CN115239558A - Low-dose lung CT image detail super-resolution reconstruction method and system - Google Patents
Low-dose lung CT image detail super-resolution reconstruction method and system Download PDFInfo
- Publication number
- CN115239558A CN115239558A CN202210846929.0A CN202210846929A CN115239558A CN 115239558 A CN115239558 A CN 115239558A CN 202210846929 A CN202210846929 A CN 202210846929A CN 115239558 A CN115239558 A CN 115239558A
- Authority
- CN
- China
- Prior art keywords
- image
- amplified
- lung
- super
- detail
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 210000004072 lung Anatomy 0.000 title claims abstract description 64
- 238000000034 method Methods 0.000 title claims abstract description 39
- 238000001914 filtration Methods 0.000 claims abstract description 39
- 238000012417 linear regression Methods 0.000 claims description 3
- 239000000758 substrate Substances 0.000 claims description 3
- 230000002685 pulmonary effect Effects 0.000 claims 1
- 238000003745 diagnosis Methods 0.000 abstract description 8
- 206010058467 Lung neoplasm malignant Diseases 0.000 abstract description 3
- 201000005202 lung cancer Diseases 0.000 abstract description 3
- 208000020816 lung neoplasm Diseases 0.000 abstract description 3
- 208000019693 Lung disease Diseases 0.000 description 9
- 238000005516 engineering process Methods 0.000 description 6
- 230000003321 amplification Effects 0.000 description 5
- 238000003199 nucleic acid amplification method Methods 0.000 description 5
- 238000004364 calculation method Methods 0.000 description 4
- 238000013459 approach Methods 0.000 description 3
- 230000009286 beneficial effect Effects 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 201000010099 disease Diseases 0.000 description 2
- 208000037265 diseases, disorders, signs and symptoms Diseases 0.000 description 2
- 230000002708 enhancing effect Effects 0.000 description 2
- 238000003384 imaging method Methods 0.000 description 2
- 230000001965 increasing effect Effects 0.000 description 2
- 230000001575 pathological effect Effects 0.000 description 2
- 230000005855 radiation Effects 0.000 description 2
- 238000010521 absorption reaction Methods 0.000 description 1
- 230000002146 bilateral effect Effects 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 239000000284 extract Substances 0.000 description 1
- 238000009499 grossing Methods 0.000 description 1
- 230000036541 health Effects 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 238000013507 mapping Methods 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 210000000056 organ Anatomy 0.000 description 1
- 238000012805 post-processing Methods 0.000 description 1
- 230000008569 process Effects 0.000 description 1
- 238000011084 recovery Methods 0.000 description 1
- 238000011524 similarity measure Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T3/00—Geometric image transformations in the plane of the image
- G06T3/40—Scaling of whole images or parts thereof, e.g. expanding or contracting
- G06T3/4053—Scaling of whole images or parts thereof, e.g. expanding or contracting based on super-resolution, i.e. the output image resolution being higher than the sensor resolution
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/0002—Inspection of images, e.g. flaw detection
- G06T7/0012—Biomedical image inspection
-
- 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
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30004—Biomedical image processing
- G06T2207/30061—Lung
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Health & Medical Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Medical Informatics (AREA)
- Nuclear Medicine, Radiotherapy & Molecular Imaging (AREA)
- Radiology & Medical Imaging (AREA)
- Quality & Reliability (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Apparatus For Radiation Diagnosis (AREA)
Abstract
A detail super-resolution reconstruction method and a system of a low-dose lung CT image comprise an image acquisition unit acquiring an original image of the lung CT image; the image guide unit obtains a detail image and a filtering image of the guide image through the guide image; the image zooming unit zooms in the guide image to obtain a magnified image; the coordinate acquisition unit sequentially acquires image blocks on the amplified image, and searches the most similar image block at the corresponding position of the filtered image to obtain the position coordinate of the most similar image block; and the image superposition unit takes out the image blocks at the corresponding positions on the detail image according to the position coordinates and superposes the image blocks on the amplified image to obtain a primary amplified image/a secondary amplified image/a tertiary amplified image, wherein the tertiary amplified image is a super-resolution image. The invention carries out reconstruction approximation on image details with different resolutions, reconstructs real image details, avoids pseudo details, restores high-resolution images, improves the resolution of low-dose lung CT images, enhances image details and improves the lung cancer diagnosis accuracy.
Description
Technical Field
The invention relates to the technical field of detail resolution of CT images, in particular to a super-resolution reconstruction method and a super-resolution reconstruction system for details of a low-dose lung CT image.
Background
On the basis of the existing CT equipment, two approaches are mainly used for improving the quality of lung CT images, namely increasing the radiation dose and the magnetic field intensity to enhance the absorption degree of organs to rays and electromagnetic waves so as to enable the equipment to directly generate high-quality images; and secondly, post-processing is carried out on the imaged medical image, so that the resolution of the original medical image is improved. The large amount of radiation dose and electromagnetic waves in the approach 1 undoubtedly bring great threat to human health, so the super-resolution reconstruction of medical images by the approach 2 is the most important way. The super-resolution reconstruction technology is utilized to generate a high-resolution image from the low-resolution image to a certain extent, so that the high-resolution image has clearer outline and richer texture information, and the position of a focus becomes easy to discover, thereby being beneficial to a clinician to reasonably reason the disease, further determining the cause of the disease and taking effective treatment measures.
At present, aiming at the problem, the traditional interpolation-based method can keep the continuity between the pixels of the original image and ensure that the interpolated image is naturally smooth; however, for the abrupt change region of the image, i.e. the pixel value describing the contour or edge of the image, it is necessary to smooth the pixel value of the portion, so that the contour and edge of the enlarged image become blurred, and the image quality is reduced.
Therefore, the problems of the prior art are to be further improved and developed.
Disclosure of Invention
The object of the invention is: aiming at the defects of the traditional interpolation algorithm, the invention provides a single CT image super-resolution method and a single CT image super-resolution system based on the combination of local self-similarity and a guided filtering technology.
The technical scheme is as follows: in order to solve the above technical problem, the present technical solution provides a low-dose lung CT image detail super-resolution reconstruction method, comprising the following steps,
s1, obtaining an original image of a lung CT image;
s2, obtaining a first detail image and a first filtering image of the original image through the original image;
s3, amplifying the original image to obtain a first amplified image;
s4, sequentially taking image blocks on the first amplified image, and searching the most similar image block at the corresponding position of the first filtered image to obtain the position coordinate of the most similar image block;
s5, extracting image blocks at corresponding positions on the first detail image according to the position coordinates, and superposing the image blocks on the first amplified image to obtain a first-stage amplified image;
s6, repeating the steps from S2 to S3 through the primary amplified image to obtain a secondary amplified image;
and S7, repeating the steps from S2 to S3 through the secondary amplified image to obtain a tertiary amplified image, namely a super-resolution image.
In the super-resolution reconstruction method for details of the low-dose lung CT image, in the step S2, a detail image and a filter image of an original image are obtained by using a guided filter technology, and the calculation is as follows,
in the formula:
to guide the image;
to output an image;
a window of pixel blocks;
，
to be a pixel point
Is a reference parameter.
The detail super-resolution reconstruction method of the low-dose lung CT image is defined according to a minimum mean square error criterion:
in the formula:
is the mean value of the guide image in the window;
to guide the variance of the image in the window;
the sum of the number of pixels in the window;
the pixel points in the image to be smoothed are the pixel points;
the average value of the image to be smoothed in the window is obtained;
is a linear regression coefficient.
The detail super-resolution reconstruction method for the low-dose lung CT image is characterized in that a large window and a small window are used as a guide filtering window and a coefficient window for calculation respectively, the guide filtering window is 9*9, and the coefficient window is 3*3.
The detail super-resolution reconstruction method for the low-dose lung CT image is characterized in that the formulas (2) and (3) can be replaced by the following formulas:
the detail super-resolution reconstruction method of the low-dose lung CT image is characterized in that one pixel point in the image is contained by a plurality of windows, and coefficients in different windows
，
Are different, taking coefficients in these windows
，
The average value of (a) is calculated,
，
respectively substitute in formula (2)
，
And obtaining a detail image of the original image by subtracting the filtered image from the original image.
In the step S3, an original image is enlarged by 1.25 times by using a bilinear interpolation method to obtain an enlarged image.
In the step S4, the most similar image blocks are searched around the corresponding position on the filtered image by utilizing the self-similarity of the local gray level image to obtain the position coordinates of the subblocks,
the specific algorithm is as follows, the position of the minimum absolute value difference of the filtering image block and the amplifying substrate image block is used as the best match, as shown in formula (1)
In the formula:
representing the gray value of a pixel point of a filtering image block;
expressing the gray value of pixel points of the amplified base image block; m, n represents the image block size, u and v represent pixel coordinates, and i and j represent relative coordinates of pixel points in the absolute difference image block.
A detail super-resolution reconstruction system of a low-dose lung CT image comprises an image acquisition unit, an image guide unit, an image scaling unit, a coordinate acquisition unit and an image superposition unit,
the image acquisition unit acquires an original image of a lung CT image;
the image guide unit obtains a detail image and a filtering image of the guide image through the guide image;
the image zooming unit zooms in the guide image to obtain a magnified image;
the coordinate acquisition unit sequentially acquires image blocks on the amplified image, and searches the most similar image block at the corresponding position of the filtered image to obtain the position coordinate of the most similar image block;
and the image superposition unit takes out the image blocks at the corresponding positions on the detail image according to the position coordinates and superposes the image blocks on the amplified image to obtain a primary amplified image/a secondary amplified image/a tertiary amplified image.
(III) the beneficial effects are as follows: the invention provides a super-resolution reconstruction method and a super-resolution reconstruction system for details of a low-dose lung CT image, which are used for performing reconstruction approximation on the details of images with different resolutions, reconstructing the details of a real image, effectively avoiding pseudo-details, reconstructing and restoring a high-resolution image to the maximum extent, improving the resolution of the low-dose lung CT image, enhancing the details of the image and improving the accuracy of lung cancer diagnosis.
Drawings
FIG. 1 is a schematic flow chart diagram of a preferred embodiment of a low-dose lung CT image detail super-resolution reconstruction method and system of the present invention;
FIG. 2 is a schematic diagram of a detailed super-resolution reconstruction method and system for a low-dose lung CT image according to the present invention;
FIG. 3 is an original low-dose lung CT image in a preferred embodiment of a low-dose lung CT image detail super-resolution reconstruction method and system of the present invention;
FIG. 4 is a super-resolution reconstructed image of a low-dose lung CT image in a preferred embodiment of the method and system for super-resolution reconstruction of details of the invention.
Detailed Description
The present invention will be described in further detail with reference to preferred embodiments, and more details are set forth in the following description in order to provide a thorough understanding of the present invention, but it is apparent that the present invention can be embodied in many other forms different from the description herein and can be similarly generalized and deduced by those skilled in the art based on the practical application without departing from the spirit of the present invention, and therefore, the scope of the present invention should not be limited by the contents of this detailed embodiment.
The drawings are schematic representations of embodiments of the invention, and it is noted that the drawings are intended only as examples and are not drawn to scale and should not be construed as limiting the true scope of the invention.
The invention is applied to the detail enhancement of the X-ray image in the CT, compared with the common image: the X-ray image of the CT image is a black and white image with less details, while the common image is visible light wave band imaging, colorful and more details, and the image characteristics of the common image and the visible light wave band imaging are completely different, so that the common image detail enhancement is not suitable for the X-ray image of the CT image, and the scheme is provided by the invention.
The following describes a preferred embodiment of a multi-scale detail enhancement method and system for low-dose lung CT images according to the present invention.
The method takes a rapid guided filtering technology and self-similar super-resolution detail reconstruction as an algorithm core, carries out reconstruction approximation on image details with different resolutions in 3 levels, fully utilizes the self-similar characteristic of a local image through reconstruction iteration of each time of smaller resolution improvement, reconstructs real image details, effectively avoids pseudo details, reduces the influence of noise on reconstruction through each guided filtering, and achieves the purpose of reconstructing and restoring a high-resolution image to the maximum extent, thereby obtaining the lung CT image with high resolution.
The method comprises the following specific steps:
the image acquisition unit acquires an original image of a lung CT image;
the image guiding unit obtains a first detail image and a first filtering image of the original image through the original image;
the image zooming unit magnifies the original image by using a bilinear interpolation method, wherein the magnification factor can be 1.25 times, and a first magnified image is obtained;
the coordinate acquiring unit sequentially acquires image blocks on the first amplified image, and searches for the most similar image block at the corresponding position of the first filtered image to obtain the position coordinate of the most similar image block;
the image superposition unit takes out image blocks at corresponding positions on the first detail image according to the position coordinates and superposes the image blocks on the first amplified image to obtain a first-stage amplified image;
then repeating the operation through the first-stage amplified image to obtain a second-stage amplified image; the above operation is repeated through the second-order magnified image to obtain a third-order magnified image, i.e. a super-resolution image, as follows,
the image guide unit obtains a second detail image and a second filtering image of the primary amplified image through the primary amplified image; the image zooming unit magnifies the first-stage magnified image by a bilinear interpolation method, wherein the magnification factor can be 1.25 times, and a second magnified image is obtained; the coordinate acquisition unit sequentially acquires image blocks on the second amplified image, and searches the most similar image block at the corresponding position of the second filtered image to obtain the position coordinate of the most similar image block; and the image superposition unit superposes the image block at the corresponding position on the second detail image on the second amplified image according to the position coordinate to obtain a second-level amplified image. The image zooming unit can be used for carrying out magnification on the original image, the first-stage magnified image and the second-stage magnified image, and determining the magnification of each stage of super-resolution reconstruction according to the super-resolution image reconstruction with the final required magnification.
The image guidance unit obtains the detail image and the filtered image through the original image, the first-level enlarged image or the second-level enlarged image, or both of them may be performed at the same time.
The image guiding unit obtains a third detail image and a third filtering image of the secondary amplified image through the secondary amplified image; the image zooming unit magnifies the second-level magnified image by a bilinear interpolation method, wherein the magnification factor can be 1.25 times, and a third magnified image is obtained; the coordinate acquiring unit sequentially acquires image blocks on the third amplified image, and searches the most similar image block at the corresponding position of the third filtered image to obtain the position coordinate of the most similar image block; and the image superposition unit superposes the image blocks at the corresponding positions on the third detail image on the third amplified image according to the position coordinates to obtain a three-level amplified image.
The image scaling unit may also magnify the image by 1.25 times using a cubic interpolation method.
The image guide unit obtains a detail image and a filtering image of an original image by utilizing a guide filtering technology, a guide filter has the functions of edge keeping and smooth filtering as well as a bilateral filter, and the specific calculation is as follows,
in the formula:
to guide the image;
is an output image;
is a pixel block window;
，
to be a pixel point
Is a reference parameter.
According to the minimum mean square error criterion, it is possible to define:
in the formula:
is the mean value of the guide image in the window;
to guide the variance of the image in the window;
the sum of the number of pixels in the window;
the pixel points in the image to be smoothed are set;
the average value of the image to be smoothed in the window;
the degree of smoothing of the filter is determined for the linear regression coefficients.
The invention improves the guiding filtering and adopts a large window and a small window as a guiding filtering window and a coefficient window to respectively calculate, greatly improves the calculating speed and increases the algorithm efficiency, the guiding filtering window is 9*9, the coefficient window is 3*3, the guiding image is an input image and specifically comprises an original image, a first-stage amplification image and a second-stage amplification image, and the formula (2) and the formula (3) can be changed into:
one pixel point in the image is contained by a plurality of windows, coefficients in different windows
，
Are different, so the coefficients in these windows are taken
，
The average value of (a) of (b),
，
respectively substitute in formula (2)
，
And obtaining output images which are respectively filtering images, and subtracting the filtering images from the original image to obtain a detail image of the original image.
The more commonalities and differences between images are, the higher the degree of similarity of images is. Therefore, commonality and dissimilarity are basic indicators of similarity measures between images. The image similarity is generally obtained by performing similarity index calculation on the characteristics such as color, texture, structure and the like representing the image information. The method for searching the most similar image block by adopting the images after the guiding filtering has two obvious advantages, namely noise interference is reduced, and image details can be better reserved by utilizing an algorithm, so that the precision of searching the most similar image is increased.
The coordinate acquiring unit sequentially takes small image blocks on the amplified base image, and searches the most similar image block around the corresponding position on the filtering image by using the self-similarity of the local gray image, and the specific operation is that the small image blocks are sequentially taken from the first amplified image/the second amplified image/the third amplified image, the local window image is 3*3, then the most similar subblock is searched by using the local gray self-similarity measurement standard in the window area (such as 5*5) of the first filtering image/the second filtering image/the third filtering image corresponding to the corresponding coordinate point of the first filtering image/the second filtering image/the third filtering image, so as to obtain the position coordinate of the subblock,
the specific algorithm is as follows, the position of the minimum absolute value difference of the filtering image block and the amplifying substrate image block is used as the best match, as shown in formula (1)
In the formula:
representing the gray value of a pixel point of a filtering image block;
expressing the gray value of pixel points of the amplified base image block; m, n denotes the image block size, u and v stand for imageThe pixel coordinates, i and j, represent the relative coordinates of the pixels in the absolute difference image block.
And after matching, the image superposition unit directly adds the image sub-blocks in the detail image corresponding to the most similar sub-block positions in the filtering image blocks to the corresponding amplification base image according to the obtained sub-block position coordinates to finish detail reconstruction of the amplification base image.
As can be seen from the images in FIGS. 2 and 3, the original low-dose lung CT image has low contrast, the image details are not well displayed, the image is masked by a gray mask, and after the super-resolution reconstruction of the image, the information such as the brightness, the contrast, the definition and the like of the image is greatly improved, so that the detail recovery of the invisible lung texture part in the original image can be obtained to a great extent.
The multi-scale detail enhancement system for the low-dose lung CT image further comprises a display unit, a selection unit and a storage unit. The storage unit is prestored with CT image examples of different types of lungs and CT reference images of healthy lungs. The selection unit searches homogeneous examples in the example lung CT image according to the obtained super-resolution image, and makes a diagnosis conjecture according to the examples for reference of a doctor. The display unit displays a super-resolution image, an exemplary lung CT image similar to the super-resolution image, and a diagnosis conjecture. The plurality of example lung CT images may be obtained by mapping image blocks in the initial example lung CT image and the super-resolution image.
The storage unit comprises different lung disease areas, each lung disease area at least comprises an initial example lung CT image, and the example lung CT images are correspondingly stored in the lung disease areas.
The multi-scale detail enhancement system for the low-dose lung CT image further comprises an example generating unit, wherein the example generating unit is used for copying the example lung CT image which is closest to the current super-resolution image in the storage unit and the initial example lung CT image of the lung disease region where the selected example lung CT image is located. Then, the example generation unit compares the current super-resolution image with the healthy lung CT reference image, extracts a pathological characteristic part line/image in the current super-resolution image, and respectively pastes the extracted pathological characteristic part line/image on the copied example lung CT image and the initial example lung CT image of the lung disease area where the example lung CT image is located to obtain two new example lung CT images.
The example generation unit stores the mapped example lung CT image, the initial example lung CT image and the current super-resolution image in the lung disease area where the selected example lung CT image is located, and updates the example image of the lung disease area.
The example generation unit automatically updates the example lung CT images in the storage unit, so that more example lung CT images are obtained, and further, when diagnosis estimation is carried out according to examples, the estimated diagnosis result is more accurate, so that misdiagnosis caused by insufficient experience of doctors is avoided, and the misdiagnosis rate is greatly reduced.
Each example lung CT image comprises a multi-angle example image, that is, each example lung CT image comprises a plurality of example images at different angles, wherein specific image angles of the plurality of example images at different angles in the example lung CT image at least comprise images at angles required by the current lung disease region for diagnosis of the lung disease.
A low-dose lung CT image detail super-resolution reconstruction method and a system thereof utilize a guide filtering technology to obtain a detail image and a filtering image of a lead-in image; amplifying the imported image by using a bilinear interpolation method to obtain an amplified base image; sequentially taking small image blocks on the amplified base image, and searching the most similar image block around the corresponding position on the filtered image by utilizing the self-similarity of the local gray image to obtain the position coordinate of the most similar image block; and finally, superimposing image blocks at corresponding positions on the detail image on the amplified base image by using the position coordinates to realize 1-level super-resolution reconstruction, repeating the process for 3 times, and realizing the super-resolution image reconstruction with required amplification factor, thereby effectively improving the resolution of the low-dose lung CT image, enhancing the image details and improving the lung cancer diagnosis accuracy.
The above description is provided for the purpose of illustrating the preferred embodiments of the present invention and will assist those skilled in the art in more fully understanding the technical solutions of the present invention. However, these examples are merely illustrative, and the embodiments of the present invention are not to be considered as being limited to the description of these examples. For those skilled in the art to which the invention pertains, several simple deductions and changes can be made without departing from the inventive concept, and all should be considered as falling within the protection scope of the invention.
Claims (9)
1. A detail super-resolution reconstruction method of a low-dose lung CT image is characterized by comprising the following steps,
s1, acquiring an original image of a lung CT image;
s2, obtaining a first detail image and a first filtering image of the original image through the original image;
s3, amplifying the original image to obtain a first amplified image;
s4, sequentially taking image blocks from the first amplified image, and searching the most similar image block at the corresponding position of the first filtered image to obtain the position coordinate of the most similar image block;
s5, extracting the image block at the corresponding position on the first detail image according to the position coordinate, and superposing the image block on the first amplified image to obtain a first-stage amplified image;
s6, repeating the steps from S2 to S3 through the primary amplified image to obtain a secondary amplified image;
and S7, repeating the steps from S2 to S3 through the secondary amplified image to obtain a tertiary amplified image, namely a super-resolution image.
2. The method for super-resolution reconstruction of low-dose lung CT image details as claimed in claim 1, wherein in step S2, the detailed image and the filtered image of the original image are obtained by using the guided filtering technique, and calculated as follows,
in the formula:
to guide the image;
is an output image;
a window of pixel blocks;
，
to be pixel points
Is a reference parameter.
3. The method of claim 2, wherein the detailed super-resolution reconstruction of the low-dose lung CT image is defined according to the minimum mean square error criterion:
in the formula:
is the mean value of the guide image in the window;
to guide the variance of the image in the window;
the sum of the number of pixels in the window;
the pixel points in the image to be smoothed are the pixel points;
the average value of the image to be smoothed in the window is obtained;
is a linear regression coefficient.
4. The method for the detail super-resolution reconstruction of the low-dose lung CT image according to claim 3, wherein the guiding filter uses a large window and a small window as a guiding filter window and a coefficient window to respectively calculate, wherein the guiding filter window is 9*9, and the coefficient window is 3*3.
5. The detail super-resolution reconstruction method of low-dose lung CT image according to claim 4, characterized in that the equations (2) and (3) can be replaced by the following equations:
6. the method of claim 5, wherein a pixel in the image is contained in multiple windows, and coefficients in different windows
，
Is different, taking coefficients in these windows
，
The average value of (a) of (b),
，
respectively substitute in formula (2)
，
And obtaining a detail image of the original image by subtracting the filtered image from the original image.
7. The method for super-resolution reconstruction of details of low-dose pulmonary CT image according to claim 1, wherein in step S3, the original image is enlarged by 1.25 times by using a bilinear interpolation method to obtain an enlarged image.
8. The method of claim 1, wherein in step S4, the local gray-scale image self-similarity is used to find the most similar image blocks around the corresponding positions on the filtered image to obtain the position coordinates of the sub-blocks,
the specific algorithm is as follows, the position of the minimum absolute value difference of the filtering image block and the amplifying substrate image block is used as the best match, as shown in formula (1)
In the formula:
representing the gray value of a pixel point of a filtering image block;
expressing the gray value of pixel points of the amplified base image block; m, n represents the image block size, u and v represent pixel coordinates, and i and j represent relative coordinates of pixel points in the absolute difference image block.
9. A low-dose lung CT image detail super-resolution reconstruction system is characterized by comprising an image acquisition unit, an image guide unit, an image zooming unit, a coordinate acquisition unit and an image superposition unit,
the image acquisition unit acquires an original image of a lung CT image;
the image guide unit obtains a detail image and a filtering image of the guide image through the guide image;
the image zooming unit zooms in the guide image to obtain a magnified image;
the coordinate acquisition unit sequentially acquires image blocks on the amplified image, and searches the most similar image block at the corresponding position of the filtered image to obtain the position coordinate of the most similar image block;
and the image superposition unit takes out the image blocks at the corresponding positions on the detail image according to the position coordinates and superposes the image blocks on the amplified image to obtain a primary amplified image/a secondary amplified image/a tertiary amplified image.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210846929.0A CN115239558A (en) | 2022-07-19 | 2022-07-19 | Low-dose lung CT image detail super-resolution reconstruction method and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210846929.0A CN115239558A (en) | 2022-07-19 | 2022-07-19 | Low-dose lung CT image detail super-resolution reconstruction method and system |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN115239558A true CN115239558A (en) | 2022-10-25 |
Family
ID=83674264
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210846929.0A Pending CN115239558A (en) | 2022-07-19 | 2022-07-19 | Low-dose lung CT image detail super-resolution reconstruction method and system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115239558A (en) |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103440676A (en) * | 2013-08-13 | 2013-12-11 | 南方医科大学 | Method for reconstruction of super-resolution coronary sagittal plane image of lung 4D-CT image based on motion estimation |
| US20170193635A1 (en) * | 2014-05-28 | 2017-07-06 | Peking University Shenzhen Graduate School | Method and apparatus for rapidly reconstructing super-resolution image |
| CN111899200A (en) * | 2020-08-10 | 2020-11-06 | 国科天成（北京）科技有限公司 | Infrared image enhancement method based on 3D filtering |
-
2022
- 2022-07-19 CN CN202210846929.0A patent/CN115239558A/en active Pending
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103440676A (en) * | 2013-08-13 | 2013-12-11 | 南方医科大学 | Method for reconstruction of super-resolution coronary sagittal plane image of lung 4D-CT image based on motion estimation |
| US20170193635A1 (en) * | 2014-05-28 | 2017-07-06 | Peking University Shenzhen Graduate School | Method and apparatus for rapidly reconstructing super-resolution image |
| CN111899200A (en) * | 2020-08-10 | 2020-11-06 | 国科天成（北京）科技有限公司 | Infrared image enhancement method based on 3D filtering |
Non-Patent Citations (1)
| Title |
|---|
| 史剑雄: "基于稀疏表示和引导滤波的图像超分辨率重建", 《中国优秀硕士学位论文全文数据库信息科技辑》, no. 01, 15 January 2019 (2019-01-15), pages 138 - 2662 * |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US20240037732A1 (en) | Method for enhancing quality and resolution of ct images based on deep learning | |
| CN110211193B (en) | Three-dimensional CT (computed tomography) interlayer image interpolation restoration and super-resolution processing method and device | |
| CN112116543B (en) | Image restoration method, system and device based on detection type generation framework | |
| CN114241077B (en) | CT image resolution optimization method and device | |
| CN113516586A (en) | Low-dose CT image super-resolution denoising method and device | |
| CN111091575A (en) | Medical image segmentation method based on reinforcement learning method | |
| Chen et al. | Double paths network with residual information distillation for improving lung CT image super resolution | |
| Yang et al. | Multi-modal medical image super-resolution fusion based on detail enhancement and weighted local energy deviation | |
| Li et al. | Unpaired low‐dose computed tomography image denoising using a progressive cyclical convolutional neural network | |
| CN114187181A (en) | Double-path lung CT image super-resolution method based on residual information refining | |
| CN109816747A (en) | A kind of metal artifacts reduction method of Cranial Computed Tomography image | |
| Hayat et al. | Combined channel and spatial attention-based stereo endoscopic image super-resolution | |
| Marker et al. | Contour-Based Surface Reconstruction using Implicit Curve Fitting, and Distance Field Filtering and Interpolation. | |
| JP7245740B2 (en) | Image processing device, image processing method and X-ray CT device | |
| CN111899166A (en) | Medical hyperspectral microscopic image super-resolution reconstruction method based on deep learning | |
| CN115239558A (en) | Low-dose lung CT image detail super-resolution reconstruction method and system | |
| Yan et al. | A multi-modal medical image fusion method in spatial domain | |
| CN116309435A (en) | Medical image specular reflection restoration method and system | |
| Li et al. | Dual-domain fusion deep convolutional neural network for low-dose CT denoising | |
| Zhao et al. | Perception-oriented generative adversarial network for retinal fundus image super-resolution | |
| Sun et al. | Clinical ultra‐high resolution CT scans enabled by using a generative adversarial network | |
| Yang | Medical application of digital image processing based on matlab | |
| CN116823660B (en) | Construction method, device and application of double-flow network model for CT image restoration | |
| CN114494498B (en) | Metal artifact removing method based on double-domain Fourier neural network | |
| CN113870137B (en) | Gradient-guided self-similarity-corrected low-dose CT image denoising method and system |
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 |