CN105354837A - X-ray digestive tract image resolution evaluation method based on human anatomic structure - Google Patents
X-ray digestive tract image resolution evaluation method based on human anatomic structure Download PDFInfo
- Publication number
- CN105354837A CN105354837A CN201510676576.4A CN201510676576A CN105354837A CN 105354837 A CN105354837 A CN 105354837A CN 201510676576 A CN201510676576 A CN 201510676576A CN 105354837 A CN105354837 A CN 105354837A
- Authority
- CN
- China
- Prior art keywords
- image
- edge
- ray
- curve
- image resolution
- 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.)
- Granted
Links
Classifications
-
- 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
- G06T2207/30168—Image quality inspection
Landscapes
- Engineering & Computer Science (AREA)
- Quality & Reliability (AREA)
- General Health & Medical Sciences (AREA)
- Medical Informatics (AREA)
- Nuclear Medicine, Radiotherapy & Molecular Imaging (AREA)
- Radiology & Medical Imaging (AREA)
- Health & Medical Sciences (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Apparatus For Radiation Diagnosis (AREA)
- Image Analysis (AREA)
- Image Processing (AREA)
Abstract
The purpose of the invention is to provide an X-ray digestive tract image resolution evaluation method based on a human anatomic structure. The X-ray digestive tract image resolution evaluation method comprises the following steps: (1) obtaining an area of interest; (2) carrying out image segmentation based on a threshold value; (3) extracting edges of an anatomic structure; (4) fitting an edge curve; (5) calculating edge smoothness. According to the X-ray digestive tract image resolution evaluation method, the edges of the anatomic structure can be automatically extracted, the smoothness can be automatically calculated and the image resolution evaluation can be quickly finished independent of subjective experience. In comparison with a method of employing a special evaluation model, the X-ray digestive tract image resolution evaluation method does not need to design a processing evaluation model, and the image resolution evaluation can be finished quickly, conveniently and exactly by only using the characteristics of the human anatomic structure. In comparison with a method of designing special low-dose X-ray imaging equipment, the X-ray digestive tract image resolution evaluation method is applied to the conventional X-ray imaging equipment.
Description
Technical field
What the present invention relates to is a kind of image processing method, specifically medical domain image processing method.
Background technology
X ray has been found, over more than 100 year, no matter in scientific research or life, be all widely used from roentgen in 1895.X-ray image is as the one of image, and the application in recent years in medical science, life science, military affairs, each industrial sector and engineering field presents the trend of sustainable growth.According to statistics, in medical diagnostic equipment, the application of X-ray occupies 60%.X-ray fluoroscopy and imaging are for a long time in hospital's widespread use.In all kinds of high precision medical equipment of hospital, medical X-ray diagnosis be apply the earliest, medical imaging means that clinical universal face is the widest.Not only the time shutter is short for x-ray, spatial resolution is high, and contains huge quantity of information in image.Therefore, although after the 1950's, other medical imaging device occurs in succession, and in the inspections such as bone, stomach and intestine, blood vessel and mammary gland, x-ray equipment inspection still has certain advantage.It substantially increases speed and the order of accuarcy of medical inspection, makes us more directly can understand the situation at some positions of health sooner.
X-ray has dual character.Biological cell can suffer damage even downright bad after a certain amount of roentgen radiation x.Utilize this effect of x-ray, tumor tissues can be destroyed by radiation therapy.But for the common person under inspection only doing X-ray routine inspection, X-ray irradiates the excessive damage that will cause normal structure, may produce stochastic effect, as carcinogenic and hereditary effect, and then impair one's health.Nowadays, people health see more and more important, therefore in enforcement x-ray, under guarantee image quality prerequisite, how to ensure that the safety problem of person under inspection is just more and more important.For preventing X-ray to the injury of human body, medical circle controls by reducing iatrogenic radiation limit value the radiation that patient suffers in x-ray, and carries out the imaging research of a large amount of low dosage X-ray.The low dosage research of current medical circle mainly concentrates on lung and checks, especially CT examination, and the research of the low dosage of common X-ray stays cool substantially.Particularly alimentary canal X-ray examination, due to its widespread use in the disease treatment such as acute abdomen and intestines and stomach process, therefore, the low dosage research for common X-ray is significant.
Belly imaging is the class that common X-ray imaging difficulty is larger.Picture quality affects due to factors such as the imaging systems by body fat thickness, imaging parameters and X-ray machine, and under identical imaging system, image quality also has larger difference.At present, the X-ray machine that hospital uses is all the parameter of manual adjustments ray tube usually, and in operation, doctor needs the size constantly changing parameter, to find the most clear, that diagnostic message is maximum image under low radiological dose prerequisite.And this dependence doctor experience often has certain subjectivity to image quality evaluation, be difficult to reach objective quantification.If operation doctor lacks experience, image is judged unclear; Or part machine " snowflake " is large, poor image quality, and radiologists needs long-time operation, causes person under inspection's irradiation time long, radiant quantity is large, can cause damage to the health of person under inspection.
Summary of the invention
The object of the present invention is to provide and automatically extract edge by image processing method, judge the smooth degree of image coboundary, complete a kind of X-ray alimentary canal Measurement for Digital Image Definition based on human anatomic structure of image quality evaluation fast, easily.
The object of the present invention is achieved like this:
A kind of X-ray alimentary canal Measurement for Digital Image Definition based on human anatomic structure of the present invention, is characterized in that:
(1) area-of-interest is obtained:
Choose a rectangle area-of-interest on image, make one of smooth edge section to be arranged in this rectangle, and preserve picture material in this rectangle;
(2) based on the Iamge Segmentation of threshold value:
In rectangular area interested in image, image is divided into two parts by smooth edge, namely to image or up and down or left and right or diagonal line segmentation, two parts are respectively prospect and the background of image;
Utilize histogram to obtain segmentation threshold, concrete grammar is as follows:
1) all for image in area-of-interest pixels divided into groups according to from 0 to 255 pixel values, every single order all calculates the number of pixels equaling this pixel value;
2) each rank number of pixels is added up, be designated as B
sum;
3) arranging background pixel number is B
bsum, will 0 be initialized as;
4) from 0 to 255, every single order pixel count is added to B
bsumin, until B
bsumbe greater than B
sum/ 2;
5) pixel value on current rank is recorded as segmentation threshold Thd;
(3) anatomical structure edge extracting:
By threshold value Thd by image binaryzation in area-of-interest, Morphological scale-space is carried out to the image after this binaryzation, comprise N continuous time to corrode and N continuous time expansion, image after Morphological scale-space is carried out edge extracting, adopt canny operator to detect, from the edge detected, select the boundary curve E of the longest edge of length as anatomical structure;
(4) boundary curve matching:
Utilize cubic spline curve to carry out matching to the edge extracted: the parametric cubic polynomial equation first getting any 2 half interval contours in edge E, then piecewise fitting is carried out to 2 half interval contours, finally process is smoothly connected to each curve and forms matched curve E
f;
(5) edge calculation smoothness:
Each point in edge curve E, calculates it to E
fminimum distance, this result is as evaluating the smooth degree of this curve, and equally also represent the sharpness of this image, concrete grammar is as follows:
1) get upper i-th point of curve E, use formula
calculate it to E
fupper each point (x
j, y
j| j=1,2,3 ...) distance;
2) E is extracted
fupper distance (x
i, y
i) closest approach (x
imin, y
imin), preserve this distance for d
imin;
3) to establish on E M point altogether, repeat above-mentioned 1), 2), calculate on E and arrive a little matched curve E
fbee-line, utilize formula
calculate mean distance and carry out evaluation edge smooth degree.
Advantage of the present invention is:
1, with use special evaluation model method compared with, the present invention without the need to designing job evaluation model, only use the characteristic of human anatomic structure just can quick, easy, complete image definition evaluation exactly.
2, compared with the method for design specialized low dosage X-ray imaging device, the present invention is applicable to traditional X-ray light imaging apparatus.
Accompanying drawing explanation
Fig. 1 a founds position plain film for human abdomen, and Fig. 1 b founds region of interest area image in the plain film of position for human abdomen;
Fig. 2 a is region of interest area image, and Fig. 2 b is histogram and segmentation threshold, and Fig. 2 c is the image of binaryzation, and Fig. 2 d is the image extracting edge;
Fig. 3 is curve-fitting results;
Fig. 4 is process flow diagram of the present invention.
Embodiment
Below in conjunction with accompanying drawing citing, the present invention is described in more detail:
Composition graphs 1 ~ 4, realization flow of the present invention is as follows:
1. obtain area-of-interest
The anatomical structure of human body natural can be demonstrated in common x-ray image.In these anatomical structures, there are some entire imaging regions, such as X-ray abdomen images, usually comprise smooth lung's lower edge.Utilizing the morphological character of these natural anatomical structures, extracting edge for judging image definition by image processing techniques.Therefore, to the first step of common x-ray image process, manually choose a rectangle area-of-interest on image, make one of smooth edge section to be arranged in this rectangle, and preserve picture material in this rectangle.
2. based on the Iamge Segmentation of threshold value
In rectangular area interested in image, image is divided into two parts of area approximation by smooth edge usually, namely to image or up and down or left and right or diagonal line segmentation.Two parts are respectively prospect and the background of image.For prospect and background area being separated, the present invention utilizes histogram to obtain segmentation threshold.Specific algorithm is as follows:
1) all for image in area-of-interest pixels divided into groups according to from 0 to 255 pixel values, every single order all calculates the number of pixels equaling this pixel value.
2) each rank number of pixels is added up, be designated as B
sum.
3) arranging background pixel number is B
bsum, will 0 be initialized as.
4) from 0 to 255, every single order pixel count is added to B
bsumin, until B
bsumbe greater than B
sum/ 2.
5) pixel value on current rank is recorded as segmentation threshold Thd.
3. anatomical structure edge extracting
Thd in step 2 as threshold value by image binaryzation in area-of-interest.Owing to there will be some little cavities in prospect after binaryzation, in background, there will be some fragmentary pixel clusters.Therefore, carry out Morphological scale-space to the image after this binaryzation, comprise N continuous time and corrode and N continuous time expansion, N value can be determined by experience.Image after Morphological scale-space is carried out edge extracting, in the present invention, adopts canny operator to detect.The boundary curve E of the longest edge of length as anatomical structure is selected from the edge detected.
4. boundary curve matching
Cubic spline curve is utilized to carry out matching to the edge extracted.First get the parametric cubic polynomial equation of any 2 half interval contours in edge E, then piecewise fitting is carried out to 2 half interval contours, finally process is smoothly connected to each curve and forms matched curve E
f.
5. edge calculation smoothness
Each point in edge curve E, calculates it to E
fminimum distance.This result, as the smooth degree evaluating this curve, equally also represents the sharpness of this image.Specific algorithm is as follows:
1) get upper i-th point of curve E, use formula (1) to calculate it to E
fupper each point (x
j, y
j| j=1,2,3 ...) distance.
2) E is extracted
fupper distance (x
i, y
i) closest approach (x
imin, y
imin), preserve this distance for d
imin.
3) to establish on E M point altogether, repetition 1), 2), calculate on E and arrive a little matched curve E
fbee-line.Utilize formula (2) to calculate mean distance and carry out evaluation edge smooth degree.
As follows to the inventive method checking:
1, configuration surroundings
Hardware comprises: (1) general X-ray machine; (2) common computer, for acquisition and processing image.Software comprises: (1) MATLAB; (2) self-compiling program of the present invention is realized.
2, x-ray image is obtained
X-ray machine is taken a width human abdomen and is found position plain film, is directly inputted in computing machine by DICOM interface by original image, sees Fig. 1 a.
3, image procossing
In MATLAB, use imread function, file path and filename are set, read in original X-ray DICOM image, by imshow function display original image.Use the getrect function of MATLAB on image, manually select lung's lower edge part as area-of-interest, see Fig. 1 b.Manually rectangle frame information is selected before utilization, use the image in the imcrop function extraction area-of-interest of MATLAB, see Fig. 2 a, calculate its grey level histogram by imhist function, and the method computed segmentation threshold value of foundation step 2 of the present invention, segmentation result is shown in Fig. 2 b.Based on this segmentation threshold, use the im2bw function of MATLAB by image binaryzation, see Fig. 2 c.Finally use the method for step 3 of the present invention, first carry out 5 times to image and corrode, expand for 5 times, finally use the edge function of MATLAB, operator is set to canny, extracts edge, sees Fig. 2 d.
4, curve
Use the method in step 4 of the present invention, two-dimensional curve matching is carried out to the edge extracted.Specific implementation is the function of the interp1 using MATLAB, and by the x of the frontier point of extraction, y coordinate is brought in function respectively, and wherein interpolation method uses " spline ", finally obtains matched curve, sees Fig. 3.
5, edge profile
Use the method in step 5 of the present invention and formula, calculate the degree of agreement of edge and the matched curve of extracting.Due to lung lower edge the smooth of the edge, therefore, the edge of extraction should coincide matched curve preferably.In this example, edge degree of agreement is 1.8528, and this value also represent the smooth degree of this curve, reflects the sharpness of present image to smooth edge imaging.
Claims (1)
1., based on an X-ray alimentary canal Measurement for Digital Image Definition for human anatomic structure, it is characterized in that:
(1) area-of-interest is obtained:
Choose a rectangle area-of-interest on image, make one of smooth edge section to be arranged in this rectangle, and preserve picture material in this rectangle;
(2) based on the Iamge Segmentation of threshold value:
In rectangular area interested in image, image is divided into two parts by smooth edge, namely to image or up and down or left and right or diagonal line segmentation, two parts are respectively prospect and the background of image;
Utilize histogram to obtain segmentation threshold, concrete grammar is as follows:
1) all for image in area-of-interest pixels divided into groups according to from 0 to 255 pixel values, every single order all calculates the number of pixels equaling this pixel value;
2) each rank number of pixels is added up, be designated as B
sum;
3) arranging background pixel number is B
bsum, will 0 be initialized as;
4) from 0 to 255, every single order pixel count is added to B
bsumin, until B
bsumbe greater than B
sum/ 2;
5) pixel value on current rank is recorded as segmentation threshold Thd;
(3) anatomical structure edge extracting:
By threshold value Thd by image binaryzation in area-of-interest, Morphological scale-space is carried out to the image after this binaryzation, comprise N continuous time to corrode and N continuous time expansion, image after Morphological scale-space is carried out edge extracting, adopt canny operator to detect, from the edge detected, select the boundary curve E of the longest edge of length as anatomical structure;
(4) boundary curve matching:
Utilize cubic spline curve to carry out matching to the edge extracted: the parametric cubic polynomial equation first getting any 2 half interval contours in edge E, then piecewise fitting is carried out to 2 half interval contours, finally process is smoothly connected to each curve and forms matched curve E
f;
(5) edge calculation smoothness:
Each point in edge curve E, calculates it to E
fminimum distance, this result is as evaluating the smooth degree of this curve, and equally also represent the sharpness of this image, concrete grammar is as follows:
1) get upper i-th point of curve E, use formula
calculate it to E
fupper each point (x
j, y
j| j=1,2,3 ...) distance;
2) E is extracted
fupper distance (x
i, y
i) closest approach (x
imin, y
imin), preserve this distance for d
imin;
3) to establish on E M point altogether, repeat above-mentioned 1), 2), calculate on E and arrive a little matched curve E
fbee-line, utilize formula
calculate mean distance and carry out evaluation edge smooth degree.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510676576.4A CN105354837B (en) | 2015-10-13 | 2015-10-13 | A kind of X-ray alimentary canal Measurement for Digital Image Definition based on human anatomic structure |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510676576.4A CN105354837B (en) | 2015-10-13 | 2015-10-13 | A kind of X-ray alimentary canal Measurement for Digital Image Definition based on human anatomic structure |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN105354837A true CN105354837A (en) | 2016-02-24 |
| CN105354837B CN105354837B (en) | 2017-12-19 |
Family
ID=55330804
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201510676576.4A Active CN105354837B (en) | 2015-10-13 | 2015-10-13 | A kind of X-ray alimentary canal Measurement for Digital Image Definition based on human anatomic structure |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN105354837B (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107424146A (en) * | 2017-06-28 | 2017-12-01 | 北京理工大学 | A kind of infrared polarization method for objectively evaluating image quality and system |
| CN109544459A (en) * | 2017-09-21 | 2019-03-29 | 中国电信股份有限公司 | Image sawtooth treating method and apparatus and computer readable storage medium |
| CN109727254A (en) * | 2018-11-27 | 2019-05-07 | 深圳市华讯方舟太赫兹科技有限公司 | Body scans image processing method, equipment and computer readable storage medium |
| CN110689947A (en) * | 2018-07-04 | 2020-01-14 | 天津天堰科技股份有限公司 | Display device and display method |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20110007334A1 (en) * | 2009-07-07 | 2011-01-13 | Xerox Corporation | Between-segment discontinuity reduction for text vectorization using dominant point classification |
| CN103473776A (en) * | 2013-09-17 | 2013-12-25 | 深圳市华因康高通量生物技术研究院 | Method and system for comparing image definition and automatic focusing control method |
| CN103679652A (en) * | 2013-11-29 | 2014-03-26 | 北京空间机电研究所 | Image restoration system capable of improving imaging quality greatly |
-
2015
- 2015-10-13 CN CN201510676576.4A patent/CN105354837B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20110007334A1 (en) * | 2009-07-07 | 2011-01-13 | Xerox Corporation | Between-segment discontinuity reduction for text vectorization using dominant point classification |
| CN103473776A (en) * | 2013-09-17 | 2013-12-25 | 深圳市华因康高通量生物技术研究院 | Method and system for comparing image definition and automatic focusing control method |
| CN103679652A (en) * | 2013-11-29 | 2014-03-26 | 北京空间机电研究所 | Image restoration system capable of improving imaging quality greatly |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107424146A (en) * | 2017-06-28 | 2017-12-01 | 北京理工大学 | A kind of infrared polarization method for objectively evaluating image quality and system |
| CN109544459A (en) * | 2017-09-21 | 2019-03-29 | 中国电信股份有限公司 | Image sawtooth treating method and apparatus and computer readable storage medium |
| CN110689947A (en) * | 2018-07-04 | 2020-01-14 | 天津天堰科技股份有限公司 | Display device and display method |
| CN109727254A (en) * | 2018-11-27 | 2019-05-07 | 深圳市华讯方舟太赫兹科技有限公司 | Body scans image processing method, equipment and computer readable storage medium |
| CN109727254B (en) * | 2018-11-27 | 2021-03-05 | 深圳市重投华讯太赫兹科技有限公司 | Human body scanning image processing method, human body scanning image processing equipment and computer readable storage medium |
Also Published As
| Publication number | Publication date |
|---|---|
| CN105354837B (en) | 2017-12-19 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108596887B (en) | Automatic segmentation method for liver tumor region image in abdominal CT sequence image | |
| Oo et al. | Brain tumor detection and segmentation using watershed segmentation and morphological operation | |
| Gan et al. | Toward accurate tooth segmentation from computed tomography images using a hybrid level set model | |
| Heutink et al. | Multi-Scale deep learning framework for cochlea localization, segmentation and analysis on clinical ultra-high-resolution CT images | |
| JP6656910B2 (en) | Medical image processing device, medical image diagnostic device, and medical image processing program | |
| Akram et al. | Artificial neural network based classification of lungs nodule using hybrid features from computerized tomographic images | |
| CN105354837A (en) | X-ray digestive tract image resolution evaluation method based on human anatomic structure | |
| Cockmartin et al. | Comparison of digital breast tomosynthesis and 2D digital mammography using a hybrid performance test | |
| Lai et al. | Automatic lung fields segmentation in CT scans using morphological operation and anatomical information | |
| van der Heyden et al. | Automated CT-derived skeletal muscle mass determination in lower hind limbs of mice using a 3D U-Net deep learning network | |
| Tseng et al. | An adaptive thresholding method for automatic lung segmentation in CT images | |
| Wei et al. | Inferior alveolar canal segmentation based on cone‐beam computed tomography | |
| Vieira et al. | Feasibility study of dose reduction in digital breast tomosynthesis using non-local denoising algorithms | |
| Jensen et al. | Improved liver lesion conspicuity with iterative reconstruction in computed tomography imaging | |
| Leyden et al. | Cross-correlation template matching for liver localisation in computed tomography | |
| Özkan et al. | Lung Segmentation algorithm for CAD system in CTA images | |
| Fu et al. | Segmentation of spinal canal region in CT images using 3D region growing technique | |
| Areeckal et al. | Fully automated radiogrammetric measurement of third metacarpal bone from hand radiograph | |
| Seong et al. | Segmentation of abnormal liver using adaptive threshold in abdominal CT images | |
| CN115410692A (en) | Apparatus and method for determining tissue boundaries | |
| Zhang et al. | Metal artifacts reduction for tomosynthesis | |
| Soares et al. | Assessing the impact of correlated noise in digital mammography: a virtual clinical trial | |
| Wu et al. | A fully automated multi-modal computer aided diagnosis approach to coronary calcium scoring of MSCT images | |
| Li et al. | Automatic heart positioning method in computed tomography scout images | |
| Mesanovic et al. | Application of lung segmentation algorithm to disease quantification from CT images |
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 |