensity values of all pixels at the patch center at pixel j and pixel k, respectively, h is a positive weighting factor, equal to the normalized inverse spatial distance between pixel j and pixel k, and ψ (||f j(μ)-fk(μ)||h) is a non-quadratic penalty function.
The application also provides application of the image reconstruction method, which is applied to image noise reduction and artifact removal or image fusion.
Sinogram data using a combination of low milliamp-seconds and sparse view protocols is recovered using a PWLS (penalty weighted least squares) based sinogram recovery method. The restored sinogram data is then reconstructed into an image using a (penalty weighted least squares based total variation) PWLS-TV method. PWLS sinogram recovery is effective as a preprocessing step for image reconstruction using a combination of low mas and sparse views.
A patch-based iterative image reconstruction regularization method uses a neighborhood patch instead of a single pixel when calculating its non-quadratic penalty function. The regularization method is more robust in distinguishing random fluctuations caused by noise from sharp edges than conventional pixel-based regularization methods.
FIG. 2 is an XCAT Phantom reconstructed by different algorithms; as can be seen from FIG. 2, the method of the application can well reduce noise and remove artifacts and preserve the detail structure of the image, can effectively improve the peak signal-to-noise ratio and the structural similarity of the image, and can recover the detail information of the image to a certain extent.
Although the application has been described with reference to specific embodiments, those skilled in the art will appreciate that many modifications are possible in the construction and detail of the application disclosed within the spirit and scope thereof. The scope of the application is to be determined by the appended claims, and it is intended that the claims cover all modifications that are within the literal meaning or range of equivalents of the technical features of the claims.
Claims (5)
1. An image reconstruction method, characterized in that: the method comprises the following steps:
step 1: reconstructing the corrected projection data; obtaining reconstruction data;
Step 2: regularizing the reconstructed data to obtain a regularized image;
step 3: reconstructing the regularized image to obtain a reconstructed image;
Step 4: taking the reconstructed image as priori information, and performing guided nuclear filtering on the reconstructed data obtained in the step 1 under the guidance of the reconstructed image to obtain a restored image; in the step 1, carrying out punishment weighted least square algorithm reconstruction on corrected projection data to obtain an ideal projection vector to be estimated; step 1, restoring the sinogram data using a combination of low milliampere seconds and sparse view protocols by using a sinogram restoration method based on a penalty weighted least squares algorithm; reconstructing the corrected projection data by using PWLS algorithm to obtain P *(n+1); performing PWLS-TV reconstruction to obtain
① PWLS method updates image from sinogram y
② PWLS-TV method image update from P
Y represents the obtained sinogram data, namely: y= (y 1,y2,...,yM)T; n is the iteration number; p represents the ideal projection vector to be estimated, β is a super parameter for balancing the fidelity, wherein H represents the projection matrix system, R (p n) is the penalty of the PWLS algorithm image roughness, the step 2 uses the recovered sinogram data for reconstructing the image by using the full variance method based on the penalty weighted least squares, and regularizes the non-quadratic penalty function of the full variance method based on the penalty weighted least squares by using a neighborhood patch instead of a single pixel, and in calculating the image roughness between adjacent pixels in the full variance method based on the penalty weighted least squares, a patch associated with each pixel is used, and the roughness function of the patch is expressed as:
Where U (μ) is a patch-based roughness function, f j(μ),fk (μ) is a feature vector consisting of intensity values of all pixels at the patch center at pixel j and pixel k, respectively, h is a positive weighting factor, equal to the normalized inverse spatial distance between pixel j and pixel k, Is a non-quadratic penalty function.
2. The image reconstruction method according to claim 1, wherein: and 3, reconstructing the regularized image by using a regularization algorithm.
3. The image reconstruction method according to claim 1, wherein: step 4, constructing a sparse matrix, and normalizing the matrix to obtain a normalized nuclear matrix; and then, guided nuclear filtering is carried out under the guidance of the reconstructed image, so as to obtain a restored image.
4. The image reconstruction method as set forth in claim 3, wherein: the matrix is constructed by adopting a k-nearest neighbor algorithm.
5. The image reconstruction method according to any one of claims 1 to 4, wherein: the image reconstruction method is applied to image noise reduction and artifact removal or image fusion.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011347074.4A CN112489153B (en) | 2020-11-26 | Image reconstruction method and application |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011347074.4A CN112489153B (en) | 2020-11-26 | Image reconstruction method and application |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN112489153A CN112489153A (en) | 2021-03-12 |
| CN112489153B true CN112489153B (en) | 2024-06-21 |
Family
ID=
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110276813A (en) * | 2019-05-06 | 2019-09-24 | 深圳先进技术研究院 | CT image rebuilding method, device, storage medium and computer equipment |
| CN111709897A (en) * | 2020-06-18 | 2020-09-25 | 深圳先进技术研究院 | Method for reconstructing positron emission tomography image based on domain transformation |
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110276813A (en) * | 2019-05-06 | 2019-09-24 | 深圳先进技术研究院 | CT image rebuilding method, device, storage medium and computer equipment |
| CN111709897A (en) * | 2020-06-18 | 2020-09-25 | 深圳先进技术研究院 | Method for reconstructing positron emission tomography image based on domain transformation |
Non-Patent Citations (2)
| Title |
|---|
| Penalized Likelihood PET Image Reconstruction Using Patch-Based Edge-Preserving Regularization;Guobao Wang 等;IEEE TRANSACTIONS ON MEDICAL IMAGING;第31卷(第12期);第2194-2204页 * |
| PWLS-PR: low-dose computed tomography image reconstruction using a patch-based regularization method based on the penalized weighted least squares total variation approach;Jing Fu 等;Original Article;第2541-2559页 * |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108961237B (en) | Low-dose CT image decomposition method based on convolutional neural network | |
| JP2020168352A (en) | Medical apparatus and program | |
| CN109785243B (en) | Denoising method and computer based on unregistered low-dose CT of countermeasure generation network | |
| CN102737392B (en) | Non-partial regularization prior reconstruction method for low-dosage X-ray captive test (CT) image | |
| CN111325686A (en) | Low-dose PET three-dimensional reconstruction method based on deep learning | |
| CN101980302A (en) | Projection data recovery-guided nonlocal mean low-dose CT reconstruction method | |
| CN103810733B (en) | A kind of statistics iterative reconstruction approach of sparse angular X ray CT image | |
| CN103810734B (en) | A kind of low dose X-ray CT data for projection restoration methods | |
| CN103679706B (en) | A kind of CT sparse angular method for reconstructing based on image anisotropy rim detection | |
| WO2022000192A1 (en) | Ct image construction method, ct device, and storage medium | |
| Zhang et al. | REDAEP: Robust and enhanced denoising autoencoding prior for sparse-view CT reconstruction | |
| CN115239588A (en) | Extremely-sparse view CT reconstruction artifact suppression method based on generation countermeasure network | |
| Liang et al. | Guest editorial low-dose CT: what has been done, and what challenges remain? | |
| Song et al. | Low-dose cardiac-gated spect studies using a residual convolutional neural network | |
| CN103793890A (en) | Method for recovering and processing energy spectrum CT images | |
| Qi et al. | Iterative image reconstruction using modified non-local means filtering for limited-angle computed tomography | |
| Feng et al. | Dual residual convolutional neural network (DRCNN) for low-dose CT imaging | |
| CN112656438B (en) | Low-dose CT projection domain denoising and reconstructing method based on curved surface total variation | |
| CN105844678A (en) | Low dose X-ray CT image reconstruction method based on completely generalized variational regularization | |
| Lv et al. | A back‐projection‐and‐filtering‐like (BPF‐like) reconstruction method with the deep learning filtration from listmode data in TOF‐PET | |
| JP7187131B2 (en) | Image generation device, X-ray computed tomography device and image generation method | |
| CN112489153B (en) | Image reconstruction method and application | |
| Zheng et al. | Improving spatial adaptivity of nonlocal means in low‐dosed CT imaging using pointwise fractal dimension | |
| CN112116677B (en) | Low-dose CT reconstruction method based on low-dimensional manifold priori | |
| WO2022109928A1 (en) | Image reconstruction method and application |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant |