CN102429679A - Computer-assisted emphysema analysis system based on chest CT (Computerized Tomography) image - Google Patents
Computer-assisted emphysema analysis system based on chest CT (Computerized Tomography) image Download PDFInfo
- Publication number
- CN102429679A CN102429679A CN2011102671651A CN201110267165A CN102429679A CN 102429679 A CN102429679 A CN 102429679A CN 2011102671651 A CN2011102671651 A CN 2011102671651A CN 201110267165 A CN201110267165 A CN 201110267165A CN 102429679 A CN102429679 A CN 102429679A
- Authority
- CN
- China
- Prior art keywords
- lung
- emphysema
- image
- module
- district
- 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
Images
Landscapes
- Apparatus For Radiation Diagnosis (AREA)
Abstract
The invention discloses an automatic lung area partitioning and computer-assisted emphysema analysis system based on a chest CT (Computerized Tomography) image. A computer-assisted emphysema analysis method based on the chest CT image comprises the following steps of: firstly, inputting a group of chest CT-series images into a system; secondly, automatically partitioning the lung area according to the following three steps of: (1) separating a trunk from a background by using an automatic threshold value; (2) extracting the profile of the lung area by applying a profile tracking method; and (3) extracting the lung area by using a boundary scanning and region filling method; and lastly, counting and analyzing the lung area by applying an emphysema quantitative diagnosis standard based on density distribution and volume fraction, displaying a statistical characteristic value for a radiology department doctor as required, determining and displaying a pathological change region with highlight, and grading the pathological change degree to realize quantitative analysis and accurate diagnosis of the emphysema. By adopting the system, the emphysema diagnosis accuracy and efficiency of the radiology department doctor can be increased, and the radiology department doctor can be assisted to perform clinical diagnosis and draw up a treatment scheme more objectively and effectively.
Description
Technical field
The invention belongs to computer analysis The Application of Technology field, be specifically related to a kind of emphysema computer-aided diagnosis system based on the breast CT image based on medical image.
Background technology
Chronic obstructive pulmonary disease is a kind ofly to be limited as the disease of characteristic with expiratory airflow, comprises chronic bronchitis and emphysema.Chronic obstructive pulmonary disease can cause pulmonary dysfunction, pulmonary hypertension, develops into to a certain degree promptly to produce hypoxemia, occurs hypercapnia and respiratory failure subsequently, causes death.According to statistics, China is chronic obstructive pulmonary disease prevalence 9% in recent years, has the chronic obstructive pulmonary disease number of the infected now up to 5,000 ten thousand, annual death toll 1,300,000, and chronic obstructive pulmonary disease has become the fourth-largest cause of disease because of disease death.
For the multiple pathological changes in patient's lung, comparatively complicated on the performance characteristic of iconography and scope, even experienced radiologist, to its do objective, analyze and diagnosis also suitable difficulty exactly.At present, both at home and abroad the CT image sign research of chronic obstructive pulmonary disease being reported morely, but mainly concentrated on the qualitative aspect of pathological changes, is that the research of quantitative aspect is less for the degree of pathological changes.Rely on visually-perceptible and the experience that extensively adopt are at present carried out method of diagnosing; Mainly be with the low density area that occurs in the lung differing in size, the pulmonary vascular markings reason is sparse and signs such as vessel branch distortion as foundation, mark or diagnose according to emphysema scope and the order of severity.Not enough below this diagnostic method qualitatively exists:
Depend on individual's perception, experience and professional ability, diagnostic result can vary with each individual;
The radiologist to read the sheet working strength big, fatiguability influences work efficiency and quality;
Can't carry out quantitative analysis and disease is carried out accurate classification;
Be unfavorable for treating or the accurate evaluation and the tracking of postoperative curative effect.
To the problem in the above-mentioned clinical etiologic diagnosis, image application is learned processing method, and the lung district is carried out quantitative analysis; Auxiliary doctor makes diagnosis; Qualitative and quantitative diagnosis advantage separately can be made full use of, on the one hand, radiologist's diagnostic work amount and labor intensity can be reduced; On the other hand, can improve accuracy, reliability and the efficient of diagnosis.Emophysematous calculating auxiliary diagnosis can be divided into for two steps, at first was accurately to extract the [district, then, according to emophysematous clinical quantification diagnostic criteria, the lung district was carried out quantitative analysis, drew diagnosis.
Lung district automatic division method mainly contains in the breast CT image at present:
(1) threshold method
O.Osman; S.Sahin, " Lung Segmentation Algorithm for CAD System in CTA Images ", World Academy of Science; Engineering and Technology; 77,306-309 (2011), and B.S.Zhao; Y.David; " Two-dimensional multi-criterion segmentation of pulmonary nodules on helical CT images ", Medical Physics, 26 (6): 889-895 (1999) etc.).Be characterized in simple, fast, but can not effectively remove zones such as trunk outside and trachea, bronchus, and definite threshold value is difficult, often rule of thumb confirms.
(2) region growing.Though region-growing method can keep the zone with diffusivity border, those structures of being surrounded by strong gradient boundaries often are left out, and it is also responsive to choosing of seed points and growth merging rule.Because region-growing method is a kind of artificial semi-automatic partition method of participating in that needs, therefore, its application receives bigger restriction, and (poplar adds; Wu prays credit; Tian Jie etc., " several kinds of realization and the comparisons of image segmentation algorithm on the CT image segmentation, " Beijing Institute of Technology's journal; 20 (6): 720-724, (2000)).
(3) based on the method (S.Sun that adds up prior model; G.McLennan; E.A.Hoffman, et al. " Model-Based Segmentation of Pathological Lungs in Volumetric CT Data, " The Third International Workshop on Pulmonary Image Analysis; 31-40, (2010)).It sets up prior model through collecting great amount of samples, adopts some coupling and deformation method, extracts lung district profile.Its advantage is to utilize the shape and the density information of sample, but sets up comparatively difficulty of model, and the some coupling is more with the deformation process spended time, is difficult to satisfy the real-time requirement of clinical practice.
(4) based on method (H.Wang, J.Zhang, the L.Wang of pattern classification; " Segmentation of thoracic CT image based on FCM clustering, " IEEE 2010 3rd International Conference on Advanced Computer Theory and Engineering (ICACTE), V3; 405-408 (2010); F.Monteiro, " Region-Based Clustering for Lung Segmentationin Low-Dose CT Images, " International Conference of Numerical Analysis and Applied Mathematics (ICNAAM); 1281,2061-2064 (2010)).These class methods are extracted effective characteristics of image, a large amount of training sample of also needs that has, and segmentation result is very strong to the dependency of sample and characteristic, and its processing time is longer.
In sum, the dividing method in lung district in the existing breast CT image, or model and computing complicacy, splitting speed is slower; Or be difficult to confirm the control parameter, and it is lower with reliability to cut apart stability, and segmentation result is inaccurate.Promptly be difficult to extract quickly and accurately lung district profile and pulmonary parenchyma, can't satisfy the requirement of computer-aided diagnosis system.
Summary of the invention
The objective of the invention is to overcome the shortcoming and defect of above-mentioned prior art, provide a kind of lung district to cut apart automatically and the emphysema computer-aided diagnosis system based on the breast CT image.
The object of the invention is realized through following technical proposals: a kind of emphysema computer-aided diagnosis system based on the breast CT image comprises:
Input module (100) is used to import breast CT image to be diagnosed, and sends extraction lung district's module (200) to;
Extract lung district's module (200), be used for cutting apart automatically the [district, and send lung district information to quantization parameter computing module (300);
Quantization parameter computing module (300) is used to calculate the statistical distribution information of the picture element density of lung district or appointed area, and geological information, and sends quantization parameter to classification diagnosis module (400) and output module (500);
Classification diagnosis module (400) is used for analyzing the data that quantize parameter calculating module (300) transmission, and analysis result is sent to output module (500);
Output module (500), with the analysis result of classification diagnosis module (400), what be positioned user input waits to diagnose the breast CT image, with particular color labelling shadow of doubt, and analysis result is shown to the user.
Extracting lung district's module (200) handles according to following step:
Step (2.1) adopts overall adaptive threshold method to separate trunk and background in the breast CT image: an at first given initial threshold, and use this threshold value image is divided into two types; Ask the average of two average densitys then and, again image is classified,, make the threshold approaches optimal value gradually, obtain threshold value accurately at last, background is separated with trunk through iterative algorithm as new threshold value;
Step (2.2) adopts contour tracing method to extract lung district profile: a pixel that at first detects a lung profile according to certain detection criterion; The certain tracking criterion of reuse is found out other pixels of objective contour, up to finding whole all pixels of lung profile; Look for all pixels of another lung profile again;
Step (2.3) adopts a kind of context marker scanning line method based on 4 neighborhoods to obtain the pixel in the [zone.
Initial threshold described in the step (2.1) is selected entire image density meansigma methods.
Said step (2.3) comprises the steps:
Step (2.3.1) is asked the boundary rectangle of area-of-interest, and generation can cover the minimum rectangular area of selection area;
Step (2.3.2) is labeled as " 1 " with the point of lung in the rectangular area, and other then is labeled as " 0 ";
Step (2.3.3) by from top to bottom, order from left to right, scan rectangle zone is if current pixel is labeled as " 0 "; Then at current line; Begin to scan from left to right from current pixel, and put the pixel of process be labeled as " 1 ", finish up to point or this every trade end;
4 neighborhoods of step (2.3.4) search current pixel find a point that is labeled as " 0 ", and as new starting point, scanning from left to right, and put through pixel and be labeled as " 1 " finishes up to point or this every trade end with this;
Behind the end of scan of step (2.3.5) rectangular area, remove the pixel that is labeled as " 1 ".
Quantization parameter in the said quantization parameter computing module (300) comprises gray-scale statistical parameter and geometric parameter, the average density, density variance, density that said gray-scale statistical parameter comprises [district or user designated area less than, more than or equal to the pixel percentage ratio of given threshold value; Said geometric parameter comprises: lung volume, region area, girth, distance and angle.
Said classification diagnosis module (400) comprises judging unit, and said judging unit judges whether to exist emphysema according to emophysematous quantification diagnostic criteria, if be judged as emphysema, then it is classified.
Said classification diagnosis module (400); According to emophysematous quantification diagnostic criteria; The applying volume fraction method scans the lung district in each CT faultage image, with each pixel in the lung district and specified density threshold; Statistics respectively greater than, be less than or equal to the pixel of assign thresholds, calculate the percentage ratio that they account for whole lung district respectively; According to emophysematous grading diagnosis standard, confirm whether the lung district is normal, if unusual, then carry out classification.
Operation principle of the present invention and process are: at first import one group of breast CT image to be diagnosed to system; Then the lung district is cut apart automatically, extracted [essence, last; According to emophysematous quantification diagnostic criteria; Discern and diagnose, and with the key area with the special color labelling, a series of relevant density and how much statistical parameters are provided as required; Zone and relevant parameter that the prompting radiologist need pay close attention to, thus accuracy, reliability and the efficient of radiologist improved to the emphysema diagnosis.
The present invention with respect to prior art, have following advantage:
(1) extracts the lung district automatically, quickly and accurately
Existing lung is distinguished segmentation method, and like traditional thresholding method, the region growing method based on the method (Snake, Level set) of deformation model, based on the method for statistical model and method based on pattern classification, all is difficult to the lung district is cut apart rapidly and accurately.And the present invention will extract lung and divide into three steps, adopt automatic threshold, profile tracking, boundary scan and zone filling labelling method, can fast and effeciently extract the lung district.
(2) exactly emphysema are quantized classification
After extracting the [district, quantize diagnostic criteria according to emphysema, the present invention is the volume calculated mark fast, realizes emophysematous quantification classification.
(3) statistics lung district quantization parameter, the labelling lesion region
Calculate the average density of lung district or appointed area quickly and easily, density variance, maximum and minimum density etc., and quantification parameters such as area, volume, distance, angle, all right particular color labelling lesion region.
Description of drawings
Fig. 1 is the structural representation of invention based on the emphysema computer-aided diagnosis system of breast CT image;
Fig. 2 is the flow chart that the present invention is based on the emphysema computer-aided diagnosis system of breast CT image;
Fig. 3 is a certain faultage image of embodiment of the invention breast CT image (a) and rectangular histogram (b) thereof;
Fig. 4 distinguishes for embodiment of the invention lung and cuts and the profile tracking results, wherein, and (a) overall adaptive threshold segmentation result; (b) left and right sides profile tracking results; (c) [essence;
Fig. 5 is the emphysema labelling of embodiment of the invention lung district after quantitative analysis, wherein, and (a) the 4th layer; (b) the 8th layer; (c) the 12nd layer; (d) the 33rd layer;
Fig. 6 for the embodiment of the invention through quantitative analysis, the emphysema volume fraction that calculates.
The specific embodiment
Below in conjunction with embodiment and accompanying drawing the present invention is done further explain, but embodiment of the present invention is not limited thereto.
As shown in Figure 1, the emphysema computer-aided diagnosis system based on the breast CT image of the present invention comprises input module 100, extracts lung district module 200, quantization parameter computing module 300, classification diagnosis module 400 and output module 500.Wherein,
Input module (100) is used to import breast CT image to be diagnosed, and sends extraction lung district's module (200) to;
Extract lung district's module (200), be used for cutting apart automatically the [district, and send lung district information to quantization parameter computing module (300);
Quantization parameter computing module (300) is used to calculate the statistical distribution information of the picture element density of lung district or appointed area, and geological information, and sends quantization parameter to classification diagnosis module (400) and output module (500);
Classification diagnosis module (400) is used for analyzing the data that quantize parameter calculating module (300) transmission, and analysis result is sent to output module (500);
Output module (500), with the analysis result of classification diagnosis module (400), what be positioned user input waits to diagnose the breast CT image, with particular color labelling shadow of doubt, and analysis result is shown to the user.
As shown in Figure 2, the present invention handles according to following steps:
One group of breast CT faultage image to be diagnosed of step (1) input.
Step (2) is extracted the [district.Breast CT image and rectangular histogram thereof from Fig. 3 can know that lung CT image mainly comprises trunk, chest wall soft tissue, pulmonary parenchyma, trachea, bronchus, vertical diaphragm, bed board and clothes etc.Its grey level histogram has three main peaks, is divided into three main region: wherein low density area is four jiaos a black background; Middle density region is the outer background of pulmonary parenchyma and trunk; High density area is thoracic wall, vertical diaphragm, trachea, bronchus etc.Low density area, promptly black background is generally a certain constant density value minimum in the image, can with simple threshold value this part be removed at an easy rate.Middle density and high density area have bigger marker space, if can find the separation between these two zones, it as threshold value, can be carried out binaryzation to image, thereby background (the outer background of the trunk of density region in referring to) is separated with trunk.
Step (2.1) adopts overall adaptive threshold method with trunk and background separation.An at first given initial threshold (according to the lung CT image characteristics, initial threshold can be selected entire image density meansigma methods) through iterative algorithm, makes the threshold approaches optimal value then gradually, obtains threshold value accurately at last, and image is implemented to cut apart.Detailed process is following:
Use this threshold value image is divided into two types, calculate the average density of two class objects respectively, ask the average of two average densitys and as new threshold value; Again image is classified; Whether more double threshold value poor, or judge whether iterations reaches maximum determine the end process process.Its process can be described as:
Step (2.1.1) is selected initial estimate T
0, given very little stop value t, and maximum iteration time N
Max
Step (2.1.2) is used T
0As threshold value image is divided into C
1With C
2Two types;
Step (2.1.3) is to C
1With C
2In all pixels calculate average densitys
Wherein f () is an image density,
With
Be respectively C
1With C
2Pixel quantity.
Step (2.1.4) is calculated new threshold value
Step (2.1.5) is calculated the difference Δ T=|T of continuous quadratic threshold value
1-T
0|, if Δ T＜t, or iterations equals N
Max, then finish; Otherwise, with new threshold value T
1Compose and give T
0, repeating step (2.1.2)-(2.1.5).
Step (2.2) utilizes the profile tracing to extract lung district profile
Lung CT image is through the processing of step (2.1); Be divided into trunk (white expression) and background (black is represented) two parts effectively; Thereby obtain the outline of trunk at an easy rate, because pulmonary parenchyma is in intrathoracic portion, therefore need obtain profile in the trunk; Profile is pulmonary's outline in the trunk, and the present invention adopts contour tracing method to extract lung outlines.Initial point can be pressed direction from left to right at line direction from the trunk middle part, finds first white point (trunk) earlier; From then on set out then, comply with a left side to right-hand row, until first black color dots to this some place of scanning; Be left lung point, with this starting point of following the tracks of as profile.Likewise, press ecto-entad, direction can be extracted the right lung profile from right to left.
The basic thought of profile tracing is to detect the contour pixel in the target according to a certain " detection criterion " earlier, finds out other pixel of objective contour again with certain tracking criterion according to a certain characteristic of these pixels.Concrete tracing process is described below:
Step (2.1.1) finds the point of lower left, and defining the initial direction of search is the upper left side;
Step (2.1.2) then is boundary point, otherwise the direction of search is turned clockwise 45 °, till finding first stain if the upper left side is a stain;
Step (2.1.3) as new boundary point, is rotated counterclockwise 90 ° in the current search direction with this stain, continues to use with next stain of quadrat method search, up to getting back to initial point.
The scanning of step (2.3) application boundary is obtained the [district with regional completion method.
After border, lung district was confirmed, the extraction of lung district interior pixels can be converted into polygonal scan conversion and regional filling problem in the computer graphics.Polygonal scan conversion has scan-line algorithm, limit completion method, fence completion method, limit sign filling algorithm and se ed filling algorithm etc.Consider it is this special object of breast CT image, task is to extract the intra-zone pixel, and it is analyzed, and therefore, the present invention adopts a kind of context marker scanning line method based on 4 neighborhoods to obtain area pixel.The main process of this method is following:
Step (2.1.1) is asked the boundary rectangle of area-of-interest, and generation can cover the minimum rectangular area of selection area.Because computing only to this rectangular area, therefore can significantly reduce amount of calculation;
Step (2.1.2) is " 1 " with the selection area boundary marker in the rectangular area, and other then is labeled as " 0 ";
Step (2.1.3) by from top to bottom, order from left to right, scan rectangle zone is if current pixel is labeled as " 0 "; Then at current line; Begin to scan from left to right from current pixel, and put the pixel of process be labeled as " 1 ", finish up to boundary point or this every trade end;
4 neighborhoods of step (2.1.4) search current pixel find a point that is labeled as " 0 ", and as new starting point, scanning from left to right, and put through pixel and be labeled as " 1 " finishes up to boundary point or this every trade end with this;
Behind the end of scan of step (2.1.5) rectangular area, remove the pixel that is labeled as " 1 ", be the zone of being asked.
So far, accomplished the automatic leaching process in [district.Like Fig. 4 is the lung district extraction result of image among Fig. 3 (a), and wherein Fig. 4 (a) is the result after adaptive threshold is cut apart, and background is separated with trunk effectively; Fig. 4 (b) is the [profile after profile is followed the tracks of; The [essence of Fig. 4 (c) for obtaining.
Step (3) quantization parameter calculates, and calculates the lung district or specifies the picture element density and how much statistical information of arbitrary region, and wherein density parameter comprises average density, density variance, the density pixel percentage ratio greater than (being less than or equal to) given threshold value; Geometric parameter comprises volume, region area, girth, distance and angle etc.Except the relevant parameter that calculates the lung district, the present invention also provides User Defined any enclosed zone, zoning quantization parameter.
Step (4) quantitative analysis and diagnosis according to quantizing the parameter that diagnostic criteria and step (3) calculate, are carried out classification diagnosis.
Relevant emophysematous clinical quantification diagnostic criteria based on the CT image, existing both at home and abroad more research, the method for generally acknowledging at present is pulmonary function quantization parameter method (H.M.John; M.D.Austin, " Pulmonary emphysema:imaging assessment of lung volume reduction surgery " .Radiology, 212 (1): 1-3 (1999); Paeonia suffruticosa, " emophysematous iconography Quantitative Research Progress ", practical radiology magazine; 22 (5): 610-613 (2006), and Shao Guangrui, Liu Cheng; Wang Tao etc., " the spiral CT dual scanning phase is inquired in emphysema diagnosis and the value in the functional evaluation, " Chinese medicine image technology; 17 (11): 1067-1069, (2001)), it mainly utilizes lung volume, average lung density, pixel index and dynamic lung density etc. to quantize index as emophysematous diagnosis basis.What average lung density reflected is the comprehensive density of aeration status, blood flow volume, extravascular Fluid amount and lung tissue, and research shows that emophysematous average density is starkly lower than normal value, and the dark inspiratory phase of CT threshold value is-953.3HU that dark expiratory phase is-914..62HU.The lung volume mainly contains two kinds of quantization methods: the emphysema degree is confirmed according to breathing the volumetrical minimizing percentage ratio of biphase lung in (1); (2) percentage ratio (volume fraction) that accounts in the full lung according to the emphysema zone carries out analyzing and diagnosing; In the CT qualitative assessment; Often diagnose emphysema as threshold value respectively, promptly recently emphysema are carried out classification less than the percentage that the lung tissue of assign thresholds accounts in the full lung, for inspiratory phase according to density with-910HU and-950HU; With-950HU is as threshold value, and volume fraction＜5% is 0 grade of an emphysema; [5%, 10%) be 1 grade; [10%, 15%) be 2 grades; Volume fraction >=15% be 3 grades (if for expiratory phase, then threshold value be-910HU).The emphysema volume fraction can also be as the important indicator of screening before disease observation, therapeutic evaluation, the lung volume reducing operation art with POE.The present invention adopts in the pulmonary function quantization parameter method and realizes lung CT image is carried out quantitative analysis and auxiliary diagnosis through statistics emphysema volume fraction.
After obtaining the [district; The volume fraction method of application of aforementioned scans lung district in each CT faultage image, with each pixel in the lung district and specified density threshold; Statistics respectively greater than, be less than or equal to the pixel of assign thresholds, calculate the percentage ratio that they account for whole lung district respectively.According to emophysematous 5 grades of diagnostic criterias, confirm whether the lung district is normal, if unusual, then carry out classification.
Wait to diagnose the breast CT image with what the classification diagnosis result was positioned user input,, analysis result is shown to the user with particular color labelling shadow of doubt, in addition, the relevant parameter that calculates in the step display (3).
As Fig. 5 be embodiment of the invention lung district after quantitative analysis, to the key area of 4 faultage images wherein, promptly emphysema have been carried out red-label; Fig. 6 for embodiment among the present invention through quantitative analysis, the emphysema volume fraction that calculates.Table 1 is added up the relevant parameter of output for embodiment
The emphysema computer-aided diagnosis system based on the breast CT image that the present invention proposes, Fig. 3-Fig. 6 and table 1 have explained processing, the analysis result of embodiment, and the relevant parameter of statistics output.In the assistant diagnosis system described in the present invention, when the lung district is implemented to cut apart, relate to some control parameters, these parameters will comprehensively be adjusted and set to concrete data characteristics, so that systematic function reaches best.The parameter that deal with data set of the present invention sets is following:
Threshold value initial estimate T0=image averaging density in the step (2.1);
Stop value t=0.01 in the step (2.1);
Maximum iteration time N in the step (2.1)
Max=500.
The present invention is through the emphysema computer-aided diagnosis system based on the breast CT image; Automatically treating diagnosis breast CT image cuts apart; Extract the lung district rapidly and accurately, carry out analyzing and processing, and the relevant density in zone and how much statistical parameters are provided as required; Thereby zone and regional relevant parameter that the prompting radiologist need pay close attention to have improved accuracy and the efficient of radiologist to the emphysema diagnosis to a certain extent.
The foregoing description is this aspect preferred implementation; But the bright embodiment of we is not restricted to the described embodiments; Other any deviates from change, the modification done under spirit of the present invention and the principle, substitutes, combination, simplify; All should be the substitute mode of equivalence, be included within protection scope of the present invention.
Claims (7)
1. based on the emphysema computer-aided diagnosis system of breast CT image, it is characterized in that, comprising:
Input module (100) is used to import breast CT image to be diagnosed, and sends extraction lung district's module (200) to;
Extract lung district's module (200), be used for cutting apart automatically the [district, and send lung district information to quantization parameter computing module (300);
Quantization parameter computing module (300) is used to calculate the statistical distribution information of the picture element density of lung district or appointed area, and geological information, and sends quantization parameter to classification diagnosis module (400) and output module (500);
Classification diagnosis module (400) is used for analyzing the data that quantize parameter calculating module (300) transmission, and analysis result is sent to output module (500);
Output module (500), with the analysis result of classification diagnosis module (400), what be positioned user input waits to diagnose the breast CT image, with particular color labelling shadow of doubt, and analysis result is shown to the user.
2. the emphysema computer-aided diagnosis system based on the breast CT image according to claim 1 is characterized in that extracting lung district module (200) and handles according to following step:
Step (2.1) adopts overall adaptive threshold method to separate trunk and background in the breast CT image: an at first given initial threshold, and use this threshold value image is divided into two types; Ask the average of two average densitys then and, again image is classified,, make the threshold approaches optimal value gradually, obtain threshold value accurately at last, background is separated with trunk through iterative algorithm as new threshold value;
Step (2.2) adopts contour tracing method to extract lung district profile: a pixel that at first detects left lung profile according to density and locus; Then; From this point, use other pixels of contour tracing method ferret out profile, up to finding left all pixels of lung profile; Similarly, can obtain all pixels of right lung profile;
Step (2.3) adopts a kind of context marker scanning line method based on 4 neighborhoods to obtain the pixel in the [zone.
3. the emphysema computer-aided diagnosis system based on the breast CT image according to claim 2 is characterized in that initial threshold is selected entire image density meansigma methods in the step (2.1).
4. the emphysema computer-aided diagnosis system based on the breast CT image according to claim 2 is characterized in that said step (2.3) comprises the steps:
Step (2.3.1) is asked the boundary rectangle of area-of-interest, and generation can cover the minimum rectangular area of selection area;
Step (2.3.2) is labeled as " 1 " with the point of lung in the rectangular area, and other then is labeled as " 0 ";
Step (2.3.3) by from top to bottom, order from left to right, scan rectangle zone is if current pixel is labeled as " 0 "; Then at current line; Begin to scan from left to right from current pixel, and put the pixel of process be labeled as " 1 ", finish up to point or this every trade end;
4 neighborhoods of step (2.3.4) search current pixel find a point that is labeled as " 0 ", and as new starting point, scanning from left to right, and put through pixel and be labeled as " 1 " finishes up to point or this every trade end with this;
Behind the end of scan of step (2.3.5) rectangular area, remove the pixel that is labeled as " 1 ".
5. the emphysema computer-aided diagnosis system based on the breast CT image according to claim 4; It is characterized in that: the quantization parameter in the quantization parameter computing module (300) comprises gray-scale statistical parameter and geometric parameter, the average density, density variance, density that said gray-scale statistical parameter comprises [district or user designated area less than, more than or equal to the pixel percentage ratio of given threshold value; Said geometric parameter comprises: lung volume, region area, girth, distance and angle.
6. the emphysema computer-aided diagnosis system based on the breast CT image according to claim 5; It is characterized in that said classification diagnosis module (400) comprises judging unit; Said judging unit is according to emophysematous quantification diagnostic criteria; Judge whether to exist emphysema,, then it is classified if be judged as emphysema.
7. the emphysema computer-aided diagnosis system based on the breast CT image according to claim 6; It is characterized in that: classification diagnosis module (400), according to emophysematous quantification diagnostic criteria, applying volume fraction method; Scan the lung district in each CT faultage image; With each pixel in the lung district and specified density threshold, statistics respectively greater than, be less than or equal to the pixel of assign thresholds, calculate the percentage ratio that they account for whole lung district respectively; According to emophysematous grading diagnosis standard, confirm whether the lung district is normal, if unusual, then carry out classification.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2011102671651A CN102429679A (en) | 2011-09-09 | 2011-09-09 | Computer-assisted emphysema analysis system based on chest CT (Computerized Tomography) image |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2011102671651A CN102429679A (en) | 2011-09-09 | 2011-09-09 | Computer-assisted emphysema analysis system based on chest CT (Computerized Tomography) image |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN102429679A true CN102429679A (en) | 2012-05-02 |
Family
ID=45978283
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN2011102671651A Pending CN102429679A (en) | 2011-09-09 | 2011-09-09 | Computer-assisted emphysema analysis system based on chest CT (Computerized Tomography) image |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN102429679A (en) |
Cited By (35)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103892854A (en) * | 2012-12-28 | 2014-07-02 | 上海联影医疗科技有限公司 | Digital medical image processing method and device |
| CN103932727A (en) * | 2013-01-22 | 2014-07-23 | 上海理工大学 | Computer-aided diagnosis system based on CT image texture tactile sense |
| CN104573309A (en) * | 2013-10-24 | 2015-04-29 | 三星电子株式会社 | Apparatus and method for computer-aided diagnosis |
| CN105232081A (en) * | 2014-07-09 | 2016-01-13 | 无锡祥生医学影像有限责任公司 | Medical ultrasound assisted automatic diagnosis device and medical ultrasound assisted automatic diagnosis method |
| CN106384342A (en) * | 2015-07-24 | 2017-02-08 | 富士通株式会社 | Computer readable storage medium, apparatus, and method for image processing |
| JP2018151791A (en) * | 2017-03-10 | 2018-09-27 | 富士通株式会社 | Similar case image search program, similar case image search apparatus, and similar case image search method |
| CN108742663A (en) * | 2018-04-03 | 2018-11-06 | 深圳蓝韵医学影像有限公司 | Exposure dose evaluation method, device and computer readable storage medium |
| CN108882907A (en) * | 2016-04-13 | 2018-11-23 | 日本医事物理股份有限公司 | Automatic estimating method, device and the computer program in the bone region in CT image |
| CN109035283A (en) * | 2018-09-05 | 2018-12-18 | 中国医科大学附属盛京医院 | It is a kind of precisely to be detected and quantitative analysis method based on the pulmonary emphysema for randomly selecting subregion |
| CN109214397A (en) * | 2018-09-14 | 2019-01-15 | 郑州大学 | The dividing method of Lung neoplasm in a kind of lung CT image |
| CN109300122A (en) * | 2018-09-14 | 2019-02-01 | 沈阳东软医疗系统有限公司 | Image procossing and Threshold, device and equipment |
| CN109394268A (en) * | 2018-12-07 | 2019-03-01 | 余姚市华耀工具科技有限公司 | Polyp extent of injury Mapping Platform |
| CN109447969A (en) * | 2018-10-29 | 2019-03-08 | 北京青燕祥云科技有限公司 | Hepatic space occupying lesion recognition methods, device and realization device |
| CN109448822A (en) * | 2018-09-21 | 2019-03-08 | 沈阳医学院附属中心医院 | CT image pulmonary emphysema automatic marking method based on image report |
| CN109754394A (en) * | 2018-12-28 | 2019-05-14 | 上海联影智能医疗科技有限公司 | 3 d medical images processing unit and method |
| CN110176009A (en) * | 2019-05-23 | 2019-08-27 | 复旦大学 | A kind of segmentation of lung images and tracking and system |
| CN110211104A (en) * | 2019-05-23 | 2019-09-06 | 复旦大学 | A kind of image analysis method and system for pulmonary masses computer aided detection |
| CN110335281A (en) * | 2018-03-28 | 2019-10-15 | 北京连心医疗科技有限公司 | A kind of tumor boundaries determine method, equipment and storage medium |
| CN110349115A (en) * | 2019-07-04 | 2019-10-18 | 邃蓝智能科技(上海)有限公司 | CT image processing system |
| CN110533658A (en) * | 2019-09-02 | 2019-12-03 | 山东大学齐鲁医院 | Intelligent pulmonary emphysema diagnostic message processing system and method, information data processing terminal |
| CN110717915A (en) * | 2019-09-25 | 2020-01-21 | 武汉联影智融医疗科技有限公司 | Segmentation method, segmentation device, computer equipment and storage medium |
| CN111212591A (en) * | 2017-10-17 | 2020-05-29 | 富士胶片株式会社 | Medical image processing apparatus and endoscope apparatus |
| CN111242931A (en) * | 2020-01-15 | 2020-06-05 | 东北大学 | Method and device for judging small airway lesion of single lung lobe |
| CN111260627A (en) * | 2020-01-15 | 2020-06-09 | 东北大学 | Pulmonary lobe-based emphysema region judgment method and device |
| CN111598853A (en) * | 2020-04-30 | 2020-08-28 | 科大讯飞股份有限公司 | Pneumonia-oriented CT image scoring method, device and equipment |
| CN111657985A (en) * | 2020-06-30 | 2020-09-15 | 杭州依图医疗技术有限公司 | Lung image information processing method, lung image information display method and readable storage medium |
| CN111739645A (en) * | 2020-05-14 | 2020-10-02 | 上海依智医疗技术有限公司 | Training method of immune-related pneumonia prediction model |
| CN111738980A (en) * | 2020-05-14 | 2020-10-02 | 上海依智医疗技术有限公司 | Medical image display method, computer equipment and storage medium |
| WO2020232998A1 (en) * | 2019-05-22 | 2020-11-26 | 平安国际智慧城市科技股份有限公司 | Medical image evaluation method and apparatus, computer device and storage medium |
| CN112190277A (en) * | 2020-11-09 | 2021-01-08 | 华中科技大学同济医学院附属协和医院 | Data fitting method for CT reexamination of new coronary pneumonia |
| CN113096139A (en) * | 2021-04-14 | 2021-07-09 | 上海市第一人民医院 | Image segmentation processing method for lung parenchyma |
| CN113425319A (en) * | 2020-03-08 | 2021-09-24 | 广东毓秀科技有限公司 | Respiratory system disease intelligent auxiliary diagnosis decision-making method for basic medical institution |
| CN113469934A (en) * | 2020-03-31 | 2021-10-01 | 西门子医疗有限公司 | Assessment of abnormal regions associated with disease from chest CT images |
| US11341734B2 (en) | 2018-12-17 | 2022-05-24 | Shanghai United Imaging Intelligence Co., Ltd. | Systems and methods for image segmentation |
| CN116958151A (en) * | 2023-09-21 | 2023-10-27 | 中国医学科学院北京协和医院 | Method, system and equipment for distinguishing adrenal hyperplasia from fat-free adenoma based on CT image characteristics |
Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101103924A (en) * | 2007-07-13 | 2008-01-16 | 华中科技大学 | Galactophore cancer computer auxiliary diagnosis method based on galactophore X-ray radiography and system thereof |
-
2011
- 2011-09-09 CN CN2011102671651A patent/CN102429679A/en active Pending
Patent Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101103924A (en) * | 2007-07-13 | 2008-01-16 | 华中科技大学 | Galactophore cancer computer auxiliary diagnosis method based on galactophore X-ray radiography and system thereof |
Non-Patent Citations (4)
| Title |
|---|
| 薛以锋等: "基于CT图像的肺结节计算机辅助诊断系统", 《中国医学物理学杂志》 * |
| 郭圣文等: "胸部CT中肺实质的提取与辅助诊断", 《中国生物医学工程学报》 * |
| 郭圣文等: "胸部CT中肺实质的自动分割与计算机辅助诊断", 《华南理工大学学报（自然科学版）》 * |
| 陈坚等: "胸部CT图像中感兴趣区域的提取与量化分析", 《中国医学物理学杂志》 * |
Cited By (55)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103892854A (en) * | 2012-12-28 | 2014-07-02 | 上海联影医疗科技有限公司 | Digital medical image processing method and device |
| CN103932727A (en) * | 2013-01-22 | 2014-07-23 | 上海理工大学 | Computer-aided diagnosis system based on CT image texture tactile sense |
| US10147223B2 (en) | 2013-10-24 | 2018-12-04 | Samsung Electronics Co., Ltd. | Apparatus and method for computer-aided diagnosis |
| CN104573309A (en) * | 2013-10-24 | 2015-04-29 | 三星电子株式会社 | Apparatus and method for computer-aided diagnosis |
| CN104573309B (en) * | 2013-10-24 | 2019-07-19 | 三星电子株式会社 | Device and method for computer-aided diagnosis |
| CN105232081A (en) * | 2014-07-09 | 2016-01-13 | 无锡祥生医学影像有限责任公司 | Medical ultrasound assisted automatic diagnosis device and medical ultrasound assisted automatic diagnosis method |
| CN106384342A (en) * | 2015-07-24 | 2017-02-08 | 富士通株式会社 | Computer readable storage medium, apparatus, and method for image processing |
| CN106384342B (en) * | 2015-07-24 | 2019-10-25 | 富士通株式会社 | Device and method for image procossing |
| CN108882907A (en) * | 2016-04-13 | 2018-11-23 | 日本医事物理股份有限公司 | Automatic estimating method, device and the computer program in the bone region in CT image |
| JP2018151791A (en) * | 2017-03-10 | 2018-09-27 | 富士通株式会社 | Similar case image search program, similar case image search apparatus, and similar case image search method |
| CN111212591B (en) * | 2017-10-17 | 2022-08-02 | 富士胶片株式会社 | Medical image processing apparatus and endoscope apparatus |
| US11574401B2 (en) | 2017-10-17 | 2023-02-07 | Fujifilm Corporation | Medical image processing apparatus and endoscope apparatus |
| CN111212591A (en) * | 2017-10-17 | 2020-05-29 | 富士胶片株式会社 | Medical image processing apparatus and endoscope apparatus |
| CN110335281A (en) * | 2018-03-28 | 2019-10-15 | 北京连心医疗科技有限公司 | A kind of tumor boundaries determine method, equipment and storage medium |
| CN108742663A (en) * | 2018-04-03 | 2018-11-06 | 深圳蓝韵医学影像有限公司 | Exposure dose evaluation method, device and computer readable storage medium |
| CN109035283A (en) * | 2018-09-05 | 2018-12-18 | 中国医科大学附属盛京医院 | It is a kind of precisely to be detected and quantitative analysis method based on the pulmonary emphysema for randomly selecting subregion |
| CN109035283B (en) * | 2018-09-05 | 2021-03-30 | 中国医科大学附属盛京医院 | Emphysema accurate detection and quantitative analysis method based on randomly selected partitions |
| CN109300122B (en) * | 2018-09-14 | 2023-04-07 | 东软医疗系统股份有限公司 | Image processing and threshold determining method, device and equipment |
| CN109214397A (en) * | 2018-09-14 | 2019-01-15 | 郑州大学 | The dividing method of Lung neoplasm in a kind of lung CT image |
| US11478163B2 (en) | 2018-09-14 | 2022-10-25 | Neusoft Medical Systems Co., Ltd. | Image processing and emphysema threshold determination |
| CN109300122A (en) * | 2018-09-14 | 2019-02-01 | 沈阳东软医疗系统有限公司 | Image procossing and Threshold, device and equipment |
| CN109448822A (en) * | 2018-09-21 | 2019-03-08 | 沈阳医学院附属中心医院 | CT image pulmonary emphysema automatic marking method based on image report |
| CN109447969A (en) * | 2018-10-29 | 2019-03-08 | 北京青燕祥云科技有限公司 | Hepatic space occupying lesion recognition methods, device and realization device |
| CN109447969B (en) * | 2018-10-29 | 2021-08-10 | 北京青燕祥云科技有限公司 | Liver occupation lesion identification method and device and implementation device |
| CN109394268B (en) * | 2018-12-07 | 2021-05-11 | 刘志红 | Polyp harm degree mapping platform |
| CN109394268A (en) * | 2018-12-07 | 2019-03-01 | 余姚市华耀工具科技有限公司 | Polyp extent of injury Mapping Platform |
| US11341734B2 (en) | 2018-12-17 | 2022-05-24 | Shanghai United Imaging Intelligence Co., Ltd. | Systems and methods for image segmentation |
| US11836925B2 (en) | 2018-12-17 | 2023-12-05 | Shanghai United Imaging Intelligence Co., Ltd. | Systems and methods for image segmentation |
| CN109754394B (en) * | 2018-12-28 | 2021-02-23 | 上海联影智能医疗科技有限公司 | Three-dimensional medical image processing device and method |
| CN109754394A (en) * | 2018-12-28 | 2019-05-14 | 上海联影智能医疗科技有限公司 | 3 d medical images processing unit and method |
| WO2020232998A1 (en) * | 2019-05-22 | 2020-11-26 | 平安国际智慧城市科技股份有限公司 | Medical image evaluation method and apparatus, computer device and storage medium |
| CN110176009B (en) * | 2019-05-23 | 2023-03-07 | 复旦大学 | Lung image segmentation and tracking method and system |
| CN110211104B (en) * | 2019-05-23 | 2023-01-06 | 复旦大学 | Image analysis method and system for computer-aided detection of lung mass |
| CN110211104A (en) * | 2019-05-23 | 2019-09-06 | 复旦大学 | A kind of image analysis method and system for pulmonary masses computer aided detection |
| CN110176009A (en) * | 2019-05-23 | 2019-08-27 | 复旦大学 | A kind of segmentation of lung images and tracking and system |
| CN110349115A (en) * | 2019-07-04 | 2019-10-18 | 邃蓝智能科技(上海)有限公司 | CT image processing system |
| CN110533658A (en) * | 2019-09-02 | 2019-12-03 | 山东大学齐鲁医院 | Intelligent pulmonary emphysema diagnostic message processing system and method, information data processing terminal |
| CN110717915A (en) * | 2019-09-25 | 2020-01-21 | 武汉联影智融医疗科技有限公司 | Segmentation method, segmentation device, computer equipment and storage medium |
| CN111260627A (en) * | 2020-01-15 | 2020-06-09 | 东北大学 | Pulmonary lobe-based emphysema region judgment method and device |
| CN111242931A (en) * | 2020-01-15 | 2020-06-05 | 东北大学 | Method and device for judging small airway lesion of single lung lobe |
| CN111242931B (en) * | 2020-01-15 | 2023-08-22 | 东北大学 | Method and device for judging small airway lesions of single lung lobes |
| CN113425319A (en) * | 2020-03-08 | 2021-09-24 | 广东毓秀科技有限公司 | Respiratory system disease intelligent auxiliary diagnosis decision-making method for basic medical institution |
| CN113469934A (en) * | 2020-03-31 | 2021-10-01 | 西门子医疗有限公司 | Assessment of abnormal regions associated with disease from chest CT images |
| CN111598853A (en) * | 2020-04-30 | 2020-08-28 | 科大讯飞股份有限公司 | Pneumonia-oriented CT image scoring method, device and equipment |
| CN111598853B (en) * | 2020-04-30 | 2024-02-13 | 讯飞医疗科技股份有限公司 | CT image scoring method, device and equipment for pneumonia |
| CN111738980A (en) * | 2020-05-14 | 2020-10-02 | 上海依智医疗技术有限公司 | Medical image display method, computer equipment and storage medium |
| CN111739645A (en) * | 2020-05-14 | 2020-10-02 | 上海依智医疗技术有限公司 | Training method of immune-related pneumonia prediction model |
| CN111738980B (en) * | 2020-05-14 | 2023-08-04 | 北京深睿博联科技有限责任公司 | Medical image display method, computer equipment and storage medium |
| CN111739645B (en) * | 2020-05-14 | 2024-01-30 | 北京深睿博联科技有限责任公司 | Training method of immune-related pneumonia prediction model |
| CN111657985A (en) * | 2020-06-30 | 2020-09-15 | 杭州依图医疗技术有限公司 | Lung image information processing method, lung image information display method and readable storage medium |
| CN112190277A (en) * | 2020-11-09 | 2021-01-08 | 华中科技大学同济医学院附属协和医院 | Data fitting method for CT reexamination of new coronary pneumonia |
| CN113096139B (en) * | 2021-04-14 | 2022-09-06 | 上海市第一人民医院 | Image segmentation processing method for lung parenchyma |
| CN113096139A (en) * | 2021-04-14 | 2021-07-09 | 上海市第一人民医院 | Image segmentation processing method for lung parenchyma |
| CN116958151A (en) * | 2023-09-21 | 2023-10-27 | 中国医学科学院北京协和医院 | Method, system and equipment for distinguishing adrenal hyperplasia from fat-free adenoma based on CT image characteristics |
| CN116958151B (en) * | 2023-09-21 | 2023-12-12 | 中国医学科学院北京协和医院 | Method, system and equipment for distinguishing adrenal hyperplasia from fat-free adenoma based on CT image characteristics |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN102429679A (en) | Computer-assisted emphysema analysis system based on chest CT (Computerized Tomography) image | |
| Gao et al. | Classification of CT brain images based on deep learning networks | |
| CN102324109B (en) | Method for three-dimensionally segmenting insubstantial pulmonary nodule based on fuzzy membership model | |
| Aggarwal et al. | Feature extraction and LDA based classification of lung nodules in chest CT scan images | |
| Mesanovic et al. | Automatic CT image segmentation of the lungs with region growing algorithm | |
| CN100484474C (en) | Galactophore cancer computer auxiliary diagnosis system based on galactophore X-ray radiography | |
| CN112184617B (en) | Spine MRI image key point detection method based on deep learning | |
| CN105653858A (en) | Image omics based lesion tissue auxiliary prognosis system and method | |
| CN108897984A (en) | Based on correlation analysis between CT images group feature and lung cancer gene expression | |
| CN105760874A (en) | CT image processing system and method for pneumoconiosis | |
| CN104616289A (en) | Removal method and system for bone tissue in 3D CT (Three Dimensional Computed Tomography) image | |
| CN104933709A (en) | Automatic random-walk CT lung parenchyma image segmentation method based on prior information | |
| Homeyer et al. | Practical quantification of necrosis in histological whole-slide images | |
| Maitra et al. | Automated digital mammogram segmentation for detection of abnormal masses using binary homogeneity enhancement algorithm | |
| CN104933701A (en) | Mammary glandular cell segmentation method based on multi-scale growth and double-strategy adhesion-removing model | |
| CN104182755A (en) | Mammary gland molybdenum target X-ray image block feature extraction method based on tower-shaped principal component analysis (PCA) | |
| He et al. | Mammographic segmentation and risk classification using a novel binary model based bayes classifier | |
| Maitra et al. | Detection of abnormal masses using divide and conquer algorithmin digital mammogram | |
| CN102360495A (en) | Pulmonary nodule segmentation method based on average intensity projection and translation gaussian model | |
| Maitra et al. | Accurate breast contour detection algorithms in digital mammogram | |
| CN111145185A (en) | Lung parenchyma segmentation method for extracting CT image based on clustering key frame | |
| CN111528907A (en) | Ultrasonic image pneumonia auxiliary diagnosis method and system | |
| Roy et al. | Heterogeneity of human brain tumor with lesion identification, localization, and analysis from MRI | |
| CN111325754A (en) | Automatic lumbar vertebra positioning method based on CT sequence image | |
| CN104835155A (en) | Fractal-based early-stage breast cancer calcification point computer auxiliary detection method |
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 | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20120502 |