CN105125228B - The image processing method that a kind of Chest X-rays DR images rib suppresses - Google Patents
The image processing method that a kind of Chest X-rays DR images rib suppresses Download PDFInfo
- Publication number
- CN105125228B CN105125228B CN201510651503.XA CN201510651503A CN105125228B CN 105125228 B CN105125228 B CN 105125228B CN 201510651503 A CN201510651503 A CN 201510651503A CN 105125228 B CN105125228 B CN 105125228B
- Authority
- CN
- China
- Prior art keywords
- image
- images
- pyramid
- chest
- rays
- 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
Landscapes
- Image Processing (AREA)
- Apparatus For Radiation Diagnosis (AREA)
Abstract
The invention discloses the image processing method that a kind of Chest X-rays DR images rib suppresses, comprise the following steps:Obtain Chest X-rays DR images;DR images are done into pyramid decomposition, down-sampled process is carried out and obtains Gaussian image pyramid S, a liter sampling process is carried out and obtains laplacian image pyramid disparity map D(s);Using the S of minimum as currently pending image I;Processing is filtered to image I with adjustable Gabor filter group, obtains reconstructed image R;Pending image I and reconstructed image R are asked poor, obtain weakening the result image E of line segment shape texture under the yardstick;By mono- times of result image E, laplacian image pyramid disparity map D (s) corresponding with the size is added, and repeats the processing procedure, until identical with original DR picture sizes, produces the image after rib suppresses;The present invention improves the vision significance of pulmonary shadow, reduces doctors'work burden, can realize and automatically process so that analytical conclusions are more objective and stably.
Description
Technical field
The present invention relates to medical information field, and in particular to the image processing method that a kind of Chest X-rays DR images rib suppresses
Method.
Background technology
DR is imaged and traditional fluoroscopy of chest is imaged by X-ray, is the Main Means of physical examination examination PUD D, such as
Lung inflammation, lump, tuberculosis etc.;But DR images are digitized videos, and imaging definition is high, and radiation is low, progressively substitutes in practice
Traditional fluoroscopy of chest imaging.The purpose of generally shooting DR pieces is to do pulmonary tuberculosis examination;Tuberculosis is drawn by mycobacterium tuberculosis
Rise, easily through the even aerosol transmission of the spittle in air;One infectiousness pulmonary tuberculosis patient, can averagely infect 10 in 1 year
To 15 people;Tuberculosis patient main body is between twenty and fifty, causes family and social labor power to be lost;The World Health Organization points out tuberculosis
It is the important public health problem in the whole world;In the whole world, annual tuberculosis captures 1,400,000 people's life;In China, there is 5,000,000 at present
Active tuberculosis patient, there are 50,000 people to die from tuberculosis every year, just had 1 people dead equivalent to every 10 minutes;According to world health
The estimation of tissue, the annual neopathy people 1,000,000 in China, the amplitude that falls progressively in year are 3%;In the high burden country of 22, whole world tuberculosis,
China is only second to India and comes second;Tuberculosis is one of Infectious Diseases of China's emphasis prevention and control, Chinese prevention from suffering from the diseases control
Point out that China is one of high burden country of 22, whole world tuberculosis and the high burden country of 27, whole world Drug resistant pulmonary tubeculosis in center processed
One of, Drug resistant pulmonary tubeculosis patient numbers rank the first in the world;Common lunger's number ranks the second in the world, and is only second to
India;Based on the pulmonary tuberculosis high incidence data of World Health Organization's issue, 67 countries including China are included it
In, the international development to China brings adverse effect.
Although China's pulmonary tuberculosis is presented, number of the infected is more, suffers from more than number of the infected and now the severe situation more than number,
Tuberculosis can prevent controlling in itself;By the promotion and international support of the Chinese government, country provide free diagnosis of tuberculosis and
Treatment;But because the pulmonary tuberculosis patient in China more than 80% is not so good as in rural area or floating population, the accessibility of medical services
Town dweller and non-current population, the compliance that patient receives long-term Canonical management is poor, and often therapeutic effect is bad;In view of tuberculosis
The harm of disease is serious, and preventing and controlling difficulty is big, and local tuberculosis prevention and treatment troops at different levels scale is still smaller, and strength and funds still can not
Preventing and treating demand is adapted to, it is necessary to strengthen technology and fund input, Combination between clinic and prevention mechanism is established, forms effective countermeasure system;
Currently, in the implementation of tuberculosis prevention and treatment, there is two weaknesses of early detection and Case management, difficult link;Firstly, since
Tuberculosis hides and the limitation of detection technique method, and diagnosis finds that patient relies primarily on patient and has symptom to go to see a doctor
Passive discovery mode;Although pulmonary tuberculosis, the seldom physical examination of patient subject crowd can be actively discovered by clapping X-ray by physical examination;
This just needs how research determines High risk group, targetedly carries out actively discovering work;At present still widely used
Sputum smear dyeing, the method for microexamination have used more than 130 years, and recall rate is low;Secondly, suffer from for the pulmonary tuberculosis made a definite diagnosis
Person's Canonical management management depigmentation rate is high, particularly difficult to multi-drug resistant tuberculosis therapy;The Canonical management compliance of patient is more
Difference, patient is largely lost in therapeutic process, and only minority can adhere to treating;Under this situation, domestic progressively application base
Key population pulmonary tuberculosis patient examination project is carried out in this public health service, and is all given free treating tuberculosis to making a definite diagnosis patient and controlled
Treat.
Chest X-rays DR pieces imaging difference under X ray using the different densities of tissue, to observe thickness and density difference
The lesion of smaller part position;But human tissue structure complexity, thoracic cavity and abdominal cavity contain the chief organ of human body, contain height
The various internal organs of density and low-density;Therefore its image mutually overlaps mutually, outstanding in the influence that DR Pian Shang lung's ribs are observed the lobe of the lung
Its is big;If it can weaken or eliminate rib image, with regard to that can be more beneficial for finding trickle lesion;In addition, the objective note of image data
Record is advantageous to the check contrast of diagnosis of disease;The roentgen dose X that patient receives perspective is also relatively bigger, the DR rabats based on X ray
Although checking the examination for being advantageous to the communicable diseases such as pulmonary tuberculosis, it has been undisputable fact that overfrequency Radiation On Human body is harmful;DR
Piece inspection is then a double-edged sword to human health;The standard formulated according to X ray according to ICRP, radiation are total
Risk factor is 0.0165/ sievert, and x-ray chest radiograph was shot less than the half second time, and exposure rate is about 0.045 mSv/second
(The mSv of 1 sievert=1000), it is very limited to the health risk of crowd.But sexual gland, eye lens, mammary gland in human body and
Thyroid gland is especially sensitive to ray, excessively frequently checks and is not beneficial to;Suppress and eliminate DR Pian Zhong lungs rib image pair
For examinee and sufferer, less rabat can be shot, doctor is obtained under the cost compared with low radiation dose focus is made a definite diagnosis.
The content of the invention
The present invention provides the image processing method that a kind of Chest X-rays DR images rib towards residents ' health physical examination suppresses.
The technical solution adopted by the present invention is:
The image processing method that a kind of Chest X-rays DR images rib suppresses, comprises the following steps:
Obtain Chest X-rays DR images;
DR images are done into pyramid decomposition, down-sampled process is carried out and obtains Gaussian image pyramid S, carry out a liter sampling process
Obtain laplacian image pyramid disparity map D(s);
Using the S of minimum as currently pending image I;
Processing is filtered to image I with adjustable Gabor filter group, obtains reconstructed image R;
Pending image I and reconstructed image R are asked poor, obtain weakening the result image of line segment shape texture under the yardstick
E;
Result image E is put and is twice, laplacian image pyramid disparity map D corresponding with the size
(s) it is added, repeats the processing procedure, until identical with original DR picture sizes, produces the image after rib suppresses.
Further, 5 layers of pyramid decomposition are done to DR images.
Further, DR images are constantly reduced into the 1/2 of original picture altitude on the basis of Gaussian smoothing, obtain one
Serial Gaussian image pyramid S;
Described liter of sampling process be specially:Each figure is enlarged into 2 times of original height, with original graph under gaussian pyramid
As asking poor, a series of laplacian image pyramid disparity map D are obtained(s).
Further, in the adjustable Gabor filter group, filter width and height are all 31 pixels, and direction
Include 16 directions from 0 to 180 °.
Further, the adjustable Gabor filter group is as follows to image I processing procedures:
Convolution is done to image I with Gabor filter group;
Convolution coefficient of each pixel under different directions is traveled through, records maximum and corresponding Gabor filter direction;
Directionality Gabor filter is superimposed according to corresponding to each pixel in maximum convolution coefficient value, obtains reconstruction image
R。
Further, after obtaining the image after rib suppresses, image is carried out to improve visual effect processing, processing procedure is such as
Under:
Lung areas profile is extracted, obtains the masking-out of lung areas;
Image is suppressed to lung's rib and does gray scale histogram equalization according to the pixels statisticses information inside masked area, is drawn
Rise gradation of image scope;
By the use of lung areas masking-out as Alpha figure layers, the image after tonal range will be drawn high and mixed with Alpha figure layers,
Obtain result.
Further, the profile of lung areas is extracted using GrabCut algorithms.
The beneficial effects of the invention are as follows:
(1)Interference of the invention by suppressing rib, so that the otherness of focus shade and the normal lobe of the lung turns into image
In leading difference, improve the vision significance of pulmonary shadow;
(2)The present invention can adapt to various Chest X-rays DR pieces, adapt to different figures and the age person of being taken, and realize full-automatic place
Reason;
(3)The present invention can effectively utilize Internet resources, the function of remote medical consultation with specialists can be achieved, so as to improve difficult and complicated illness
Consultation of doctors reliability;
(4)The present invention can be as the basis of computer-aided diagnosis, and the image after removal rib interference contributes to follow-up
Automate the design of lesion method of discrimination;
(5)The present invention can significantly reduce the live load of doctor, and improve overall recognition accuracy and treatment effeciency;
And the influence differentiated due to doctors experience deficiency and experience difference to the state of an illness can be reduced, make analytical conclusions more objective and steady
It is fixed;
(6)The present invention can be integrated with existing Medical Devices and informatization and network resource, be set without purchasing additional dedicated
It is standby, the completely compatible traditional approach of mode of operation, make migration work acceptant.The utilization rate for reducing equipment is improved simultaneously,
Avoid the idleness of equipment and the wasting of resources.
Brief description of the drawings
Fig. 1 is schematic flow sheet of the present invention.
Equipment connection diagram in Fig. 2 embodiments.
Embodiment
The present invention will be further described with specific embodiment below in conjunction with the accompanying drawings.
The image processing method that a kind of Chest X-rays DR images rib suppresses as shown in Figure 1, comprises the following steps:
Obtain Chest X-rays DR images;
DR images are done into pyramid decomposition, down-sampled process is carried out and obtains Gaussian image pyramid S, carry out a liter sampling process
Obtain laplacian image pyramid disparity map D(s);
Using the S of minimum as currently pending image I;
Processing is filtered to image I with adjustable Gabor filter group, obtains reconstructed image R;
Pending image I and reconstructed image R are asked poor, obtain weakening the result image of line segment shape texture under the yardstick
E;
Result image E is put and is twice, laplacian image pyramid disparity map D corresponding with the size
(s) it is added, repeats the processing procedure, until identical with original DR picture sizes, produces the image after rib suppresses.
Wherein the specific method of pyramid decomposition is recorded in " Burt, Peter and Adelson, Ted, " The
Laplacian Pyramid as a Compact Image Code", IEEE Trans. Communications, 9:4,
1983, 532-540.”;The processing method of adjustable Gabor filter group be recorded in " Fischer, S., Sroubek, F.,
Perrinet, L., Redondo, R. and Cristóbal, G. "Self invertible Gabor wavelets".
International Journal of Computer Vision, 75, 2007: 231-246.”。
Further, 5 layers of pyramid decomposition are done to DR images.
Further, the down-sampled process is specially:DR images are constantly reduced into original on the basis of Gaussian smoothing
Come the 1/2 of picture altitude, obtain a series of Gaussian image pyramid S;
Described liter of sampling process be specially:Each figure is enlarged into 2 times of original height, with original graph under gaussian pyramid
As asking poor, a series of laplacian image pyramid disparity map D are obtained(s).
Further, in the adjustable Gabor filter group, filter width and height are all 31 pixels, and direction
Include 16 directions from 0 to 180 °.
Further, the adjustable Gabor filter group is as follows to image I processing procedures:
Convolution is done to image I with Gabor filter group;
Convolution coefficient of each pixel under different directions is traveled through, records maximum and corresponding Gabor filter direction;
Directionality Gabor filter is superimposed according to corresponding to each pixel in maximum convolution coefficient value, obtains reconstruction image
R。
Further, after obtaining the image after rib suppresses, image is carried out to improve visual effect processing, processing procedure is such as
Under:
Lung areas profile is extracted, obtains the masking-out of lung areas;
Image is suppressed to lung's rib and does gray scale histogram equalization according to the pixels statisticses information inside masked area, is drawn
Rise gradation of image scope;
By the use of lung areas masking-out as Alpha figure layers, the image after tonal range will be drawn high and mixed with Alpha figure layers,
Obtain result.
Further, the profile of lung areas is extracted using GrabCut algorithms.
The circular of wherein GrabCut algorithms is recorded in " C. Rother, V. Kolmogorov, and A.
Blake, GrabCut: Interactive foreground extraction using iterated graph cuts,
In ACM Trans. Graph., vol. 23, pp. 309-314,2004. ".
Apply the present invention to basic scale physical examination point and serious infectious diseases examination, equipment connection diagram such as Fig. 2 institutes
Show, be mainly made up of computer automation processing function, Chest X-rays DICOM view data is digitized by reading, by data transfer
It is as follows to the multiple dimensioned extraction of lung's rib and suppression module, specific works step:
1)The computer for being equipped with pulmonary tuberculosis examination and lesion localization automation module is connected to by physical examination point staff
Medical image data server, and configure the parameter of DICOM image files reading;
2)System is according to time point, the examination of automatic running pulmonary tuberculosis and lesion localization automation module;
3)Pulmonary tuberculosis examination and lesion localization automation module accesses medical image data storehouse server, therefrom inquire about and do not divide
Analysis newly enters DICOM data;
4)System calls lung's rib suppression module, adaptively extracts the line segment dress line that rib is similar in lung images
Region is managed, and is targetedly suppressed;
5)System uses multiple dimensioned processing strategy, makes the rib of critical region by slightly to smart being suppressed and removing;
6)System calls lung outlines extraction module, and extraction may be the masking-out image of lung areas from lung image;
7)Using masking-out image as Alpha passages, enhancing lung remove image after rib to comparing and detail textures, and
Alpha Blending are with original image;
8)Exported fused images as system;
9)The GTG of fused images is converted into pseudo-colours simultaneously, also serves as system output, to prompt doctor to pay close attention to main points;
10)The subscriber terminal equipment that computer mark image and original image are used by network transmission to doctor, by
Doctor confirms mark correctness.If doctor thinks correct judgment, " confirmation " button is clicked directly on typing and is submitted in word
Hold;If doctor thinks to be out of one's reckoning, " modification " button is clicked on, to open the program interface manually marked;
11)In the period of the not subscriber terminal equipment service to doctor, system calls adaptive state update module automatically,
The parameter improvement of existing grader and in-depth are trained with the image texture content according to the new hand labeled of doctor.
Alpha Blending processing methods are recorded in " Wallace, Bruce. " Merging and
transformation of raster images for cartoon animation". SIGGRAPH Computer
Graphics 15 (3), 1981: 253–262.”。
More than in each step, system can prompt doctor to operate in a manner of patterned, by Computer Automatic Recognition and
Dynamic learning, operation keyboard and the frequency of mouse are needed to reduce doctor, so as to improve treatment effeciency and improve Consumer's Experience, made withered
Dry mark and checking work becomes easily to allow people to receive;In addition, system uses B/S framework, as long as making doctor have user name and close
Code can arbitrarily carry out mark and the assessment of pulmonary tuberculosis image on the computer of connection internet, make workbench from part
The dedicated network of change expands to wide area universal network;Not only it is beneficial to the work and coordination of doctor, and is advantageous to local bodyguard portion
Door and assurance and data analysis and excavation of the disease control unit to grass-roots work.
The present invention can aid in doctor to carry out pulmonary tuberculosis infectious disease screening;It is automatic using computer image processing technology
The rib region in DICOM view data is detected, and its pattern of suppression for being born from adaptation only includes soft tissue texture to generate
Effect image, to solve extensive resident's physical examination at present, to produce data volume too big, and doctor is in finite time hand inspection one by one
It is difficult to the problem for keeping high precision test;It utilizes the advantage of medical information, and adapting to supervisor's factor of healthcare givers causes
Deviation, the change of physical examination point, operating personnel computer level difference the problems such as.Whole processing procedure is simple and convenient, improves
The treatment effeciency of pulmonary tuberculosis examination, while the work load of physical examination point medical personnel is reduced, it is suitable for lacking pulmonary tuberculosis Chest X-rays
The basic medical unit of image quided experience, or even can be as the basic technology of the vehicle-mounted Chest X-rays examination solution of mobile.
Thus it is more beneficial for promoting for the further normalization of the extensive physical examination of resident and standardization of major infectious diseases.
The present invention is without manual intervention, various Chest X-rays DR images of processing that can be full-automatic;It can be dropped using new technical meanses
The workload of low medical personnel's desk checking Chest X-rays image, and lifting body focus Detection accuracy and serious infectious diseases monitoring effect
Rate, to work out the information that the relevant policies of the program decisions of the prevention and control of infectious disease and adjustment masses' hygiene and health provide preciousness
Basis;And using DICOM data as process object, based on existing medical imaging device and computer and internet, do not relate to
And the improvement of particular hardware.
The foregoing is merely illustrative of the preferred embodiments of the present invention, is not intended to limit the invention, all essences in the present invention
All any modification, equivalent and improvement made within refreshing and principle etc., should be included in the scope of the protection.
Claims (5)
1. the image processing method that a kind of Chest X-rays DR images rib suppresses, it is characterised in that comprise the following steps:
Obtain Chest X-rays DR images;
DR images are done into pyramid decomposition, down-sampled process is carried out and obtains a series of Gaussian image pyramid S, liter sampled
Journey obtains a series of laplacian image pyramid disparity map D (s);
Using the S of minimum as currently pending image I;
Processing is filtered to image I with adjustable Gabor filter group, obtains reconstructed image R;
Pending image I and reconstructed image R are asked poor, obtain weakening the result image E of line segment shape texture under the yardstick;
Result image E is put and is twice, laplacian image pyramid disparity map D (s) phase corresponding with the size
Add, repeat the processing procedure, until identical with original DR picture sizes, produce the image after rib suppresses;
The adjustable Gabor filter group is as follows to image I processing procedures:
Convolution is done to image I with Gabor filter group;
Convolution coefficient of each pixel under different directions is traveled through, records maximum convolution coefficient value and corresponding Gabor filter
Direction;
Directionality Gabor filter is superimposed according to corresponding to each pixel in maximum convolution coefficient value, obtains reconstructed image R;
After obtaining the image after rib suppresses, image is carried out to improve visual effect processing, processing procedure is as follows:
Lung areas profile is extracted, obtains the masking-out of lung areas;
Image is suppressed to lung's rib and does gray scale histogram equalization according to the pixels statisticses information inside masked area, draws high figure
As tonal range;
By the use of lung areas masking-out as Alpha figure layers, the image after tonal range will be drawn high and mixed with Alpha figure layers, obtained
Result.
2. the image processing method that a kind of Chest X-rays DR images rib according to claim 1 suppresses, it is characterised in that to DR
Image does 5 layers of pyramid decomposition.
3. the image processing method that a kind of Chest X-rays DR images rib according to claim 1 or 2 suppresses, it is characterised in that
The down-sampled process is specially:DR images are constantly reduced into the 1/2 of original picture altitude on the basis of Gaussian smoothing, obtained
To a series of Gaussian image pyramid S;
Described liter of sampling process be specially:Currently processed image is enlarged into 2 times of original height, with corresponding gaussian pyramid
Lower original image asks poor, obtains a series of laplacian image pyramid disparity map D (s).
4. the image processing method that a kind of Chest X-rays DR images rib according to claim 1 suppresses, it is characterised in that described
In adjustable Gabor filter group, filter width and height are all 31 pixels, and 16 sides that direction includes from 0 to 180 °
To.
5. the image processing method that a kind of Chest X-rays DR images rib according to claim 1 suppresses, it is characterised in that use
GrabCut algorithms extract the profile of lung areas.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201510651503.XA CN105125228B (en) | 2015-10-10 | 2015-10-10 | The image processing method that a kind of Chest X-rays DR images rib suppresses |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201510651503.XA CN105125228B (en) | 2015-10-10 | 2015-10-10 | The image processing method that a kind of Chest X-rays DR images rib suppresses |
Publications (2)
Publication Number | Publication Date |
---|---|
CN105125228A CN105125228A (en) | 2015-12-09 |
CN105125228B true CN105125228B (en) | 2018-04-06 |
Family
ID=54711076
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201510651503.XA Active CN105125228B (en) | 2015-10-10 | 2015-10-10 | The image processing method that a kind of Chest X-rays DR images rib suppresses |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN105125228B (en) |
Families Citing this family (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN106023200B (en) * | 2016-05-19 | 2019-06-28 | 四川大学 | A kind of x-ray chest radiograph image rib cage suppressing method based on Poisson model |
CN106408024B (en) * | 2016-09-20 | 2019-06-21 | 四川大学 | For the lobe of the lung contour extraction method of DR piece |
CN108510446B (en) * | 2018-04-10 | 2023-03-14 | 四川和生视界医药技术开发有限公司 | Method and device for superimposing retinal images |
Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN1729481A (en) * | 2002-12-18 | 2006-02-01 | 皇家飞利浦电子股份有限公司 | Method of processing an input image by means of multi-resolution |
CN101111864A (en) * | 2005-01-31 | 2008-01-23 | 皇家飞利浦电子股份有限公司 | Pyramidal decomposition for multi-resolution image filtering |
CN101241593A (en) * | 2007-02-06 | 2008-08-13 | 英特维数位科技股份有限公司 | Picture layer image processing unit and its method |
CN102646269A (en) * | 2012-02-29 | 2012-08-22 | 中山大学 | Image processing method and device based on Laplace pyramid |
CN104166974A (en) * | 2013-08-01 | 2014-11-26 | 上海联影医疗科技有限公司 | CT locating film image enhancing method and CT locating film image enhancing device |
Family Cites Families (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US7961925B2 (en) * | 2006-11-14 | 2011-06-14 | Siemens Aktiengesellschaft | Method and system for dual energy image registration |
TW200830217A (en) * | 2007-01-08 | 2008-07-16 | Corel Tw Corp | Image processing device and method by using differences of different scaled images as layered images |
-
2015
- 2015-10-10 CN CN201510651503.XA patent/CN105125228B/en active Active
Patent Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN1729481A (en) * | 2002-12-18 | 2006-02-01 | 皇家飞利浦电子股份有限公司 | Method of processing an input image by means of multi-resolution |
CN101111864A (en) * | 2005-01-31 | 2008-01-23 | 皇家飞利浦电子股份有限公司 | Pyramidal decomposition for multi-resolution image filtering |
CN101241593A (en) * | 2007-02-06 | 2008-08-13 | 英特维数位科技股份有限公司 | Picture layer image processing unit and its method |
CN102646269A (en) * | 2012-02-29 | 2012-08-22 | 中山大学 | Image processing method and device based on Laplace pyramid |
CN104166974A (en) * | 2013-08-01 | 2014-11-26 | 上海联影医疗科技有限公司 | CT locating film image enhancing method and CT locating film image enhancing device |
Non-Patent Citations (5)
Title |
---|
Background Modeling using Mixture of Gaussians and Laplacian Pyramid Decomposition;Minyong Wan,et al;《2011 International Conference of Soft Computing and Pattern Recognition(SoCPaR)》;20111231;第33-38页 * |
The Laplacian Pyramid as a Compact Image Code;Peter J.BURT,et al;《IEEE TRANSACTIONS ON COMMUNICATIONS》;19830430;第COM-31卷(第4期);第532-540页 * |
一种改进的多尺度医学图像增强算法;陈爱斌等;《Proceedings of the 32nd Chinese Control Conference》;20130728;第8252-8255页 * |
基于Gabor滤波器虚部的CT脑血管医学图像边缘特征提取方法;雷印胜等;《天津大学学报》;20070731;第40卷(第7期);第833-838页 * |
改进拉普拉斯金字塔模型的高动态图色调映射方法;梁云等;《计算机辅助设计与图形学学报》;20141231;第26卷(第12期);第2182-2188页 * |
Also Published As
Publication number | Publication date |
---|---|
CN105125228A (en) | 2015-12-09 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN106408024B (en) | For the lobe of the lung contour extraction method of DR piece | |
US12011310B2 (en) | Image quality compliance tool | |
CN106529555B (en) | DR (digital radiography) sheet lung contour extraction method based on full convolution network | |
CN104574316B (en) | A kind of digitized X-ray machine system | |
US10867375B2 (en) | Forecasting images for image processing | |
Na et al. | Guided filter‐based images fusion algorithm for CT and MRI medical images | |
US20190095760A1 (en) | Learning assistance device, method of operating learning assistance device, learning assistance program, learning assistance system, and terminal device | |
CN105125228B (en) | The image processing method that a kind of Chest X-rays DR images rib suppresses | |
KR20200137768A (en) | A Method and Apparatus for Segmentation of Orbital Bone in Head and Neck CT image by Using Deep Learning and Multi-Graylevel Network | |
Wang et al. | Wound image analysis system for diabetics | |
CN116703901B (en) | Lung medical CT image segmentation and classification device and equipment | |
CN111096835A (en) | Orthosis design method and system | |
CN111524109A (en) | Head medical image scoring method and device, electronic equipment and storage medium | |
CN105956386A (en) | Health indicator index classification system and method based on chest radiography of healthy people | |
Solves‐Llorens et al. | A complete software application for automatic registration of x‐ray mammography and magnetic resonance images | |
Xie et al. | Artificial intelligence–based computed tomography processing framework for surgical telementoring of congenital heart disease | |
Xie et al. | Contextual loss based artifact removal method on CBCT image | |
CN106228520B (en) | Image enchancing method and device | |
CN108038840A (en) | A kind of image processing method, device, image processing equipment and storage medium | |
EP3665643B1 (en) | X-ray image processing method and system and computer storage medium | |
He et al. | Medical image analysis of multiple myeloma based on convolutional neural network | |
WO2020209826A1 (en) | Deep reinforcement learning for computer assisted reading and analysis | |
Sadykov et al. | Computer diagnosis of tumors in mammograms | |
Gottschalk et al. | Deep learning based metal inpainting in the projection domain: Initial results | |
CN106127712A (en) | Image enchancing method and device |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
C10 | Entry into substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant |