CN103679685B - Image processing system and image processing method - Google Patents
Image processing system and image processing method Download PDFInfo
- Publication number
- CN103679685B CN103679685B CN201210335088.3A CN201210335088A CN103679685B CN 103679685 B CN103679685 B CN 103679685B CN 201210335088 A CN201210335088 A CN 201210335088A CN 103679685 B CN103679685 B CN 103679685B
- Authority
- CN
- China
- Prior art keywords
- mrow
- msub
- image
- level set
- contour curve
- 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.)
- Expired - Fee Related
Links
Abstract
A kind of image processing system and image processing method are provided.A kind of image processing system for breast image includes:Image capturing device, for obtaining two-dimentional breast ultrasound ripple image;Lesion detection device, the tumor region that the two-dimentional breast ultrasound ripple Image detection for being obtained from image capturing unit includes tumor of breast;Ridge detector, for the two-dimentional breast ultrasound ripple Image detection ridge information from acquisition;Lesion segmentation device, for being partitioned into tumor of breast from the tumor region of detection based on level set, wherein, using the tumor region of detection as initial profile curve level set function is established as the two-dimentional breast ultrasound ripple image, and the ridge information iteration detected according to ridge detection unit limit the level set movements of contour curve, to minimize the energy function of the contour curve.
Description
Technical field
The application is related to a kind of image processing system and its image processing method for breast ultrasound ripple image, especially relates to
And it is a kind of from breast ultrasound ripple Image detection ridge information, tumor of breast is gone out from breast ultrasound ripple Image Segmentation based on level set,
In lesion segmentation processing, carry out limit levels collection using the ridge information of detection and develop, so as to remove fat from breast ultrasound ripple image
Region and the image processing techniques of shadow region.
Background technology
Breast cancer is the second largest killer of women, and early detection is to reduce the key of the death rate (40% or more).It is super
Sound wave is used for breast imaging as the supplemental diagnostics test of Mammogram (X ray) more and more, when breast X-ray is examined
Looking into may occur when sensitiveness reduces or when Mammogram has unacceptable radiation risk, and itself is also served as
First Line imaging technique is used.Therefore, computer-aided diagnosis (CAD) system can help to lack experience doctor avoids
Mistaken diagnosis, reduces the quantity of benign lesion biopsy on the premise of not mistaken diagnosis cancer, and reduces the change of various detections.
In computer assisted breast ultrasound ripple diagnostic system, core technology includes lesion detection and segmentation, and tumour
Segmentation is to determine that tumour is benign or pernicious key.At present, almost all of lesion segmentation approach assumes that tumor region
It is uniform, and tumor region and non-tumor region have different intensity.But existing lesion segmentation approach is difficult to gram
Take problems with.
1st, in breast ultrasound ripple image, the intensity of fat region and the intensity of tumor region are closely similar.If both
It is closer to each other, then it is difficult to differentiate between fat region and tumor region.
2nd, some tumor regions have a dark shade, and shadow region is just below tumour.Local strength's information deficiency
So that tumour and Shadow segmentation to be opened.
Therefore, it is necessary to a kind of lesion segmentation technology that more accurately can be removed fat region and shadow region.
The content of the invention
It is an object of the invention to provide a kind of image processing system and its image procossing for breast ultrasound ripple image
Method, when use level diversity method carries out lesion segmentation processing, based on the ridge information from breast ultrasound ripple Image detection to water
Flat collection evolution is limited, so as to be eliminated as much as the fat region close with tumor region and shade in lesion segmentation processing
Region.
According to an aspect of the present invention, there is provided a kind of image processing system for breast image, including:Image capturing
Device, for obtaining two-dimentional breast ultrasound ripple image;Lesion detection device, for the two-dimentional breast ultrasound obtained from image capturing unit
Ripple Image detection includes the tumor region of tumor of breast;Ridge detector, for the two-dimentional breast ultrasound ripple Image detection from acquisition
Ridge information;Lesion segmentation device, for being partitioned into tumor of breast from the tumor region of detection based on level set, wherein, with detection
Tumor region is that the two-dimentional breast ultrasound ripple image establishes level set function as initial profile curve, and detects list according to ridge
The level set movements of contour curve are limited, to minimize the energy function of the contour curve ridge information iteration of member detection.
The level set can be Chan-Vese Level Set Methods in method.
Lesion segmentation device ridge restriction factor R can be used to limit the level set of the contour curve when carrying out dividing processing
Develop:
Wherein, IR_NormIt is normalized ridge image, H is Hai Weisai (Heaviside) function, and G is Gaussian function, Φ (x,
Y) be contour curve level set function.
Following level set movements function can be used to perform level set movements for lesion segmentation device:
Wherein, δε=H ', ε → 0 are limit entries.
Lesion segmentation device can be used as stopping when iteratively limiting the level set movements of contour curve using one of following conditions
The condition of iteration:Iterations reach predetermined number, contour curve in current iteration change less than predetermined threshold value,
Made a mistake in iteration.
Image capturing device can obtain the two-dimentional breast ultrasound ripple image from supersonic imaging device.
According to another aspect of the present invention, there is provided a kind of image processing method for breast image, including:Obtain two dimension
Breast ultrasound ripple image;The two-dimentional breast ultrasound ripple Image detection obtained from image capturing unit includes the tumor area of tumor of breast
Domain;From the two-dimentional breast ultrasound ripple Image detection ridge information of acquisition;Mammary gland is partitioned into from the tumor region of detection based on level set
Tumour, wherein, establish level set as the two-dimentional breast ultrasound ripple image using the tumor region of detection as initial profile curve
Function, and the ridge information iteration detected according to ridge detection unit limit the level set movements of contour curve, with described in minimum
The energy function of contour curve.
The Level Set Method can be Chan-Vese Level Set Methods.
When carrying out dividing processing, ridge restriction factor R can be used to limit the level set movements of the contour curve:
Wherein, IR_NormIt is normalized ridge image, H is Hai Weisai (Heaviside) function, and G is Gaussian function, Φ (x,
Y) be contour curve level set function.
Following level set movements function can be used to perform level set movements:
Wherein, δε=H ', ε → 0 are limit entries.
When iteratively limiting the level set movements of contour curve the condition for stopping iteration being used as using one of following conditions:
Iterations reaches the change of predetermined number, contour curve in current iteration less than occurring in predetermined threshold value, iteration
Mistake.
The two-dimentional breast ultrasound ripple image can be obtained from supersonic imaging device.
Brief description of the drawings
By the description carried out below in conjunction with the accompanying drawings, above and other purpose of the invention and feature will become more clear
Chu, wherein:
Fig. 1 is the logic diagram for the image processing system for showing the exemplary embodiment according to the present invention;
Fig. 2 is the flow chart for the image processing method for showing the exemplary embodiment according to the present invention;
Fig. 3~Fig. 8 shows the example of the breast ultrasound image handled according to the image processing method of the present invention.
Embodiment
Hereinafter, with reference to the accompanying drawings to describing embodiments of the invention in detail.
Fig. 1 is the logic diagram for the image processing system for showing the exemplary embodiment according to the present invention.
Reference picture 1, image capturing device 110, tumour are included according to the image processing system of the exemplary embodiment of the present invention
Detector 120, ridge detector 130 and lesion segmentation device 140.
Image capturing device 110 is used to obtain two-dimentional breast ultrasound ripple image.Image capturing device 110 can surpass from connected
Acoustic imaging equipment obtains the two-dimentional breast ultrasound ripple image, can also read the two-dimentional mammary gland from information storage medium and surpass
Sound wave image.
Lesion detection device 120 is used for the two-dimentional breast ultrasound ripple Image detection bag for obtaining/reading from image capturing unit 110
Tumor region containing tumor of breast.According to the exemplary embodiment of the present invention, in detection process, the tumor region is determined
For the rectangular area comprising tumor of breast.610 in rectangle frame 310 and Fig. 6 and Fig. 8 in Fig. 3 and Fig. 5 indicate respectively basis
The tumor region of the exemplary embodiment detection of the present invention.The tumor region can also be confirmed as including its of tumor of breast
The region of his shape.
Ridge detector 130 is used for the two-dimentional breast ultrasound ripple Image detection ridge information from acquisition.Known ridge inspection can be used
Survey technology performs the detection of ridge.For example, the detection to two-dimentional breast ultrasound ripple image application different zoom ratio, to calculate difference
The response of scaling parameter simultaneously detects ridge;According to the intensity of response, the optimal pantograph ratio for a ridge can be automatically chosen
Example parameter.Then, whole ridges under different zoom ratio are combined into final ridge testing result.The lines delineated in Fig. 4
The ridge that the breast ultrasound ripple Image detection shown from Fig. 3 goes out is indicated, and what the lines instruction delineated in Fig. 7 was shown from Fig. 6
The ridge that breast ultrasound ripple Image detection goes out.
Lesion segmentation device 140 is used to be partitioned into tumor of breast from the tumor region of detection based on level set.In the segmentation
In processing, lesion segmentation device 140 is using the tumor region that lesion detection device 120 detects as initial profile curve as the two dimension breast
Level set function is established in gland ultrasonograph, and the ridge information iteration detected according to ridge detection unit 130 limit contour curve
Level set movements, to minimize the energy function of the contour curve.
Fig. 2 is the flow chart for the image processing method for showing the exemplary embodiment according to the present invention.
Reference picture 2, in step S210, image capturing device 110 obtains two-dimentional breast ultrasound ripple image.
In step S220, lesion detection device 120 includes tumor of breast from the two-dimentional breast ultrasound ripple Image detection of acquisition
Tumor region.According to the exemplary embodiment of the present invention, in detection process, lesion detection device 120 is true by the tumor region
It is set to the rectangular area for including tumor of breast.
In step S230, ridge detector 130 is from the two-dimentional breast ultrasound ripple Image detection ridge information obtained in step S210.
Known ridge detection technique can be used to perform the detection of ridge.
In step S240, lesion segmentation device 140 is partitioned into breast based on level set from the tumor region detected in step S220
Adenoncus knurl.
It will be detailed below being based on level set from step according to the lesion segmentation device 140 of the exemplary embodiment of the present invention
The tumor region of rapid S220 detections is partitioned into the processing of tumor of breast.
First, it is to establish level set function in the step S210 two-dimentional breast ultrasound ripple images obtained.
In level set, the contour curve of closure is expressed as C={ (x, y) | Φ (x, y) } by lesion segmentation device 140.It is horizontal
Set function Φ (x, y) is defined as symbolic measurement (SDF):
For the point inside contour curve, Φ (x, y) > 0;
For the point outside contour curve, Φ (x, y) < 0;
For the point on contour curve, Φ (x, y)=0.
According to the image processing method of the present invention, the tumor region that lesion segmentation device 140 detects lesion detection device 120 is made
For the initial profile curve of institute's level set function.
Chan-Vese Level Set Methods are the Level Set Methods based on region, and its target is to minimize energy function:
F(c1, c2, C) and=μ Length (C)+vArea (inside (C))
+λ1∫inside(C)|u0(x, y)-c1|2dx dy
+λ2∫outside(C)|u0(x, y)-c2|2dx dy
Wherein, Section 1 Length (C) is contour curve C length, and Section 2 Area (C) is inside contour curve C
Area, Section 3 are the uniformities of contour curve C interior zones, and Section 4 is the uniformity of contour curve C perimeters.μ0
(x, y) is image intensity, c1It is the average strength inside contour curve C, c2It is the average strength outside contour curve C, μ,
v、λ1And λ2It is constant parameter.
Secondly, limit the level of contour curve the ridge information iteration that lesion segmentation device 140 detects according to ridge detection unit
Collection develops, to minimize the energy function of the contour curve.
Level set movements function is defined as by lesion segmentation device 140:
Wherein, δε=H ', ε → 0 are limit entries, are Heaviside function H derivatives.
That is, in level set movements, the shock response of contour curve is detected, only the point close to contour curve is held
The adjustment of row level set value, so as to reduce the energy function of contour curve.
According to the exemplary embodiment of the present invention, in order to remove shadow region and fat region from the tumor region of detection,
Lesion segmentation device 140 is developed using the ridge image limit levels collection obtained by the detection of ridge detector 130.Therefore, limited using ridge
Factor R processed limits the level set movements of the contour curve C:
Wherein, IR_NormIt is normalized ridge image, H is Heaviside function, and G is Gaussian function, and Φ (x, y) is profile song
The level set function of line.
In each iteration, F (c are calculated using below equation respectively1, c2, C) in c1And c2:
Wherein, H is Hai Weisai (Heaviside) function.In tumor region, H (Φ (x, y))=1, and in tumor region
Outside, H (Φ (x, y))=0.
Therefore, lesion segmentation device 140 uses above-mentioned ridge restriction factor R iteratively limit levels collection evolution functions, every time
Tumor's profiles of the contour curve that iteration develops out closer to reality.Lesion segmentation device 140 constantly performs the iteration, until
Untill one of following condition occurs:It is bent that iterations reaches predetermined number (such as 1000 times), the profile in current iteration
The change of line be less than predetermined threshold value (for example, the change of profile point be less than 1% or less than 20 points), make a mistake in iteration
(for example, it is in expansion trend that contour curve, which changes).
Fig. 3 is the example for showing to carry out lesion segmentation according to common Chan-Vese Level Set Methods.In figure 3, rectangle
310 be the tumor region of detection, and profile 320 carries out lesion segmentation according to common Chan-Vese Level Set Methods and obtained
As a result, profile 330 is the lesion segmentation benchmark result that doctor is partitioned into according to clinical experience.General Chan-Vese level set sides
Method carries out Jaccard index=0.38 for the result (region 310) that lesion segmentation obtains.
Fig. 4 shows the ridge image detected on same breast ultrasound ripple image.
Fig. 5 is the example for showing to carry out the result of lesion segmentation according to the image processing method of the present invention.In Figure 5, examine
The tumor region 310 of survey and constant as the profile 330 of lesion segmentation benchmark.According to Fig. 4 ridge image to Chan-Vese water
Flat collection differentiation carries out limitation and performs lesion segmentation, obtains profile 520.As can be seen that the profile 330 of profile 520 and benchmark is more
It is close, its Jaccard index=0.70.
Similarly, Fig. 6-Fig. 8 is shown with the tumor's profiles that common Chan-Vese level sets carry out lesion segmentation acquisition
620 (its Jaccard index=0.49) and being developed using the present invention using ridge information limitation Chan-Vese level sets are swollen
The tumor's profiles 820 (its Jaccard index=0.86) that knurl segmentation obtains.
Thus, it will be seen that water of the present invention by the ridge information in the breast ultrasound ripple image of detection to the image
Flat collection differentiation is limited, and fat region and shadow region can be can remove in the processing of lesion segmentation, so as to obtain more
Accurate lesion segmentation result.
Although show and describing the present invention with reference to preferred embodiment, it will be understood by those skilled in the art that not
In the case of departing from the spirit and scope of the present invention that are defined by the claims, these embodiments can be carried out various modifications and
Conversion.
Claims (12)
1. a kind of image processing system for breast image, including:
Image capturing device, for obtaining two-dimentional breast ultrasound ripple image;
Lesion detection device, the tumour that the two-dimentional breast ultrasound ripple Image detection for being obtained from image capturing device includes tumor of breast
Region;
Ridge detector, for the two-dimentional breast ultrasound ripple Image detection ridge information from acquisition;
Lesion segmentation device, for being partitioned into tumor of breast from the tumor region of detection based on level set, wherein, with the tumour of detection
Region is that the two-dimentional breast ultrasound ripple image establishes level set function as initial profile curve, and is detected according to ridge detector
Ridge information iteration limit the level set movements of contour curve, to minimize the energy function of the contour curve.
2. image processing system as claimed in claim 1, wherein, the Level Set Method is Chan-Vese Level Set Methods.
3. image processing system as claimed in claim 2, wherein, lesion segmentation device is limited when carrying out dividing processing using ridge
Factor R processed limits the level set movements of the contour curve:
<mrow>
<msub>
<mrow>
<mo>&lsqb;</mo>
<mfrac>
<mrow>
<mo>&part;</mo>
<mi>&phi;</mi>
</mrow>
<mrow>
<mo>&part;</mo>
<mi>t</mi>
</mrow>
</mfrac>
<mo>&rsqb;</mo>
</mrow>
<mrow>
<mi>N</mi>
<mi>e</mi>
<mi>w</mi>
</mrow>
</msub>
<mo>=</mo>
<mfrac>
<mrow>
<mo>&part;</mo>
<mi>&phi;</mi>
</mrow>
<mrow>
<mo>&part;</mo>
<mi>t</mi>
</mrow>
</mfrac>
<mo>&times;</mo>
<mi>R</mi>
</mrow>
<mrow>
<mi>R</mi>
<mo>=</mo>
<msup>
<mrow>
<mo>(</mo>
<mi>G</mi>
<mo>&CircleTimes;</mo>
<msub>
<mi>I</mi>
<mrow>
<mi>R</mi>
<mo>_</mo>
<mi>N</mi>
<mi>o</mi>
<mi>r</mi>
<mi>m</mi>
</mrow>
</msub>
<mo>)</mo>
</mrow>
<mrow>
<mn>1</mn>
<mo>-</mo>
<mi>H</mi>
<mrow>
<mo>(</mo>
<mi>&phi;</mi>
<mo>(</mo>
<mrow>
<mi>x</mi>
<mo>,</mo>
<mi>y</mi>
</mrow>
<mo>)</mo>
<mo>)</mo>
</mrow>
</mrow>
</msup>
</mrow>
Wherein, IR_NormIt is normalized ridge image, H is Heaviside function, and G is Gaussian function, and Φ (x, y) is contour curve
Level set function.
4. image processing system as claimed in claim 3, wherein, lesion segmentation device is held using following level set movements function
Row level set movements:
<mrow>
<mfrac>
<mrow>
<mo>&part;</mo>
<mi>&phi;</mi>
</mrow>
<mrow>
<mo>&part;</mo>
<mi>t</mi>
</mrow>
</mfrac>
<mo>=</mo>
<msub>
<mi>&delta;</mi>
<mi>&epsiv;</mi>
</msub>
<mrow>
<mo>(</mo>
<mi>&phi;</mi>
<mo>)</mo>
</mrow>
<mo>&lsqb;</mo>
<mi>&mu;</mi>
<mi>d</mi>
<mi>i</mi>
<mi>v</mi>
<mrow>
<mo>(</mo>
<mfrac>
<mrow>
<mo>&dtri;</mo>
<mi>&phi;</mi>
</mrow>
<mrow>
<mo>|</mo>
<mo>&dtri;</mo>
<mi>&phi;</mi>
<mo>|</mo>
</mrow>
</mfrac>
<mo>)</mo>
</mrow>
<mo>-</mo>
<mi>v</mi>
<mo>-</mo>
<msub>
<mi>&lambda;</mi>
<mn>1</mn>
</msub>
<msup>
<mrow>
<mo>(</mo>
<msub>
<mi>u</mi>
<mn>0</mn>
</msub>
<mo>-</mo>
<msub>
<mi>c</mi>
<mn>1</mn>
</msub>
<mo>)</mo>
</mrow>
<mn>2</mn>
</msup>
<mo>+</mo>
<msub>
<mi>&lambda;</mi>
<mn>2</mn>
</msub>
<msup>
<mrow>
<mo>(</mo>
<msub>
<mi>u</mi>
<mn>0</mn>
</msub>
<mo>-</mo>
<msub>
<mi>c</mi>
<mn>2</mn>
</msub>
<mo>)</mo>
</mrow>
<mn>2</mn>
</msup>
<mo>&rsqb;</mo>
</mrow>
Wherein, δε=H ', H ' are Heaviside function H derivatives, and ε → 0 is limit entry, u0It is image intensity, c1It is in contour curve
The average strength in portion, c2It is the average strength outside contour curve, μ, ν, λ1And λ2It is constant parameter.
5. image processing system as claimed in claim 4, wherein, lesion segmentation device is iteratively limiting the level of contour curve
When collection develops the condition for stopping iteration being used as using one of following conditions:Iterations reaches predetermined number, in current iteration
The change of contour curve made a mistake less than predetermined threshold value, in iteration.
6. such as the image processing system any one of claim 1-5, wherein, image capturing device is from supersonic imaging device
Obtain the two-dimentional breast ultrasound ripple image.
7. a kind of image processing method for breast image, including:
Obtain two-dimentional breast ultrasound ripple image;
The two-dimentional breast ultrasound ripple Image detection obtained from image capturing device includes the tumor region of tumor of breast;
From the two-dimentional breast ultrasound ripple Image detection ridge information of acquisition;
Tumor of breast is partitioned into from the tumor region of detection based on level set, wherein, initial wheel is used as using the tumor region of detection
Wide curve is that the two-dimentional breast ultrasound ripple image establishes level set function, and the ridge information iteration detected according to ridge detector
The level set movements of contour curve are limited, to minimize the energy function of the contour curve.
8. image processing method as claimed in claim 7, wherein, the Level Set Method is Chan-Vese Level Set Methods.
9. image processing method as claimed in claim 8, wherein, when carrying out dividing processing, limited using ridge restriction factor R
The level set movements of the contour curve:
<mrow>
<msub>
<mrow>
<mo>&lsqb;</mo>
<mfrac>
<mrow>
<mo>&part;</mo>
<mi>&phi;</mi>
</mrow>
<mrow>
<mo>&part;</mo>
<mi>t</mi>
</mrow>
</mfrac>
<mo>&rsqb;</mo>
</mrow>
<mrow>
<mi>N</mi>
<mi>e</mi>
<mi>w</mi>
</mrow>
</msub>
<mo>=</mo>
<mfrac>
<mrow>
<mo>&part;</mo>
<mi>&phi;</mi>
</mrow>
<mrow>
<mo>&part;</mo>
<mi>t</mi>
</mrow>
</mfrac>
<mo>&times;</mo>
<mi>R</mi>
</mrow>
<mrow>
<mi>R</mi>
<mo>=</mo>
<msup>
<mrow>
<mo>(</mo>
<mi>G</mi>
<mo>&CircleTimes;</mo>
<msub>
<mi>I</mi>
<mrow>
<mi>R</mi>
<mo>_</mo>
<mi>N</mi>
<mi>o</mi>
<mi>r</mi>
<mi>m</mi>
</mrow>
</msub>
<mo>)</mo>
</mrow>
<mrow>
<mn>1</mn>
<mo>-</mo>
<mi>H</mi>
<mrow>
<mo>(</mo>
<mi>&phi;</mi>
<mo>(</mo>
<mrow>
<mi>x</mi>
<mo>,</mo>
<mi>y</mi>
</mrow>
<mo>)</mo>
<mo>)</mo>
</mrow>
</mrow>
</msup>
</mrow>
Wherein, IR_NormIt is normalized ridge image, H is Heaviside function, and G is Gaussian function, and Φ (x, y) is contour curve
Level set function.
10. image processing method as claimed in claim 9, wherein, perform level set using following level set movements function
Develop:
<mrow>
<mfrac>
<mrow>
<mo>&part;</mo>
<mi>&phi;</mi>
</mrow>
<mrow>
<mo>&part;</mo>
<mi>t</mi>
</mrow>
</mfrac>
<mo>=</mo>
<msub>
<mi>&delta;</mi>
<mi>&epsiv;</mi>
</msub>
<mrow>
<mo>(</mo>
<mi>&phi;</mi>
<mo>)</mo>
</mrow>
<mo>&lsqb;</mo>
<mi>&mu;</mi>
<mi>d</mi>
<mi>i</mi>
<mi>v</mi>
<mrow>
<mo>(</mo>
<mfrac>
<mrow>
<mo>&dtri;</mo>
<mi>&phi;</mi>
</mrow>
<mrow>
<mo>|</mo>
<mo>&dtri;</mo>
<mi>&phi;</mi>
<mo>|</mo>
</mrow>
</mfrac>
<mo>)</mo>
</mrow>
<mo>-</mo>
<mi>v</mi>
<mo>-</mo>
<msub>
<mi>&lambda;</mi>
<mn>1</mn>
</msub>
<msup>
<mrow>
<mo>(</mo>
<msub>
<mi>u</mi>
<mn>0</mn>
</msub>
<mo>-</mo>
<msub>
<mi>c</mi>
<mn>1</mn>
</msub>
<mo>)</mo>
</mrow>
<mn>2</mn>
</msup>
<mo>+</mo>
<msub>
<mi>&lambda;</mi>
<mn>2</mn>
</msub>
<msup>
<mrow>
<mo>(</mo>
<msub>
<mi>u</mi>
<mn>0</mn>
</msub>
<mo>-</mo>
<msub>
<mi>c</mi>
<mn>2</mn>
</msub>
<mo>)</mo>
</mrow>
<mn>2</mn>
</msup>
<mo>&rsqb;</mo>
</mrow>
Wherein, δε=H ', H ' are Heaviside function H derivatives, and ε → 0 is limit entry, u0It is image intensity, c1It is in contour curve
The average strength in portion, c2It is the average strength outside contour curve, μ, ν, λ1And λ2It is constant parameter.
11. image processing method as claimed in claim 10, wherein, when iteratively limiting the level set movements of contour curve
The condition for stopping iteration being used as using one of following conditions:It is bent that iterations reaches predetermined number, the profile in current iteration
The change of line, which is less than in predetermined threshold value, iteration, to make a mistake.
12. such as the image processing method any one of claim 7-11, wherein, obtain from supersonic imaging device described in
Two-dimentional breast ultrasound ripple image.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201210335088.3A CN103679685B (en) | 2012-09-11 | 2012-09-11 | Image processing system and image processing method |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201210335088.3A CN103679685B (en) | 2012-09-11 | 2012-09-11 | Image processing system and image processing method |
Publications (2)
Publication Number | Publication Date |
---|---|
CN103679685A CN103679685A (en) | 2014-03-26 |
CN103679685B true CN103679685B (en) | 2018-03-27 |
Family
ID=50317140
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201210335088.3A Expired - Fee Related CN103679685B (en) | 2012-09-11 | 2012-09-11 | Image processing system and image processing method |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN103679685B (en) |
Families Citing this family (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN104517116A (en) * | 2013-09-30 | 2015-04-15 | 北京三星通信技术研究有限公司 | Device and method for confirming object region in image |
CN103942799B (en) * | 2014-04-25 | 2017-02-01 | 哈尔滨医科大学 | Breast ultrasounography image segmentation method and system |
CN105374025B (en) * | 2015-09-30 | 2018-05-04 | 上海联影医疗科技有限公司 | Breast image acquisition methods and device, mammography system |
US10297024B2 (en) | 2015-09-30 | 2019-05-21 | Shanghai United Imaging Healthcare Co., Ltd. | System and method for determining a breast region in a medical image |
CN108154509B (en) * | 2018-01-12 | 2022-11-11 | 平安科技(深圳)有限公司 | Cancer identification method, device and storage medium |
CN110033456B (en) * | 2019-03-07 | 2021-07-09 | 腾讯科技(深圳)有限公司 | Medical image processing method, device, equipment and system |
Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN1870006A (en) * | 2005-04-19 | 2006-11-29 | 西门子共同研究公司 | Effective nuclear density assess for horizontal collection divided shapes and brightness prior |
CN101334895A (en) * | 2008-08-07 | 2008-12-31 | 清华大学 | Image division method aiming at dynamically intensified mammary gland magnetic resonance image sequence |
CN101366661A (en) * | 2007-08-06 | 2009-02-18 | 卡尔斯特里姆保健公司 | Linear structure verification in medical applications |
CN101599174A (en) * | 2009-08-13 | 2009-12-09 | 哈尔滨工业大学 | Method for outline extraction of level set medical ultrasonic image area based on edge and statistical nature |
CN102354396A (en) * | 2011-09-23 | 2012-02-15 | 清华大学深圳研究生院 | Method for segmenting image with non-uniform gray scale based on level set function |
CN102542556A (en) * | 2010-12-30 | 2012-07-04 | 汕头大学 | Method for automatically extracting ultrasonic breast tumor image |
Family Cites Families (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US7773806B2 (en) * | 2005-04-19 | 2010-08-10 | Siemens Medical Solutions Usa, Inc. | Efficient kernel density estimation of shape and intensity priors for level set segmentation |
-
2012
- 2012-09-11 CN CN201210335088.3A patent/CN103679685B/en not_active Expired - Fee Related
Patent Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN1870006A (en) * | 2005-04-19 | 2006-11-29 | 西门子共同研究公司 | Effective nuclear density assess for horizontal collection divided shapes and brightness prior |
CN101366661A (en) * | 2007-08-06 | 2009-02-18 | 卡尔斯特里姆保健公司 | Linear structure verification in medical applications |
CN101334895A (en) * | 2008-08-07 | 2008-12-31 | 清华大学 | Image division method aiming at dynamically intensified mammary gland magnetic resonance image sequence |
CN101599174A (en) * | 2009-08-13 | 2009-12-09 | 哈尔滨工业大学 | Method for outline extraction of level set medical ultrasonic image area based on edge and statistical nature |
CN102542556A (en) * | 2010-12-30 | 2012-07-04 | 汕头大学 | Method for automatically extracting ultrasonic breast tumor image |
CN102354396A (en) * | 2011-09-23 | 2012-02-15 | 清华大学深圳研究生院 | Method for segmenting image with non-uniform gray scale based on level set function |
Non-Patent Citations (2)
Title |
---|
C-V模型在乳腺X影像分割中的研究和应用;熊思;《湖北第二师范学院学报》;20110220;第28卷(第2期);全文 * |
超声乳腺肿瘤的全自动SVM检测与水平集分割算法;徐静等;《计算机辅助设计与图形学学报》;20120515;第24卷(第5期);第663-665页 * |
Also Published As
Publication number | Publication date |
---|---|
CN103679685A (en) | 2014-03-26 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
Midya et al. | Influence of CT acquisition and reconstruction parameters on radiomic feature reproducibility | |
Varghese et al. | Texture analysis of imaging: what radiologists need to know | |
CN103679685B (en) | Image processing system and image processing method | |
Ganeshan et al. | Quantifying tumour heterogeneity with CT | |
Goo | A computer-aided diagnosis for evaluating lung nodules on chest CT: the current status and perspective | |
Mesanovic et al. | Automatic CT image segmentation of the lungs with region growing algorithm | |
US8781202B2 (en) | Tumor classification based on an analysis of a related ultrasonic attenuation map | |
Firjani et al. | A diffusion-weighted imaging based diagnostic system for early detection of prostate cancer | |
US7929741B2 (en) | System and method for automated detection of mucus plugs within bronchial tree in MSCT images | |
Wang et al. | Mammographic mass segmentation: embedding multiple features in vector-valued level set in ambiguous regions | |
US9014447B2 (en) | System and method for detection of lesions in three-dimensional digital medical image | |
Xu et al. | Quantifying the margin sharpness of lesions on radiological images for content‐based image retrieval | |
US8090172B2 (en) | Robust segmentation of breast and muscle in MRI | |
CN115843373A (en) | Multi-scale local level set ultrasonic image segmentation method fusing Gabor wavelets | |
Zhou et al. | Grading of hepatocellular carcinoma based on diffusion weighted images with multiple b‐values using convolutional neural networks | |
US20150190120A1 (en) | Method and system for processing ultrasonic imaging data | |
Kaur et al. | Computer-aided diagnosis of renal lesions in CT images: a comprehensive survey and future prospects | |
US8577101B2 (en) | Change assessment method | |
US20090069665A1 (en) | Automatic Lesion Correlation in Multiple MR Modalities | |
CN101989351A (en) | Suspected lung nodule image enhancement directional scale filtering method | |
Tan et al. | Chest wall segmentation in automated 3D breast ultrasound scans | |
Liu et al. | A new background distribution‐based active contour model for three‐dimensional lesion segmentation in breast DCE‐MRI | |
KR101307162B1 (en) | Computer Aided System and method for Breast Mass Detection and Analysis in Microwave Tomography | |
Ma et al. | Radiomics analysis of breast lesions in combination with coronal plane of ABVS and strain elastography | |
US20090069666A1 (en) | Correction of Intensity Inhomogeneity in Breast MRI |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
C10 | Entry into substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant | ||
CF01 | Termination of patent right due to non-payment of annual fee | ||
CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20180327 |