CN105184741A - Three-dimensional CBCT (cone-beam computed tomography) image denoising method on the basis of improved nonlocal means - Google Patents
Three-dimensional CBCT (cone-beam computed tomography) image denoising method on the basis of improved nonlocal means Download PDFInfo
- Publication number
- CN105184741A CN105184741A CN201510468704.6A CN201510468704A CN105184741A CN 105184741 A CN105184741 A CN 105184741A CN 201510468704 A CN201510468704 A CN 201510468704A CN 105184741 A CN105184741 A CN 105184741A
- Authority
- CN
- China
- Prior art keywords
- projection
- data
- block
- dimensional cbct
- dimensional
- 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
Abstract
The present invention discloses a three-dimensional CBCT (cone-beam computed tomography) image denoising method on the basis of improved nonlocal means. The three-dimensional CBCT image denoising method on the basis of the improved nonlocal means comprises: obtaining projection data of three-dimensional CBCT images having different angles, projection data of a three-dimensional CBCT image having an angle corresponding to a set of projection data of the three-dimensional CBCT images, calculating edge information of the three-dimensional CBCT images, and dividing background subblocks and texture subblocks of the three-dimensional CBCT images; respectively calculating a noise standard deviation of a background region and a mean of the average gradient values of the texture subblocks in the three-dimensional CBCT images; respectively calculating filtering intensity values of the projection data of the three-dimensional CBCT images having different angles according to the edge information, the noise standard deviation of the background region and the mean of the average gradient values of the texture subblocks in the three-dimensional CBCT images; and searching for other pixel points in the three-dimensional CBCT images similar to filtered pixel points, and calculating the similarity among other pixel points similar to the filtered pixel points to achieve the three-dimensional CBCT image denoising according to the filtering intensity values of the projection data of the three-dimensional CBCT images having different angles.
Description
Technical field
The present invention relates to image processing field, particularly relating to a kind of three-dimensional CBCT scene image partition method based on improving non-local mean.
Background technology
When carrying out radiation therapy, because breathing, the wriggling of histoorgan, daily Set-up errors, target area contraction etc. exist larger impact to treatment plan, need to utilize various image documentation equipment to monitor in real time tumour, irradiation field is made tightly to follow target area to realize accurate treatment, therefore image guided radiation therapy (ImageGuidedRadiationTherapy, IGRT) also just becomes current state-of-the-art radiation therapy means.Pencil-beam Computed tomography technology (Cone-BeamComputedTomography, CBCT) be widely used in image guided radiation therapy system with its superior performance, its application is mainly reflected in following two aspects: one is rebuild image instruct and put position information with the drift condition in plan CT image radiotherapy district by obtaining three-dimensional CBCT; Two be by CBCT time-series image between deformable registration operation realize the real-time follow-up of pathology and tumour.
But there is a large amount of noises in CBCT image, it is fuzzy that the existence of noise makes that the soft tissue contrast of image reduces, image border becomes, and this adds the difficulty of drawing target outline undoubtedly, have impact on the Obtaining Accurate of pendulum position information.Although noise is caused by the decay of quantum greatly, the radiation of high dose can obtain image more clearly, also can have a strong impact on the health status of patient.So, under the irradiation of low dosage, how to obtain high-quality CBCT image in real time there is very important Research Significance.
In recent years, for the noise existed in CBCT image, multiple have the denoise algorithm about CBCT image to be in succession suggested, and comprises wavelet filtering method, non-local mean Denoising Algorithm, three-dimensional bits matching method, noise figure method etc.In the filtering method of these novelties, non-local mean (the NonlocalMeans that the people such as Buades propose, NLM) denoise algorithm obtains very large accreditation with its superior performance, this algorithm make use of the self-similarity of image, the global information of image has been taken into full account when estimating pixel, as much as possiblely in a search window search other pixels similar to filtered pixel, utilize the similarity between them to reach the effect of denoising.But this algorithm still exists Similarity Measure amount excessive and filtering parameter the problem such as to choose.
For calculated amount this problem excessive, multiple quick NLM denoise algorithm is in succession suggested and achieves good effect, and the people such as such as Wang utilize the symmetry of regional area distance to reduce the complexity of algorithm; The people such as Liu utilize integrogram and Fast Fourier Transform (FFT) to accelerate etc. NLM algorithm.In addition, NLM denoise algorithm contains three key parameters: search window radius, similar windows radius and filter strength value, and these three parameters all have the impact of larger impact, particularly filter strength value on the denoising performance of NLM algorithm.Lot of domestic and international researcher have also been made relevant research for the problem of parameter choose, but the scholar be applied in CBCT image is also few.
Summary of the invention
In order to solve the shortcoming of prior art, the invention provides a kind of three-dimensional CBCT scene image partition method based on improving non-local mean, the method realizes the denoising of self-adaptation non-local mean, to obtain the CBCT image of better quality carrying out to CBCT data for projection that noise criteria difference is estimated, on the basis of improving filter strength value.
For achieving the above object, the present invention is by the following technical solutions:
Based on the three-dimensional CBCT scene image partition method improving non-local mean, comprising:
Step (1): the data for projection obtaining the three-dimensional CBCT image of different angles, the data for projection of the corresponding one group of three-dimensional CBCT image of data for projection of the three-dimensional CBCT image of each angle, ask for the marginal information of three-dimensional CBCT image, and mark off background sub-block and the Streak block of three-dimensional CBCT image;
Step (2): calculate the average of Streak block average gradient value in the noise criteria difference of the middle background area of three-dimensional CBCT image and three-dimensional CBCT image respectively;
Step (3): according to the average of the marginal information of three-dimensional CBCT image, the noise criteria difference of background area and Streak block average gradient value, ask for the filter strength value of the data for projection of the three-dimensional CBCT image of each angle respectively;
Step (4): search other pixels similar to filtered pixel in three-dimensional CBCT image, according to the filter strength value of the data for projection of the three-dimensional CBCT image of each angle, calculate similarity between other similar pixels of filtered pixel to realize three-dimensional CBCT scene image partition.
The algorithm based on sub-block segmentation is adopted to carry out the background sub-block and the Streak block that divide three-dimensional CBCT image in described step (1).
Divide the background sub-block of three-dimensional CBCT image and the detailed process of Streak block, comprising:
The data for projection of three-dimensional CBCT image to be divided into size be several is the sub-block of square formation arrangement, calculates the variance of each sub-block;
According to the sub-block variance threshold values preset, carrying out dividing background sub-block and Streak block, when the variance of sub-block is less than default sub-block variance threshold values, is then background sub-block; When the variance of sub-block is greater than default sub-block variance threshold values, then it is Streak block.
The filter strength value of the data for projection of the three-dimensional CBCT image asked in described step (3), comprising:
The data for projection of selected one group of three-dimensional CBCT image, according to the noise criteria difference of the data for projection of the three-dimensional CBCT image of this group and the ratio of marginal information, obtains the filter strength value h of the data for projection of the three-dimensional CBCT image of this group
1:
Wherein, I
1represent one group of selected data for projection, it can be used as first group of data for projection; σ
1represent first group of data for projection I
1noise criteria poor, E (I
1) represent first group of shadow data I
1marginal information.
The filter strength value of the data for projection of the three-dimensional CBCT image asked in described step (3), also comprises:
Ask for the filter strength value h of the data for projection of i-th group of CBCT data for projection
i:
Wherein, h
i-1be the filter strength value of the data for projection of the i-th-1 group CBCT data for projection, G
ibe the average of Streak block average gradient value in i-th group of CBCT data for projection, G
i-
1be the average of Streak block average gradient value in the i-th-1 group CBCT data for projection, σ
ithe noise criteria being i-th group of CBCT data for projection is poor; σ
i-1the noise criteria being the i-th-1 group CBCT data for projection is poor; I=2,3,4 ... n, n are the group number obtaining CBCT data for projection.
Described default sub-block variance threshold values v
thfor:
Wherein,
for the mean value of all sub-block variances that the data for projection of three-dimensional CBCT image divides; α is proportionality constant.
The process of searching other pixels similar to filtered pixel in described step (4) is:
According to the average weighted method of all grey scale pixel values in three-dimensional CBCT image, obtain the gray-scale value of pixel in three-dimensional CBCT image:
In formula, I is three-dimensional CBCT image; Pixel p and pixel q are any pixel in three-dimensional CBCT image; V (q) represents the gray-scale value of pixel q; W (p, q) represents the similarity of pixel p and pixel q.
The similarity w (p, q) of pixel p and pixel q meets:
0≤w(p,q)≤1(8)
In formula, Z (p) represents normaliztion constant; N
pand N
qrepresenting the similar window of square centered by pixel p with pixel q respectively, window size is (2d+1) × (2d+1), d is similar windows radius; || ||
2, αfor Gauss's weighted euclidean distance function; h
jrepresent the filtering strength of jth group CBCT data for projection, j=1,2,3 ... n; N is the group number obtaining CBCT data for projection.
Beneficial effect of the present invention is:
(1) the present invention utilize that the noise criteria under different crevice projection angle is poor, the average of Streak block average gradient value and the marginal information of first group of data for projection, determine the filter strength value adapted with data for projection, the useful information effectively in retaining projection data also improves the denoising performance of NLM algorithm;
(2) the present invention is also according to the threshold value in dividing background region, accurately divides background area, further increases the validity of NLM algorithm, is conducive to follow-up pendulum position acquisition of information and tumour real-time follow-up.
Accompanying drawing explanation
Fig. 1 is the process flow diagram of three-dimensional CBCT scene image partition method of the present invention.
Embodiment
Below in conjunction with Figure of description and specific embodiment, the present invention is described:
As shown in Figure 1, the three-dimensional CBCT scene image partition method based on improving non-local mean of the present invention, comprising:
Step (1): the data for projection obtaining the three-dimensional CBCT image of different angles, the data for projection of the corresponding one group of three-dimensional CBCT image of data for projection of the three-dimensional CBCT image of each angle, ask for the marginal information of three-dimensional CBCT image, and mark off background sub-block and the Streak block of three-dimensional CBCT image;
Step (2): calculate the average of Streak block average gradient value in the noise criteria difference of the middle background area of three-dimensional CBCT image and three-dimensional CBCT image respectively;
Step (3): according to the average of the marginal information of three-dimensional CBCT image, the noise criteria difference of background area and Streak block average gradient value, ask for the filter strength value of the data for projection of the three-dimensional CBCT image of each angle respectively;
Step (4): search other pixels similar to filtered pixel in three-dimensional CBCT image, according to the filter strength value of the data for projection of the three-dimensional CBCT image of each angle, calculate similarity between other similar pixels of filtered pixel to realize three-dimensional CBCT scene image partition.
Further, the algorithm based on sub-block segmentation is adopted to carry out the background sub-block and the Streak block that divide three-dimensional CBCT image in step (1).Wherein, divide the background sub-block of three-dimensional CBCT image and the detailed process of Streak block, comprising:
The data for projection of three-dimensional CBCT image to be divided into size be several is the sub-block of square formation arrangement, calculates the variance of each sub-block;
According to the sub-block variance threshold values preset, carrying out dividing background sub-block and Streak block, when the variance of sub-block is less than default sub-block variance threshold values, is then background sub-block; When the variance of sub-block is greater than default sub-block variance threshold values, then it is Streak block.
The sub-block variance threshold values v preset
thfor:
Wherein,
for the mean value of all sub-block variances that the data for projection of three-dimensional CBCT image divides; α is proportionality constant.
In the present embodiment, the data for projection of three-dimensional CBCT image is divided into the sub-block that size is 10*10, calculates the mean value of all sub-block variances
Sub-block variance is sorted from high to low, calculates the mean value from 0.95*L sub-BOB(beginning of block) to the variance of L sub-block
wherein L is the sum of variance; According to the sub-block variance threshold values preset, carrying out dividing background sub-block and Streak block, when the variance of sub-block is less than default sub-block variance threshold values, is then background sub-block; When the variance of sub-block is greater than default sub-block variance threshold values, then it is Streak block; Wherein, proportionality constant α, is obtained by following formula:
Wherein, K=5.
The present invention adopts x ray generator to irradiate three-dimensional CBCT image, and x ray generator shines body around throwing and does 360 ° of rotation throwing photographs, and on average shine once with regard to throwing at interval of about 0.5 °, acquisition organizes CBCT data for projection more.The marginal information of adjacent projections data is more close.
Further, the filter strength value of the data for projection of the three-dimensional CBCT image asked in step (3), comprising:
The data for projection of selected one group of three-dimensional CBCT image, according to the noise criteria difference of the data for projection of the three-dimensional CBCT image of this group and the ratio of marginal information, obtains the filter strength value h1 of the data for projection of the three-dimensional CBCT image of this group:
Wherein, I
1represent one group of selected data for projection, it can be used as first group of data for projection; σ
1represent first group of data for projection I
1noise criteria poor, E (I
1) represent first group of shadow data I
1marginal information.
Further, the filter strength value of the data for projection of the three-dimensional CBCT image asked in step (3), also comprises:
Ask for the filter strength value hi of the data for projection of i-th group of CBCT data for projection:
Wherein, h
i-1be the filter strength value of the data for projection of the i-th-1 group CBCT data for projection, G
ibe the average of Streak block average gradient value in i-th group of CBCT data for projection, G
i-1be the average of Streak block average gradient value in the i-th-1 group CBCT data for projection, σ
ithe noise criteria being i-th group of CBCT data for projection is poor; σ
i-1the noise criteria being the i-th-1 group CBCT data for projection is poor; I=2,3,4 ... n, n are the group number obtaining CBCT data for projection.
Further, the process of searching other pixels similar to filtered pixel in step (4) is:
According to the average weighted method of all grey scale pixel values in three-dimensional CBCT image, obtain the gray-scale value of pixel in three-dimensional CBCT image:
In formula, I is three-dimensional CBCT image; Pixel p and pixel q are any pixel in three-dimensional CBCT image; V (q) represents the gray-scale value of pixel q; W (p, q) represents the similarity of pixel p and pixel q.
Further, the similarity w (p, q) of pixel p and pixel q meets:
0≤w(p,q)≤1(8)
In formula, Z (p) represents normaliztion constant; N
pand N
qrepresenting the similar window of square centered by pixel p with pixel q respectively, window size is (2d+1) × (2d+1), d is similar windows radius; || ||
2, αfor Gauss's weighted euclidean distance function; h
jrepresent the filtering strength of jth group CBCT data for projection, j=1,2,3 ... n; N is the group number obtaining CBCT data for projection.
As can be seen from above, the present invention is according to the average of the marginal information of three-dimensional CBCT image, the noise criteria difference of background area and Streak block average gradient value, ask for the filter strength value of the data for projection of the three-dimensional CBCT image of each angle respectively, obtain suitable filter strength value, reach denoising effect good, and retain detailed information in three-dimensional CBCT image.
By reference to the accompanying drawings the specific embodiment of the present invention is described although above-mentioned; but not limiting the scope of the invention; one of ordinary skill in the art should be understood that; on the basis of technical scheme of the present invention, those skilled in the art do not need to pay various amendment or distortion that creative work can make still within protection scope of the present invention.
Claims (8)
1., based on the three-dimensional CBCT scene image partition method improving non-local mean, it is characterized in that, comprising:
Step (1): the data for projection obtaining the three-dimensional CBCT image of different angles, the data for projection of the corresponding one group of three-dimensional CBCT image of data for projection of the three-dimensional CBCT image of each angle, ask for the marginal information of three-dimensional CBCT image, and mark off background sub-block and the Streak block of three-dimensional CBCT image;
Step (2): calculate the average of Streak block average gradient value in the noise criteria difference of the middle background area of three-dimensional CBCT image and three-dimensional CBCT image respectively;
Step (3): according to the average of the marginal information of three-dimensional CBCT image, the noise criteria difference of background area and Streak block average gradient value, ask for the filter strength value of the data for projection of the three-dimensional CBCT image of each angle respectively;
Step (4): search other pixels similar to filtered pixel in three-dimensional CBCT image, according to the filter strength value of the data for projection of the three-dimensional CBCT image of each angle, calculate similarity between other similar pixels of filtered pixel to realize three-dimensional CBCT scene image partition.
2. a kind ofly as claimed in claim 1 to it is characterized in that based on improving the three-dimensional CBCT scene image partition method of non-local mean, in described step (1), adopting the algorithm based on sub-block segmentation to carry out the background sub-block and the Streak block that divide three-dimensional CBCT image.
3. a kind of based on improving the three-dimensional CBCT scene image partition method of non-local mean as claimed in claim 2, it is characterized in that, divide the background sub-block of three-dimensional CBCT image and the detailed process of Streak block, comprising:
The data for projection of three-dimensional CBCT image to be divided into size be several is the sub-block of square formation arrangement, calculates the variance of each sub-block;
According to the sub-block variance threshold values preset, carrying out dividing background sub-block and Streak block, when the variance of sub-block is less than default sub-block variance threshold values, is then background sub-block; When the variance of sub-block is greater than default sub-block variance threshold values, then it is Streak block.
4. a kind of based on improving the three-dimensional CBCT scene image partition method of non-local mean as claimed in claim 1, it is characterized in that, the filter strength value of the data for projection of the three-dimensional CBCT image asked in described step (3), comprising:
The data for projection of selected one group of three-dimensional CBCT image, according to the noise criteria difference of the data for projection of the three-dimensional CBCT image of this group and the ratio of marginal information, obtains the filter strength value h of the data for projection of the three-dimensional CBCT image of this group
1:
Wherein, I
1represent one group of selected data for projection, it can be used as first group of data for projection; σ
1represent first group of data for projection I
1noise criteria poor, E (I
1) represent first group of shadow data I
1marginal information.
5. a kind of based on improving the three-dimensional CBCT scene image partition method of non-local mean as claimed in claim 1, it is characterized in that, the filter strength value of the data for projection of the three-dimensional CBCT image asked in described step (3), also comprises:
Ask for the filter strength value h of the data for projection of i-th group of CBCT data for projection
i:
Wherein, h
i-1be the filter strength value of the data for projection of the i-th-1 group CBCT data for projection, G
ibe the average of Streak block average gradient value in i-th group of CBCT data for projection, G
i-1be the average of Streak block average gradient value in the i-th-1 group CBCT data for projection, σ
ithe noise criteria being i-th group of CBCT data for projection is poor; σ
i-1the noise criteria being the i-th-1 group CBCT data for projection is poor; I=2,3,4 ... n, n are the group number obtaining CBCT data for projection.
6. a kind of based on improving the three-dimensional CBCT scene image partition method of non-local mean as claimed in claim 3, it is characterized in that, described default sub-block variance threshold values v
thfor:
Wherein,
for the mean value of all sub-block variances that the data for projection of three-dimensional CBCT image divides; α is proportionality constant.
7. a kind ofly as claimed in claim 1 it is characterized in that based on improving the three-dimensional CBCT scene image partition method of non-local mean, the process of searching other pixels similar to filtered pixel in described step (4) is:
According to the average weighted method of all grey scale pixel values in three-dimensional CBCT image, obtain the gray-scale value of pixel in three-dimensional CBCT image:
In formula, I is three-dimensional CBCT image; Pixel p and pixel q are any pixel in three-dimensional CBCT image; V (q) represents the gray-scale value of pixel q; W (p, q) represents the similarity of pixel p and pixel q.
8. a kind ofly as claimed in claim 1 it is characterized in that based on improving the three-dimensional CBCT scene image partition method of non-local mean, the similarity w (p, q) of pixel p and pixel q meets:
0≤w(p,q)≤1(8)
In formula, Z (p) represents normaliztion constant; N
pand N
qrepresenting the similar window of square centered by pixel p with pixel q respectively, window size is (2d+1) × (2d+1), d is similar windows radius; || ||
2, αfor Gauss's weighted euclidean distance function; h
jrepresent the filtering strength of jth group CBCT data for projection, j=1,2,3 ... n; N is the group number obtaining CBCT data for projection.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201510468704.6A CN105184741A (en) | 2015-08-03 | 2015-08-03 | Three-dimensional CBCT (cone-beam computed tomography) image denoising method on the basis of improved nonlocal means |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201510468704.6A CN105184741A (en) | 2015-08-03 | 2015-08-03 | Three-dimensional CBCT (cone-beam computed tomography) image denoising method on the basis of improved nonlocal means |
Publications (1)
Publication Number | Publication Date |
---|---|
CN105184741A true CN105184741A (en) | 2015-12-23 |
Family
ID=54906796
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201510468704.6A Pending CN105184741A (en) | 2015-08-03 | 2015-08-03 | Three-dimensional CBCT (cone-beam computed tomography) image denoising method on the basis of improved nonlocal means |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN105184741A (en) |
Cited By (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN106203522A (en) * | 2016-07-15 | 2016-12-07 | 西安电子科技大学 | Hyperspectral image classification method based on three-dimensional non-local mean filtering |
CN109767396A (en) * | 2019-01-04 | 2019-05-17 | 北京朗视仪器有限公司 | A kind of oral cavity CBCT image de-noising method based on image dynamic partition |
CN112381727A (en) * | 2020-11-03 | 2021-02-19 | 中航航空电子有限公司 | Image denoising method and device, computer equipment and storage medium |
CN113763275A (en) * | 2021-09-09 | 2021-12-07 | 深圳市文立科技有限公司 | Adaptive image noise reduction method and system and readable storage medium |
Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN103150712A (en) * | 2013-01-18 | 2013-06-12 | 清华大学 | Image denoising method based on projection sequential data similarity |
US20150178899A1 (en) * | 2012-09-05 | 2015-06-25 | Peking University | Non-local mean-based video denoising method and apparatus |
-
2015
- 2015-08-03 CN CN201510468704.6A patent/CN105184741A/en active Pending
Patent Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20150178899A1 (en) * | 2012-09-05 | 2015-06-25 | Peking University | Non-local mean-based video denoising method and apparatus |
CN103150712A (en) * | 2013-01-18 | 2013-06-12 | 清华大学 | Image denoising method based on projection sequential data similarity |
Non-Patent Citations (2)
Title |
---|
刘静等: "基于分块处理的三维非局部均值降噪算法", 《计算机工程》 * |
王公堂等: "基于自适应非局部均值的CBCT投影数据去噪算法", 《光电子.激光》 * |
Cited By (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN106203522A (en) * | 2016-07-15 | 2016-12-07 | 西安电子科技大学 | Hyperspectral image classification method based on three-dimensional non-local mean filtering |
CN109767396A (en) * | 2019-01-04 | 2019-05-17 | 北京朗视仪器有限公司 | A kind of oral cavity CBCT image de-noising method based on image dynamic partition |
CN112381727A (en) * | 2020-11-03 | 2021-02-19 | 中航航空电子有限公司 | Image denoising method and device, computer equipment and storage medium |
CN112381727B (en) * | 2020-11-03 | 2023-12-15 | 中航航空电子有限公司 | Image denoising method and device, computer equipment and storage medium |
CN113763275A (en) * | 2021-09-09 | 2021-12-07 | 深圳市文立科技有限公司 | Adaptive image noise reduction method and system and readable storage medium |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
Park et al. | Unpaired image denoising using a generative adversarial network in X-ray CT | |
CN108053417B (en) | lung segmentation device of 3D U-Net network based on mixed rough segmentation characteristics | |
Kearney et al. | An unsupervised convolutional neural network-based algorithm for deformable image registration | |
CN107123095B (en) | PET image reconstruction method and imaging system | |
Sheng et al. | Denoised and texture enhanced MVCT to improve soft tissue conspicuity | |
Xie et al. | Scatter artifacts removal using learning-based method for CBCT in IGRT system | |
CN103279929B (en) | A kind of prediction of the CT image metallic traces based on integral cosine and artifact minimizing technology | |
CN100508891C (en) | Maximum posteriori optimizing image rebuilding method in PET imaging | |
CN108520542B (en) | Reconstruction method for time phase matching of PET/CT data | |
CN103034989B (en) | A kind of low dosage CBCT image de-noising method based on high-quality prior image | |
CN103150712A (en) | Image denoising method based on projection sequential data similarity | |
CN105184741A (en) | Three-dimensional CBCT (cone-beam computed tomography) image denoising method on the basis of improved nonlocal means | |
EP3785222B1 (en) | Systems and methods for image processing | |
CN113256753A (en) | PET image region-of-interest enhancement reconstruction method based on multitask learning constraint | |
Yan et al. | Improved mask R-CNN for lung nodule segmentation | |
Van Der Heyden et al. | A Monte Carlo based scatter removal method for non-isocentric cone-beam CT acquisitions using a deep convolutional autoencoder | |
CN102024267A (en) | Low-dose computed tomography (CT) image processing method based on wavelet space directional filtering | |
CN103226815A (en) | Low dose CT image filtering method | |
CN111260583A (en) | Multi-discriminant-based multi-analysis network missing CT projection data estimation method | |
Liao et al. | Noise estimation for single-slice sinogram of low-dose X-ray computed tomography using homogenous patch | |
CN109285176A (en) | A kind of cerebral tissue dividing method cut based on regularization figure | |
CN105488460A (en) | Physiological feature based image processing method | |
CN104091356A (en) | X-ray medical image objective reconstruction based on independent component analysis | |
Shi et al. | Fast shading correction for cone-beam CT via partitioned tissue classification | |
CN106447650B (en) | Human thyroglobulin method for measuring weight based on SPECT planar imaging |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
C10 | Entry into substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20151223 |
|
RJ01 | Rejection of invention patent application after publication |