CN112085808B - CT image reconstruction method, system, equipment and medium - Google Patents
CT image reconstruction method, system, equipment and medium Download PDFInfo
- Publication number
- CN112085808B CN112085808B CN202010962224.6A CN202010962224A CN112085808B CN 112085808 B CN112085808 B CN 112085808B CN 202010962224 A CN202010962224 A CN 202010962224A CN 112085808 B CN112085808 B CN 112085808B
- Authority
- CN
- China
- Prior art keywords
- image
- pixel
- iteration
- processing
- roughness
- 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 100
- 238000009499 grossing Methods 0.000 claims abstract description 29
- 230000008569 process Effects 0.000 claims abstract description 23
- 238000012545 processing Methods 0.000 claims description 100
- 230000006870 function Effects 0.000 claims description 72
- 238000004422 calculation algorithm Methods 0.000 claims description 30
- 238000007499 fusion processing Methods 0.000 claims description 26
- 230000014509 gene expression Effects 0.000 claims description 19
- 230000005540 biological transmission Effects 0.000 claims description 17
- 238000005457 optimization Methods 0.000 claims description 5
- 238000007476 Maximum Likelihood Methods 0.000 claims description 4
- 238000004590 computer program Methods 0.000 claims description 4
- 238000012887 quadratic function Methods 0.000 claims description 4
- 238000012804 iterative process Methods 0.000 claims description 2
- 238000010586 diagram Methods 0.000 description 7
- 238000004364 calculation method Methods 0.000 description 5
- 230000007547 defect Effects 0.000 description 4
- 230000000694 effects Effects 0.000 description 4
- 238000005259 measurement Methods 0.000 description 4
- 238000013459 approach Methods 0.000 description 2
- 230000015556 catabolic process Effects 0.000 description 2
- 238000006731 degradation reaction Methods 0.000 description 2
- 238000011161 development Methods 0.000 description 2
- 238000003745 diagnosis Methods 0.000 description 2
- 230000004927 fusion Effects 0.000 description 2
- 238000003384 imaging method Methods 0.000 description 2
- 230000014759 maintenance of location Effects 0.000 description 2
- 230000003287 optical effect Effects 0.000 description 2
- 238000011160 research Methods 0.000 description 2
- 230000000717 retained effect Effects 0.000 description 2
- 238000013179 statistical model Methods 0.000 description 2
- 230000000007 visual effect Effects 0.000 description 2
- 206010028980 Neoplasm Diseases 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 201000011510 cancer Diseases 0.000 description 1
- 238000013170 computed tomography imaging Methods 0.000 description 1
- 238000001514 detection method Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000003706 image smoothing Methods 0.000 description 1
- 230000036039 immunity Effects 0.000 description 1
- 230000001788 irregular Effects 0.000 description 1
- 230000003902 lesion Effects 0.000 description 1
- 239000011159 matrix material Substances 0.000 description 1
- 230000006371 metabolic abnormality Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 239000013307 optical fiber Substances 0.000 description 1
- 210000000056 organ Anatomy 0.000 description 1
- 230000002265 prevention Effects 0.000 description 1
- 230000005855 radiation Effects 0.000 description 1
- 229910052704 radon Inorganic materials 0.000 description 1
- SYUHGPGVQRZVTB-UHFFFAOYSA-N radon atom Chemical compound [Rn] SYUHGPGVQRZVTB-UHFFFAOYSA-N 0.000 description 1
- 230000008707 rearrangement Effects 0.000 description 1
- 230000004044 response Effects 0.000 description 1
- 239000004065 semiconductor Substances 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T11/00—2D [Two Dimensional] image generation
- G06T11/003—Reconstruction from projections, e.g. tomography
- G06T11/008—Specific post-processing after tomographic reconstruction, e.g. voxelisation, metal artifact correction
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T5/00—Image enhancement or restoration
- G06T5/50—Image enhancement or restoration by the use of more than one image, e.g. averaging, subtraction
-
- G06T5/70—
-
- 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
- G06T2207/20212—Image combination
- G06T2207/20221—Image fusion; Image merging
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30004—Biomedical image processing
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2210/00—Indexing scheme for image generation or computer graphics
- G06T2210/41—Medical
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2211/00—Image generation
- G06T2211/40—Computed tomography
- G06T2211/424—Iterative
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Apparatus For Radiation Diagnosis (AREA)
Abstract
The invention provides a CT image reconstruction method, a CT image reconstruction system, CT image reconstruction equipment and a CT image reconstruction medium. The method comprises the following steps: carrying out image iterative reconstruction on the image, judging whether the iterated image meets the iteration stopping condition, and stopping iteration if the iterated image meets the iteration stopping condition; if not, continuing iteration; each iteration process sequentially comprises the following steps: updating, smoothing and fusing; the updating process comprises the steps of solving the image roughness by calculating the intensity difference of the adjacent patches, and further obtaining the maximized punishment likelihood function. The invention can give consideration to the image quality and the image generation efficiency on the premise of ensuring the safety, can output images with higher quality at a higher speed, and simultaneously retains the fine characteristics of the images, thereby having great significance in the field of medical images.
Description
Technical Field
The present invention relates to the field of CT image processing, and in particular, to a method, system, apparatus, and medium for CT image reconstruction.
Background
The CT technology has the characteristics of rapidness, accuracy, no wound, no pain and the like, and can clearly obtain the attenuation information of different tissues of a human body on X-rays on a millimeter scale, thereby providing the information of human organ tissues for diagnosis and prevention of clinicians. The CT imaging comprises the following specific processes: the detector records projection data of X-rays passing through the object under different visual angles, and the linear attenuation coefficient diagram of the fault can be obtained by reconstructing the projection data. CT reconstruction algorithms are mainly divided into two types, namely an analytical reconstruction algorithm and an iterative reconstruction algorithm.
The most widely applied algorithm in the analytical reconstruction algorithm is filtered back-projection (FBP), and the FBP performs image reconstruction based on the Radon transform, but the FBP does not consider the space-time heterogeneity of the system response, nor the noise of the instrument during measurement, and does not have anti-noise performance, so that the reconstruction quality of the CT image cannot be ensured when the signal-to-noise ratio of projection data is low.
The iterative reconstruction algorithm comprises algebraic reconstruction and statistical reconstruction. The algebraic reconstruction mainly comprises algebraic reconstruction algorithms and new algorithms obtained by further expanding the algebraic reconstruction algorithms on the basis. The maximum likelihood-expectation maximization (Maximum likelihood-expectation maximization, ML-EM) method in statistical reconstruction is currently widely used in clinic and practice because it has better performance in lesion detection than conventional algorithms.
In the transmission process of X-rays, part of energy can be transferred to a human body, so that the human body is damaged, metabolism abnormality of the human body is induced, even cancer is caused, and the problem of low-dose CT gradually becomes a research hot spot. Reducing the X-ray tube current intensity to reduce the exposure dose at each view angle is one of the most common methods of reducing the CT dose. Reducing the tube current intensity reduces the radiation dose for a single scan, but reduces the signal-to-noise ratio of the projection data, increasing the noise intensity exponentially. The filtered backprojection method, because it does not have noise immunity, results in a reconstructed low dose CT image that is severely degraded and contains a significant amount of streak artifacts. Therefore, iterative reconstruction algorithms are commonly employed for low dose CT image reconstruction.
However, the conventional image iterative reconstruction method is based on pixels for image reconstruction, and a large amount of iterative processing is required, so that the reconstruction process is complex and tedious. The maximum likelihood-expectation maximization method can degrade the image quality with the increase of the iteration times, generate 'checkerboard artifact', and seriously influence the reconstruction effect. Because it is processed on a fine pixel-by-pixel basis, fine texture features in some images are eliminated, which tend to affect the determination of CT images. In addition, the large amount of noise caused by reducing the CT dose can cause edge effects, so that the image edge holding capacity is poor, and the distinction of the image edges is seriously influenced.
In summary, the prior art generally has the defects of complex image reconstruction process, large amount of noise and artifact in the reconstructed image, poor edge holding capability, unreserved original characteristic textures in the image and the like, so that the problems of long reconstruction time, poor quality of the reconstructed image, missing image content and the like are caused, and diagnosis and treatment of a patient are seriously influenced. Therefore, there is a need for an image reconstruction method that shortens the reconstruction process, improves the image quality, and preserves the image characteristics.
Disclosure of Invention
Based on the problems existing in the prior art, the invention provides a CT image reconstruction method, a CT image reconstruction system, CT image reconstruction equipment and CT image reconstruction media. The specific scheme is as follows:
a CT image reconstruction method comprising the steps of: s1, acquiring an initialization image according to projection data; s2, taking the initialized image as an image to be reconstructed, and carrying out image iterative processing on the image to be reconstructed; s3, judging whether an iteration stopping condition is met, if so, stopping iteration, and outputting an image after iteration processing; if not, returning to the step S2, and carrying out iterative processing again by taking the image after iterative processing as the image to be reconstructed. In the step S2, the image iterative process further includes the steps of: s21, updating, namely solving the image roughness by calculating the intensity difference of the adjacent patches so as to obtain a maximized punishment likelihood function; s22, smoothing processing, which comprises smoothing and denoising the image from S21; s23, fusion processing, including area fusion processing of the images from S22, is carried out, and the images are fused into images pixel by pixel.
Further, in the update processing of S21, the patch for calculating the image roughness includes a square pixel area centered on one pixel, which is constituted by a plurality of pixels, and all the patches for calculating the image roughness are the same in size.
Further, in the updating process of S21, the expression of the maximum penalty likelihood function is:
wherein μ is a predicted image value, Φ (μ) is an objective function, L (y|μ) is a likelihood function, β is a regularization parameter controlling data fidelity and spatial smoothness, U (μ) is an image roughness, a maximized penalty likelihood function is obtained by solving for the image roughness, the image roughness expression is:
wherein, ψ (μ) p -μ q ) Is a penalty function, mu p -μ q Omega is the distance between pixel p and adjacent pixel q pq Is a weight factor of the distance between pixel p and pixel q in the neighborhood, n p Is the total number of detectors, N p Is the total number of image pixels.
Further, in the update processing of S21, the image roughness based on the patch is calculated by the following method;
the pixel j and the pixel k are respectively the center pixels of two adjacent patches, the image roughness is solved by calculating the intensity difference between the two patches, and then the maximum punishment likelihood function is obtained, and the pixel distance expression of the pixel j and the pixel k is as follows:
the image roughness expressions of the pixel j and the pixel k are:
wherein: f (f) j (mu) is a feature vector composed of all pixel intensity values of the patch centered at the pixel j, f k (mu) is a feature vector composed of all pixel intensity values of the patch centered at the pixel k, j l Represents the first pixel, k, in a patch centered on pixel j l Represents the first pixel, n, in a patch centered on pixel k j Identifying the total number of detectors, N j Indicating the number of all pixel points in the patch, h l Is the normalized inverse space from pixel to pixelPositive weighting factor of distance, ψ (||f) j (μ)-f k (μ)|| h ) Is a penalty function.
In particular, the penalty function expression is:
wherein δ is the attenuation factor;
when f j (μ)-f k (μ)|| h When the delta is less than the delta, the punishment function approximates a quadratic function;
when f j (μ)-f k (μ)|| h When I > delta, the penalty function approximates an absolute function.
Further, the updating process of S21 further includes optimizing a transmission algorithm;
the optimization transmission algorithm includes constructing a proxy function at each iteration, and minimizing the objective function Φ (μ).
Still further, the optimized transmission algorithm includes a expectation-maximization algorithm, and the acquisition is performed according to the expectation-maximization algorithmWhere n+1 is the number of iterations, j represents the center pixel sequence number of the patch, and EM represents the desired maximization.
Specifically, the step S22 includes smoothing by regularization, and acquiringWherein Re g represents the regularization method.
Specifically, the fusion process of S23 includes fusing pixels into an image, and obtaining the next iteration image estimation according to the fusion process:
when β=0, the above equation reduces to the maximum likelihood expectation maximization equation.
A CT image reconstruction system comprising: an initializing unit: the method comprises the steps of acquiring an initialization image according to projection data; iteration unit: the initialization image is used as an image to be reconstructed, and the image to be reconstructed is subjected to iterative processing; and a judgment output unit: the method comprises the steps of judging whether an iteration stopping condition is met, and if so, outputting an image after iteration processing; if not, sending the image after the iteration processing to the iteration unit, and carrying out the iteration processing again as an image to be reconstructed; the iteration unit comprises an updating processing unit, a smoothing processing unit and a fusion processing unit; the update processing unit further includes a roughness calculation unit that calculates an image roughness based on the patch.
Further, the update processing unit further includes: and the optimization transmission unit is used for constructing a proxy function when the image is iterated each time and carrying out minimization processing on the objective function phi (mu).
A computer device, the computer device comprising: one or more processors; a memory for storing one or more programs; the one or more programs, when executed by the one or more processors, cause the one or more processors to implement the CT image reconstruction method described above.
A computer readable storage medium having stored thereon a computer program which when executed by a processor implements the CT image reconstruction method described above.
The invention has the following beneficial effects:
aiming at the technical problems of low-dose CT image imaging quality and speed, the invention provides an image reconstruction method, an image reconstruction system, image reconstruction equipment and an image reconstruction medium, which solve the defects that the prior art generally has complex image reconstruction process, a large amount of noise and artifacts exist in a reconstructed image, the edge retaining capability is poor, the original characteristic texture in the image is not retained and the like, can eliminate the noise, inhibit the artifacts, retain the edge and fine characteristics, reduce the iteration times and simultaneously ensure the image quality, has the characteristics of short reconstruction process, high image quality, comprehensive image characteristic retention and the like, and has important significance for the development of the medical influence field.
In order to make the above objects, features and advantages of the present invention more comprehensible, preferred embodiments accompanied with figures are described in detail below.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings that are needed in the embodiments will be briefly described below, it being understood that the following drawings only illustrate some embodiments of the present invention and therefore should not be considered as limiting the scope, and other related drawings may be obtained according to these drawings without inventive effort for a person skilled in the art.
FIG. 1 is a flow chart of a CT image reconstruction method of the present invention;
FIG. 2 is a flowchart of a CT image reconstruction method of the present invention;
FIG. 3 is a block diagram of a CT image reconstruction system of the present invention;
FIG. 4 is a specific block diagram of a CT image reconstruction system of the present invention;
fig. 5 is a schematic diagram of a computer device to which the CT image reconstruction method of the present invention is applied.
Detailed Description
The CT image reconstruction process comprises the following steps: firstly, recording projection data of X-rays passing through a scanned object under different visual angles by a detector; secondly, constructing projection data into a required CT initial image, and then carrying out image iterative reconstruction processing, wherein in general, the image needs to be subjected to iterative processing for a plurality of times; and then judging whether the reconstructed image after each iteration meets the iteration stop condition, if so, outputting the reconstructed image meeting the iteration stop condition, and if not, continuing to reconstruct the image iteration.
Definition: assuming measurement errors and noise effects, the X-ray CT measurement is expressed as follows:
y＝Hx+e
wherein: y= (y) 1 ,y 2 ,...,y M ) T Representation systemCalibrating and logarithmically converting projection data; x= (x 1 ,x 2 ,…,x N ) T Representing the attenuation coefficient to be estimated; h denotes the image vector to be solved, T denotes the matrix transpose, e denotes the measurement error and noise. The purpose of CT image reconstruction is to estimate the attenuation coefficient x from the measured value y of H.
The invention mainly carries out image reconstruction by punishment likelihood method. The penalty likelihood method estimates the reconstructed image by maximizing a penalty likelihood function, and the specific expression is as follows:
Φ(μ)＝L(y|μ)-βU(μ)
where L (y|μ) is a likelihood function, U (μ) is an image roughness penalty term, Φ (μ) can be understood as an objective function, β is a regularization parameter controlling data fidelity and spatial smoothness, μ is a predicted image value to be solved.
The likelihood function is a function of parameters in the statistical model that represent the likelihood in the model parameters. Likelihood functions have a significant role in statistical inference, such as maximum likelihood estimation (ML), which is a function of finding the maximum value of a likelihood function, and representing corresponding parameters can make a statistical model most reasonable, and in practical application, the logarithm of the likelihood function is generally taken as the function of finding the maximum value. In the above expression, as β approaches 0, the reconstructed image approaches the maximum likelihood estimate.
The general image roughness is measured based on the intensity difference between two adjacent pixels, expressed as:
wherein, ψ (μ) p -μ q ) Is a penalty function, mu p -μ q Omega is the distance between pixel p and adjacent pixel q pq Is the weighting factor of the distance between pixel p and pixel q in the neighborhoodSon, n p Is the total number of detectors, N p Is the total number of image pixels.
Example 1
Aiming at the defects of the prior art, the embodiment provides a CT image reconstruction method. The specific scheme of the embodiment is as follows:
an image reconstruction method, the method comprising: s1, acquiring an initialization image according to projection data; s2, taking the initialized image as an image to be reconstructed, and carrying out image iterative processing on the image to be reconstructed; s3, judging whether an iteration stopping condition is met, if so, stopping iteration, and outputting an image after iteration processing; if not, returning to the step S2, and carrying out iterative processing again by taking the image after iterative processing as the image to be reconstructed. The specific flow chart is shown in figure 1 of the accompanying drawings. Wherein the image iterative processing in S2 includes performing, on the image: s21, updating, namely solving the image roughness by calculating the intensity difference of the adjacent patches so as to obtain a maximized punishment likelihood function; s22, smoothing processing, namely performing smoothing denoising processing on the image from the S21; s23, performing fusion processing, namely performing region fusion processing on the image from S22, wherein the fusion processing comprises fusing the images pixel by pixel. The specific flow is shown in figure 2 of the specification.
And carrying out iterative processing on the image to be reconstructed. The image to be reconstructed includes an initialized image and an image after the iteration process that does not satisfy the iteration stop condition, that is, the initialized image of S1 and the image after the iteration process in S3 but that does not satisfy the iteration stop condition. Specifically, an image update process is performed on an image. The image update processing section estimates an image using a penalty likelihood function, more specifically, estimates a reconstructed image by maximizing the penalty likelihood function, as follows:
Φ(μ)＝L(y|μ)-βU(μ)
where L (y|μ) is a likelihood function, U (μ) is an image roughness penalty term, Φ (μ) can be understood as an objective function, β is a regularization parameter controlling data fidelity and spatial smoothness, μ is a predicted image value. Obtaining a maximized punishment likelihood function by solving for image roughness, wherein the image roughness expression is as follows:
wherein, ψ (μ) p -μ q ) Is a penalty function, mu p -μ q Omega is the distance between pixel p and adjacent pixel q pq Is a weight factor for the distance between pixel p and pixel q in the neighborhood.
Further, the S21 image update process further includes performing calculation of image roughness based on the patch. The patch includes a pixel block formed by a plurality of adjacent pixels, and the shape of the pixel block may be regular, such as square, or irregular. Preferably, the embodiment selects a regular pixel block, specifically, selects a patch with one pixel as the center and other pixels around. In 2D modeling, the patches include 3×3, 5×5, etc., and may be a single pixel. The patch may include complete pixels or incomplete pixels, i.e. square areas formed by cutting pixels, and preferably, the complete pixels are selected in this embodiment, i.e. the patch includes only complete pixels, for example, 9 complete pixels in a 3×3 patch, and one pixel is used as a center, and the other 8 pixels surround the center pixel. Different sized patches affect the efficiency of image processing. The patch has a larger area relative to a single pixel, and more texture features on the pixel are preserved, without excessive fine features being eliminated due to the smoothing process. Preferably, the patches in the image are the same size. Selecting a pixel j and a pixel k to calculate the image roughness according to the requirement of a punishment likelihood method, wherein the pixel j and the pixel k are respectively the center pixels of two adjacent patches, solving the image roughness by calculating the intensity difference between the two patches, and further obtaining a maximized punishment likelihood function, and the distance expression between the pixel j and the adjacent pixel k is as follows:
wherein: j (j) l Represents the first pixel, k, in a patch centered on pixel j l Represents the first pixel, n, in a patch centered on pixel k j Identifying the total number of detectors, N j Indicating the number of all pixel points in the patch, h l Is the positive weighting factor of the normalized inverse spatial distance between pixels:
according to the calculation expression of the image roughness, the image roughness based on the surface patch is obtained as follows:
wherein f j (mu) is a feature vector composed of all pixel intensity values of the patch centered at the pixel j, f k (μ) is a feature vector consisting of all pixel intensity values of the patch centered at the pixel k, ψ (||f) j (μ)-f k (μ)|| h ) Is a penalty function. ψ (||f) j (μ)-f k (μ)|| h ) The penalty function includes a quadratic function and a hyperbolic function, and preferably, the penalty function expression selected in this embodiment is:
when f j (μ)-f k (μ)|| h When I < the penalty function approximates a quadratic function;
when f j (μ)-f k (μ)|| h When I > the penalty function approximates an absolute function.
Wherein δ is a penalty factor that is not differentiable at zero, has a continuous second derivative, and can ensure that the edges of the reconstructed image are preserved, forming a sharp edge. The regularization method based on the patches calculates the image roughness by comparing the similarity between the patches, has stronger edge holding capacity, reduces iteration times, shortens image reconstruction time, and reserves fine texture features of the image. Compared with the traditional prior based on single pixels, the method has better robustness in distinguishing edges, and can also reduce noise in a reconstructed image and retain more detail information.
Further, the S21 image update process also includes an optimized transmission algorithm whose basic idea includes constructing a proxy function at each iteration, i.e., constructing the proxy function Q (μ; μ) of the image μ at the nth image iteration n ) The objective function Φ (μ) is subjected to the minimization process by satisfying the following two conditions:
Q(μ；μ n )-Q(μ n ；μ n )≤Φ(μ)-Φ(μ n )
wherein,represents the gradient with respect to μ, the objective function Φ (μ) =l (y|μ) - βu (μ);
converting the maximum value of phi (mu) to Q (mu; mu) n ) Is carried into the maximization penalty likelihood function to obtain:
phi (mu) with updated mu n+1 Monotonically increasing: phi (mu) n+1 )≥Φ(μ n )。
Each image iteration is updated as follows:
a well-known Expectation Maximization (EM) algorithm is a special case of an optimized transmission algorithm. Based on the prior study, the invention selects an Expectation Maximization (EM) algorithm to update the image to obtain an estimated image value based on the expectation maximization, namelyWhere n+1 is the number of iterations, j represents the center pixel sequence number of the patch, and EM represents the desired maximization.
Specifically, in S22, the image from S21 is subjected to image smoothing processing. Specifically, the regularization method is adopted to carry out smoothing treatment on the image, a regularization term is added in a likelihood function, and the proportion of parameters is controlled through the size of a threshold (or called regularization term coefficient), so that the solution of the problem is weighted between the fitting precision of the model and the sparsity measure. The core purpose of the regularization method is to limit the size of the parameter space to reduce the complexity of the model, and limit the value range of the solution by adding a regularization term to ensure good properties of the solution, such as uniqueness. In the smoothing process of S22, a predicted image value is obtained based on the regularization method, that isWhere n is the number of iterations, j represents the center pixel number of the patch, and Re g represents the regularization method.
Specifically, in S23, the image from S22 is subjected to image fusion processing. Specifically, the image after the update processing and the smoothing processing is subjected to pixel-by-pixel image fusion, and pixels in the image are fused, so that the image can be fused into a patch, such as a patch with a 3×3 structure and a 5×5 structure. According to the result obtained in the update processing of S21And S22Obtaining the next iteration image estimation, wherein the estimated image value expression is as follows:
where β is a regularization parameter controlling data fidelity and spatial smoothness, and when β=0, the above updated equation is reduced to a maximum likelihood-expectation maximization formula. The maximum likelihood-expectation maximization method degrades image quality with increasing iteration number, creating "checkerboard artifacts". The iteration is terminated in advance and penalty terms or some priori knowledge are integrated in the likelihood function, which overcomes the problems of the maximum likelihood-expectation maximization method to some extent. When the iteration is not converged and the reconstructed image with the same effect is to be achieved, the method and the device have fewer iteration times, and avoid degradation of image quality. The projection data is converted into the image through one-time image iterative reconstruction, so that the iteration times are small, the image quality is high, and the iteration efficiency is high.
Example 2
In order to overcome the shortcomings of the prior art, a CT image reconstruction system is provided in this embodiment, and the step modules in a CT image reconstruction method in embodiment 1 are embodied to form a system, where a system block diagram is shown in fig. 3 of the specification, and specifically includes:
an initializing unit: the method comprises the steps of acquiring an initialization image according to projection data;
iteration unit: the method comprises the steps of using an initialization image as an image to be reconstructed, and carrying out image iterative processing on the image to be reconstructed;
and a judgment output unit: the method comprises the steps of judging whether an iteration stopping condition is met, and if so, outputting an image after iteration processing; if not, the image after the iteration processing is sent to an iteration unit, and the iteration processing is carried out again.
The iteration unit comprises an updating processing unit, a smoothing processing unit and a fusion processing unit; the image sequentially passes through an updating processing unit, a smoothing processing unit and a fusion processing unit; the updating processing unit is used for solving the image roughness by calculating the intensity difference of the adjacent patches so as to obtain a maximized punishment likelihood function; a smoothing processing unit for performing smoothing denoising processing on the image from the updating processing unit; and the fusion processing unit is used for carrying out region fusion processing on the image from the smoothing processing unit, and the region fusion processing comprises pixel-by-pixel fusion into an image.
The updating processing unit further comprises a roughness calculating unit, wherein the roughness calculating unit is used for storing an algorithm for calculating the roughness of the image based on the surface patch, and performing image roughness calculation based on the surface patch on the image.
The initialization unit establishes an image model according to the acquired projection data of the projection object, forms an initialization image and sends the initialization image to the iteration unit; the updating processing unit of the iteration unit receives the initialization image sent by the initialization unit, takes the initialization image as an image to be reconstructed, performs image updating processing on the image to be reconstructed, and sends the updated image to the smoothing processing unit after updating processing; the smoothing processing unit receives the image sent by the updating processing unit, performs smoothing processing on the image based on a regularization method, and sends the processed image to the fusion processing unit; the fusion processing unit receives the image sent by the smoothing processing unit, performs region fusion processing on the image, comprises fusing the image pixel by pixel, and sends the processed image to the judgment output unit to complete one round of iterative processing; and the judging output unit receives the image sent by the iteration unit, judges whether the iteration stopping condition is met, outputs the image after the iteration processing if the iteration stopping condition is met, and returns the image after the iteration processing to the iteration unit to perform a new round of iteration processing if the iteration stopping condition is not met. A specific system block diagram is shown in fig. 4 of the specification.
The updating processing unit further comprises an optimizing transmission unit, wherein the optimizing transmission unit is provided with an optimizing transmission algorithm, and the optimizing transmission algorithm is used for constructing a proxy function when the image iterates each time and carrying out minimizing processing on an objective function phi (mu).
In order to solve the technical problems of low-dose CT image imaging quality and speed, the invention combines the existing researches and provides a CT image reconstruction system which can take the image quality and the image generation efficiency into consideration on the premise of ensuring the safety and can output (e.g. display) images with higher quality at a higher speed. And carrying out iterative reconstruction processing on the image by adopting an iteration unit, sequentially passing the image through an updating processing unit, a smoothing processing unit and a fusion processing unit, and finally judging whether the iteration stopping condition is met by an output unit. The embodiment optimizes punishment likelihood reconstruction by using an optimized transmission algorithm, solves the artifacts and noise caused by low dosage, avoids degradation of image quality and improves the image quality; the image roughness is calculated by adopting the image roughness calculation method based on the surface patch, so that the reconstruction time is shortened while the image quality is ensured, and the fine features and the edge features of the image are reserved.
Example 3
Fig. 5 is a schematic structural diagram of a computer device according to embodiment 3 of the present invention. The computer device 12 shown in fig. 5 is merely an example and should not be construed as limiting the functionality and scope of use of embodiments of the present invention.
As shown in FIG. 5, the computer device 12 is in the form of a general purpose computing device. Components of computer device 12 may include, but are not limited to: one or more processors or processing units 16, a system memory 28, a bus 18 that connects the various system components, including the system memory 28 and the processing units 16. Computer device 12 typically includes a variety of computer system readable media. Such media can be any available media that is accessible by device computer 12 and includes both volatile and nonvolatile media, removable and non-removable media. The system memory 28 may include computer system readable media in the form of volatile memory.
The computer device 12 may also communicate with one or more external devices 14 (e.g., keyboard, pointing device, display, etc.), one or more devices that enable a user to interact with the computer device 12, and/or any device that enables the computer device 12 to communicate with one or more other computing devices.
The processing unit 16 executes various functional applications and data processing by running a program stored in the system memory 28, for example, implements a CT image reconstruction method provided in embodiment 1 of the present invention, the method including:
s1, acquiring an initialization image according to projection data; s2, taking the initialized image as an image to be reconstructed, and carrying out image iterative processing on the image to be reconstructed; s3, judging whether an iteration stopping condition is met, if so, stopping iteration, and outputting an image after iteration processing; if not, returning to the step S2, and carrying out iterative processing again by taking the image after iterative processing as the image to be reconstructed. Wherein, the image iteration processing in S2 includes performing an image:
s21, updating, namely acquiring a maximized penalty likelihood function by calculating the image roughness based on the surface patches; s22, smoothing processing, namely performing smoothing denoising processing on the image from the S21; s23, performing fusion processing on the image from the S22, wherein the fusion processing comprises fusing the images pixel by pixel.
The embodiment applies a CT image reconstruction method to specific computer equipment, stores the method in the memory, and runs the method to reconstruct the CT image when the executor executes the memory, so that the method is fast and convenient to use and wide in application range.
Of course, those skilled in the art will understand that the processor may also implement the technical solution of the CT image reconstruction method provided in any embodiment of the present invention.
Example 4
Embodiment 4 provides a computer-readable storage medium having stored thereon a computer program which, when executed by a processor, implements the steps of a CT image reconstruction method as provided by any of the embodiments of the present invention, the method comprising:
s1, acquiring an initialization image according to projection data; s2, taking the initialized image as an image to be reconstructed, and carrying out image iterative processing on the image to be reconstructed; s3, judging whether an iteration stopping condition is met, if so, stopping iteration, and outputting an image after iteration processing; if not, returning to the step S2, and carrying out iterative processing again by taking the image after iterative processing as the image to be reconstructed. Wherein, the image iteration processing in S2 includes performing an image:
s21, updating, namely acquiring a maximized penalty likelihood function by calculating the image roughness based on the surface patches; s22, smoothing processing, namely performing smoothing denoising processing on the image from the S21; s23, performing fusion processing, namely performing region fusion processing on the image from S22, wherein the fusion processing comprises fusing the images pixel by pixel.
The computer storage media of embodiments of the invention may take the form of any combination of one or more computer-readable media. The computer readable medium may be a computer readable signal medium or a computer readable storage medium. The computer readable storage medium may be, for example, but not limited to: an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or a combination of any of the foregoing. More specific examples (a non-exhaustive list) of the computer-readable storage medium would include the following: an electrical connection having one or more wires, a portable computer diskette, a hard disk, a Random Access Memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing. In this document, a computer readable storage medium may be any tangible medium that can contain, or store a program for use by or in connection with an instruction execution system, apparatus, or device.
The embodiment applies a CT image reconstruction method to a computer readable storage medium, and a computer program is stored on the computer readable storage medium, and when the program is executed by a processor, the steps of the CT image reconstruction method provided by the invention are realized, so that the CT image reconstruction method is simple, convenient and quick, easy to store and difficult to lose.
In summary, the invention provides an image reconstruction method, an image reconstruction system, image reconstruction equipment and an image reconstruction medium, which solve the defects that the prior art generally has complex image reconstruction process, a great amount of noise and artifacts exist in reconstructed images, the edge retaining capability is poor, original characteristic textures in the images are not retained and the like, can eliminate the noise, inhibit the artifacts, retain the edges and fine characteristics, reduce the iteration times, simultaneously ensure the image quality, and have the characteristics of short reconstruction process, high image quality, comprehensive image characteristic retention and the like. The CT reconstruction method is applied to a specific system, computer equipment and a computer storage medium, and is materialized, so that the CT reconstruction method has important significance for the development of the medical influence field.
It will be appreciated by those of ordinary skill in the art that the modules or steps of the invention described above may be implemented in a general purpose computing device, they may be centralized on a single computing device, or distributed over a network of computing devices, or they may alternatively be implemented in program code executable by a computer device, such that they are stored in a memory device and executed by the computing device, or they may be separately fabricated as individual integrated circuit modules, or multiple modules or steps within them may be fabricated as a single integrated circuit module. Thus, the present invention is not limited to any specific combination of hardware and software.
Note that the above is only a preferred embodiment of the present invention and the technical principle applied. It will be understood by those skilled in the art that the present invention is not limited to the particular embodiments described herein, but is capable of various obvious changes, rearrangements and substitutions as will now become apparent to those skilled in the art without departing from the scope of the invention. Therefore, while the invention has been described in connection with the above embodiments, the invention is not limited to the embodiments, but may be embodied in many other equivalent forms without departing from the spirit or scope of the invention, which is set forth in the following claims.
The foregoing disclosure is merely illustrative of some embodiments of the invention, and the invention is not limited thereto, as modifications may be made by those skilled in the art without departing from the scope of the invention.
Claims (12)
1. A CT image reconstruction method comprising the steps of:
s1, acquiring an initialization image according to projection data;
s2, taking the initialized image as an image to be reconstructed, and carrying out image iterative processing on the image to be reconstructed;
s3, judging whether an iteration stopping condition is met, if so, stopping iteration, and outputting an image after iteration processing; if not, returning to the step S2, and carrying out iterative processing again by taking the image after iterative processing as the image to be reconstructed;
it is characterized in that the method comprises the steps of,
in the step S2, the image iterative process further includes the steps of:
s21, updating, namely solving the image roughness by calculating the intensity difference of the adjacent patches so as to obtain a maximized punishment likelihood function;
s22, smoothing processing, including smoothing denoising processing on the image from the S21;
s23, fusion processing, including carrying out region fusion processing on the image from the S22;
in the update processing of S21, the image roughness based on the patch is calculated by the following method;
the pixel j and the pixel k are respectively the center pixels of two adjacent patches, the image roughness is solved by calculating the intensity difference between the two patches, and then the maximum punishment likelihood function is obtained, and the pixel distance expression of the pixel j and the pixel k is as follows:
the image roughness expressions of the pixel j and the pixel k are:
wherein: f (f) j (mu) is a feature vector composed of all pixel intensity values of the patch centered at the pixel j, f k (mu) is a feature vector composed of all pixel intensity values of the patch centered at the pixel k, j l Represents the first pixel, k, in a patch centered on pixel j l Represents the first pixel, n, in a patch centered on pixel k j Identifying the total number of detectors, N j Indicating the number of all pixel points in the patch,h l is the positive weighting factor of the normalized inverse spatial distance from pixel to pixel, ψ (||f) j (μ)-f k (μ)‖ h ) Is a penalty function.
2. The CT image reconstruction method according to claim 1, wherein in the update processing of S21, the patches for calculating the image roughness include a square pixel region centered on one pixel composed of a plurality of pixels, and all the patches for calculating the image roughness are the same size.
3. The CT image reconstruction method according to claim 1, wherein in the update process of S21, the expression of the maximizing penalty likelihood function is:
Φ(μ)＝L(y|μ)-βU(μ)
wherein μ is a predicted image value, Φ (μ) is an objective function, L (y|μ) is a likelihood function, β is a regularization parameter controlling data fidelity and spatial smoothness, U (μ) is an image roughness, a maximized penalty likelihood function is obtained by solving for the image roughness, the image roughness expression is:
wherein, ψ (μ) p -μ q ) Is a penalty function, mu p -μ q Omega is the distance between pixel p and adjacent pixel q pq Is a weight factor of the distance between pixel p and pixel q in the neighborhood, n p Is the total number of detectors, N p Is the total number of image pixels.
4. The CT image reconstruction method as recited in claim 1, wherein the penalty function expression is:
wherein δ is the attenuation factor;
when f j (μ)-f k (μ)‖ h When the delta is less than the delta, the punishment function approximates a quadratic function;
when f j (μ)-f k (μ)‖ h When I > delta, the penalty function approximates an absolute function.
5. The CT image reconstruction method according to any one of claims 1 to 4, further comprising optimizing a transmission algorithm in the update process of S21;
the optimization transmission algorithm includes constructing a proxy function at each iteration, and minimizing the objective function Φ (μ).
6. The CT image reconstruction method as recited in claim 5, wherein the optimized transmission algorithm includes a expectation maximization algorithm, and wherein the optimization transmission algorithm is obtained according to the expectation maximization algorithm
Where n+1 is the number of iterations, j represents the center pixel sequence number of the patch, and EM represents the desired maximization.
7. The method of CT image reconstruction as recited in claim 6, wherein said step S22 includes smoothing the image from step S21 using a regularization method, involving a patch μ centered on pixel j j Is obtained according to the regularization method
Wherein Re g represents the regularization method.
8. The method of claim 7, wherein the merging of S23 includes merging the images from S22 into an image pixel by pixel, and obtaining the estimated image value at the n+1th iteration after merging pixel by pixel:
when β=0, the above equation reduces to the maximum likelihood expectation maximization equation.
9. A CT image reconstruction system comprising:
an initializing unit: the method comprises the steps of acquiring an initialization image according to projection data;
iteration unit: the initialization image is used as an image to be reconstructed, and the image to be reconstructed is subjected to iterative processing;
and a judgment output unit: the method comprises the steps of judging whether an iteration stopping condition is met, and if so, outputting an image after iteration processing; if not, sending the image after the iteration processing to the iteration unit, and carrying out the iteration processing again as an image to be reconstructed;
it is characterized in that the method comprises the steps of,
the iteration unit comprises an updating processing unit, a smoothing processing unit and a fusion processing unit;
the updating processing unit further comprises a roughness calculating unit, wherein the roughness calculating unit calculates image roughness based on the surface patch, and the image roughness based on the surface patch is calculated by the following method;
the pixel j and the pixel k are respectively the center pixels of two adjacent patches, the image roughness is solved by calculating the intensity difference between the two patches, and then the maximum punishment likelihood function is obtained, and the pixel distance expression of the pixel j and the pixel k is as follows:
the image roughness expressions of the pixel j and the pixel k are:
wherein: f (f) j (mu) is a feature vector composed of all pixel intensity values of the patch centered at the pixel j, f k (mu) is a feature vector composed of all pixel intensity values of the patch centered at the pixel k, j l Represents the first pixel, k, in a patch centered on pixel j l Represents the first pixel, n, in a patch centered on pixel k j Identifying the total number of detectors, N j Indicating the number of all pixel points in the patch, h l Is the positive weighting factor of the normalized inverse spatial distance from pixel to pixel, ψ (||f) j (μ)-f k (μ)‖ h ) Is a penalty function.
10. The CT image reconstruction system according to claim 9, wherein the update processing unit further comprises:
and the optimization transmission unit is used for constructing a proxy function when the image is iterated each time and carrying out minimization processing on the objective function phi (mu).
11. A computer device, the computer device comprising:
one or more processors;
a memory for storing one or more programs;
the one or more programs, when executed by the one or more processors, cause the one or more processors to implement the CT image reconstruction method as recited in any one of claims 1-8.
12. A computer readable storage medium, on which a computer program is stored, characterized in that the program, when being executed by a processor, implements the CT image reconstruction method as claimed in any of the claims 1-8.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010962224.6A CN112085808B (en) | 2020-09-14 | 2020-09-14 | CT image reconstruction method, system, equipment and medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010962224.6A CN112085808B (en) | 2020-09-14 | 2020-09-14 | CT image reconstruction method, system, equipment and medium |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN112085808A CN112085808A (en) | 2020-12-15 |
| CN112085808B true CN112085808B (en) | 2024-04-05 |
Family
ID=73737032
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010962224.6A Active CN112085808B (en) | 2020-09-14 | 2020-09-14 | CT image reconstruction method, system, equipment and medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112085808B (en) |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106056644A (en) * | 2016-05-24 | 2016-10-26 | 深圳先进技术研究院 | Data processing method and data processing device for CT scanning |
| CN106683144A (en) * | 2016-12-30 | 2017-05-17 | 上海联影医疗科技有限公司 | Image iteration reconstruction method and device |
| CN109685871A (en) * | 2018-12-14 | 2019-04-26 | 深圳先进技术研究院 | Image rebuilding method, computer equipment and computer readable storage medium |
| CN109712209A (en) * | 2018-12-14 | 2019-05-03 | 深圳先进技术研究院 | The method for reconstructing of PET image, computer storage medium, computer equipment |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US10621756B2 (en) * | 2017-01-13 | 2020-04-14 | Canon Medical Systems Corporation | Apparatus and method for correcting bias in low-count computed tomography projection data |
-
2020
- 2020-09-14 CN CN202010962224.6A patent/CN112085808B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106056644A (en) * | 2016-05-24 | 2016-10-26 | 深圳先进技术研究院 | Data processing method and data processing device for CT scanning |
| CN106683144A (en) * | 2016-12-30 | 2017-05-17 | 上海联影医疗科技有限公司 | Image iteration reconstruction method and device |
| CN109685871A (en) * | 2018-12-14 | 2019-04-26 | 深圳先进技术研究院 | Image rebuilding method, computer equipment and computer readable storage medium |
| CN109712209A (en) * | 2018-12-14 | 2019-05-03 | 深圳先进技术研究院 | The method for reconstructing of PET image, computer storage medium, computer equipment |
Also Published As
| Publication number | Publication date |
|---|---|
| CN112085808A (en) | 2020-12-15 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Wang et al. | PET image reconstruction using kernel method | |
| Rohkohl et al. | Improving best‐phase image quality in cardiac CT by motion correction with MAM optimization | |
| US9256967B2 (en) | Systems and methods for partial volume correction in PET penalized-likelihood image reconstruction | |
| WO2020118844A1 (en) | Reconstruction method for pet image, computer storage medium, and computer device | |
| US9595121B2 (en) | Systems and methods for PET penalized-likelihood image reconstruction with spatially varying smoothing parameters | |
| Dai et al. | Reconstruction of tomographic images from limited range projections using discrete Radon transform and Tchebichef moments | |
| Sun et al. | An iterative projection‐based motion estimation and compensation scheme for head x‐ray CT | |
| Sawatzky et al. | A variational framework for region-based segmentation incorporating physical noise models | |
| JP5590548B2 (en) | X-ray CT image processing method, X-ray CT program, and X-ray CT apparatus equipped with the program | |
| JPH10510080A (en) | Reconstruction processing of 3D images with improved contrast and resolution, and application of the processing to the production of attenuated plots | |
| Xu et al. | Statistical projection completion in X-ray CT using consistency conditions | |
| JP2007244871A (en) | Method for reconstructing image from sinogram and computer readable medium for storing computer program commands to perform the method and apparatus for reconstructing image from sinogram | |
| US20170340287A1 (en) | Method And Apparatus For Motion Correction In CT Imaging | |
| WO2022000192A1 (en) | Ct image construction method, ct device, and storage medium | |
| EP1861824A2 (en) | Method and device for the iterative reconstruction of tomographic images | |
| Hashemi et al. | Accelerated compressed sensing based CT image reconstruction | |
| Tiwari et al. | Computed tomography reconstruction on distributed storage using hybrid regularization approach | |
| Chen et al. | Joint-MAP tomographic reconstruction with patch similarity based mixture prior model | |
| Qiu et al. | New iterative cone beam CT reconstruction software: parameter optimisation and convergence study | |
| CN112085808B (en) | CT image reconstruction method, system, equipment and medium | |
| CN111968192A (en) | Construction method of CT image, CT device and storage medium | |
| Karimi et al. | A hybrid stochastic-deterministic gradient descent algorithm for image reconstruction in cone-beam computed tomography | |
| Mesquita et al. | Choosing the ART relaxation parameter for Clear-PEM 2D image reconstruction | |
| WO2022052114A1 (en) | Ct image reconstruction method and system, and device and medium | |
| US10515467B2 (en) | Image reconstruction system, method, and computer program |
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 |