CN103093457B - Ultrasonic blood vessel border detection System and method for - Google Patents
Ultrasonic blood vessel border detection System and method for Download PDFInfo
- Publication number
- CN103093457B CN103093457B CN201210584625.8A CN201210584625A CN103093457B CN 103093457 B CN103093457 B CN 103093457B CN 201210584625 A CN201210584625 A CN 201210584625A CN 103093457 B CN103093457 B CN 103093457B
- Authority
- CN
- China
- Prior art keywords
- blood vessel
- image
- border
- pixel
- module
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Landscapes
- Ultra Sonic Daignosis Equipment (AREA)
Abstract
A kind of ultrasonic blood vessel border detection system, including: image zooming-out module, for extracting blood vessel characteristic image from ultrasonoscopy；Middle epicardial border detection subsystem, for by the border being detected tunica media, adventitia in described blood vessel characteristic image by region growing；Tube chamber border detection subsystem, for detecting the border of vessel lumen further from the middle epicardial border interior zone that described middle epicardial border detection subsystem is detected by cluster.The present invention also provides for a kind of corresponding ultrasonic blood vessel boundary detection method.Image in ultrasonic blood vessel is made the image of epicardial border both sides in blood-vessel image be prone to differentiate and make tube chamber boundaries on either side image in blood-vessel image be prone to differentiate by cluster mode by the way of region growing by the ultrasonic blood vessel border detection System and method for of the present invention, thus the middle epicardial border detected and tube chamber boundary coordinate are accurate.
Description
Technical field
The present invention relates to a kind of medical image processing system and method, in particular it relates to a kind of base
In the vessel lumen border of ultrasonoscopy and middle epicardial border detecting system and method.
Background technology
Cardiovascular and cerebrovascular disease has become the number one killer of human health.Atherosclerosis and complication thereof cause
The common cause of cardiovascular and cerebrovascular disease.The size and shape of vessel lumen border and middle epicardial border is that clinic is examined
The important indicator of atherosclerosis is weighed in disconnected.
Know vessel lumen border and the positional information of China and foreign countries' membrane junction for convenience, enter usually through ultrasonic examination
The shooting of row blood-vessel image.Existing mode passes through handmarking's vessel lumen border and middle epicardial border, by
Manually in captured ivus image (Intravascular Ultrasound, IVUS), touch off blood vessel
Tube chamber border and the position of middle epicardial border.Ivus image owing to shooting for same patient is frequent
Comprise thousands of width image, not only time-consuming by the method for artificial manual markings, and poor repeatability, it is easily subject to
Experience and the impact of subjective factors.
Additionally, existing use image procossing detection vessel lumen border and the method for middle epicardial border, exist
Ultrasonoscopy resolution is low, be attended by the shortcoming such as serious speckle noise and calcification artifact so that detection blood vessel
There is bigger difficulty in tube chamber border and middle epicardial border.
Summary of the invention
In view of this, it is necessary to a kind of ultrasonic blood vessel border detection system with higher resolution is provided.
Additionally, there is a need to provide a kind of ultrasonic blood vessel boundary detection method accordingly with higher resolution.
A kind of ultrasonic blood vessel border detection system, including:
Image zooming-out module, for extracting blood vessel characteristic image from ultrasonoscopy；
Middle epicardial border detection subsystem, for being detected by region growing by described blood vessel characteristic image
Go out the border of tunica media, adventitia；
Tube chamber border detection subsystem, for the China and foreign countries detected from described middle epicardial border detection subsystem
Further by the border of cluster detection vessel lumen in membrane boundary interior zone.
Alternatively, described ultrasonic blood vessel border detection system farther includes polar coordinate transform module, and being used for will
The blood vessel characteristic image extracted carries out polar coordinate transform, to obtain the polar coordinate change of described blood vessel characteristic image
Changing image, described middle epicardial border detection subsystem, described tube chamber border detection subsystem are respectively according to described
Polar coordinate transform image detects the middle epicardial border of described blood vessel and the border of described vessel lumen.
Alternatively, during described polar coordinate transform module takes the rectangular coordinate of this blood vessel characteristic image, coordinate isPixel as co-ordinate zero point, this blood vessel characteristic image is carried out polar coordinate transform, its
In: h be the height of ultrasonoscopy, w be this ultrasonoscopy width,Represent round numbers downward to this value.
Alternatively, described ultrasonic blood vessel border detection system farther includes rectangular coordinates transformation module, is used for
Epicardial border in blood vessel under the polar coordinate system of detection and vessel lumen border are transformed to rectangular coordinate.
Alternatively, described ultrasonic blood vessel border detection system farther includes filtering and noise reduction module, for institute
State blood vessel characteristic image to be filtered and denoising.
Alternatively, described filtering and noise reduction module carries out medium filtering to remove noise to described blood vessel characteristic image.
Alternatively, described filtering and noise reduction module carries out the square that template window is predefined size of described medium filtering
Shape window.
Alternatively, described filtering and noise reduction module carries out the most described medium filtering to described blood vessel characteristic image.
Alternatively, described filtering and noise reduction module carries out the most described medium filtering to described blood vessel characteristic image
Template window is the rectangular window template of 3 × 3,4 × 4,5 × 5,6 × 6.
Alternatively, described middle epicardial border detection subsystem includes:
Region growing module, is used for setting up seed point set, and based on initial pixel point by described blood vessel feature
The pixel conformed to a predetermined condition in image adds described seed point set to；
Incremental modular, concentrates pixel number according to the increment of predetermined condition for calculating described seed points；
Mark module, for the pixel region corresponding to the described increment that labelling is maximum.
Alternatively, described initial pixel point is that on described blood vessel characteristic image, coordinate is the pixel of (1,1).
Alternatively, described predetermined condition is: be the pixel of neighborhood 8 connection for seed point set.
Alternatively, described predetermined condition is: this pixel and the initial pixel point that coordinate in filtering image is (1,1)
Between gray scale difference less than or equal to predetermined threshold value T.
Alternatively, the multiple values in described predetermined threshold value T is given range.
Alternatively, described predetermined threshold value T is the continuous integral number between 1 to 255.
Alternatively, described region growing module is set up according to multiple predetermined threshold values T and is obtained corresponding multiple district
Territory S and corresponding pixel number N.
Alternatively, described region growing module sets up the multiple seed point sets of formation, institute according to multiple predetermined conditions
State incremental modular and calculate the pixel number dullness according to the plurality of predetermined condition of the plurality of seed points concentration
The increment increased and produce.
Alternatively, in the plurality of seed point set, each pixel forms corresponding image-region, described labelling mould
The region that block is formed for the pixel corresponding to increment that labelling is maximum.
Alternatively, described ultrasonic blood vessel border detection system farther includes rectification module, for described mark
The region of note module institute labelling is corrected.
Alternatively, described rectification module uses Mathematical Morphology Method to correct the region of institute's labelling.
Alternatively, described rectification module carries out closing operation of mathematical morphology, described form to the region of described institute labelling
Learning closed operation uses radius to be R1Round die block computing, radius R1Calculating formula be:Its
Middle NjThe number of the pixel included by the region of described institute labelling.
Alternatively, the hole in described region after described rectification module deletes described closing operation of mathematical morphology.
Alternatively, described rectification module carries out closing operation of mathematical morphology, described form to the region of described institute labelling
Learning closed operation uses radius to be R2Round die block computing, radius R2Calculating formula be:Its
Middle NjThe number of the pixel included by the region of described institute labelling.
Alternatively, described tube chamber border detection subsystem includes:
Internal selection module, adventitia in the blood vessel detected according to described middle epicardial border detection subsystem
Border is by the image in the region selected in described blood vessel characteristic image in described blood vessel within epicardial border；
Cluster module, for obtaining the specific characteristic vector of each pixel in described images of interior regions, and root
Carry out clustering computing according to described specific characteristic vector.
Alternatively, the pixel of described tube chamber boundaries on either side area image is clustered by described cluster module respectively,
So that the some set of the cluster of described cluster is formed cluster areas respectively.
Alternatively, described specific characteristic vector includes the gray value of each pixel and each picture in described interior zone
The standard deviation of gray value in vegetarian refreshments 8 territory.
Alternatively, the appointment spy that described cluster computing is the pixel of (1,1) with coordinate on described blood vessel characteristic image
Levying the take up an official post characteristic vector of a pixel of epicardial border in vectorial and described blood vessel is cluster centre.
Alternatively, described cluster module utilizes K means clustering method to pixel each in described interior zone
Specific characteristic vector calculates and clusters, and each pixel cluster forms cluster areas respectively.
Alternatively, described ultrasonic blood vessel border detection system farther includes rectification module, for described poly-
Described in the representative that the generic module cluster each pixel of gained is formed, the cluster areas of vessel lumen interior zone is carried out
Correct.
Alternatively, described rectification module uses Mathematical Morphology Method to correct described cluster areas.
Alternatively, described rectification module carries out closing operation of mathematical morphology to described cluster areas, and described morphology is closed
Computing uses radius to be R3Round die block computing, described radius R3Calculating formula be: R3=0.02 ×
Polarwidth, wherein polarwidth is described blood vessel characteristic image pole of gained after polar coordinate transform converts
The horizontal size of coordinate diagram picture.
Alternatively, the hole of described cluster areas after described rectification module deletes closing operation of mathematical morphology.
Alternatively, described rectification module carries out closing operation of mathematical morphology to described cluster areas, and described morphology is closed
Computing uses radius to be R4Round die block computing, radius R4Calculating formula be: R4=0.03 × polarwidth,
Wherein polarwidth is through described blood vessel characteristic image polar coordinate image of gained after polar coordinate transform converts
Horizontal size.
Alternatively, described ultrasonic blood vessel border detection system farther includes Boundary Extraction module, is used for extracting
Epicardial border and the border of described vessel lumen in the blood vessel of described blood vessel characteristic image.
A kind of ultrasonic blood vessel boundary detection method, is carried out including such as above-mentioned ultrasonic blood vessel boundary detection method
Step.
Image in ultrasonic blood vessel is passed through region growing by the ultrasonic blood vessel border detection System and method for of the present invention
Mode makes the image of epicardial border both sides in blood-vessel image be prone to differentiate and make vessel graph by cluster mode
In Xiang, tube chamber boundaries on either side image is prone to differentiate, thus the middle epicardial border detected and tube chamber boundary coordinate are accurate
Really.
Accompanying drawing explanation
Fig. 1 is the function structure chart of a kind of embodiment of ultrasonic blood vessel border detection system of the present invention；
Fig. 2 be example ultrasonic blood vessel in image；
Fig. 3 be Fig. 2 of example ultrasonic blood vessel in image through this Fig. 1 image zooming-out module 102 extract after
The blood vessel characteristic image region of gained；
Fig. 4 is that the blood vessel characteristic image of Fig. 3 of example is after the polar coordinate transform module 104 of Fig. 1 converts
The polar coordinate image of gained；
Fig. 5 is that the polar coordinate image of Fig. 4 of example make use of 3 successively through the filtering and noise reduction module 106 of Fig. 1
The rectangular window module of × 3,4 × 4,5 × 5,6 × 6 carries out the filtering image of gained after 4 medium filterings；
Fig. 6 be example with threshold value T of monotone increasing as abscissa, accordingly by the incremental modular 124 of Fig. 1
The relation curve of increment A is vertical coordinate the one group of threshold value-increment calculated.
Fig. 7 is that the filter profile shown in Fig. 5 of example carries out seed through the region growing module 122 of Fig. 1
After the growth of point set, incremental modular 124 carry out incremental computations, the district marked by mark module 126
Territory；
Fig. 8 is the marked region of Fig. 7 gained of example institute after the rectification module 108 of Fig. 1 is corrected
The regional graphics obtained；
Fig. 9 is the middle epicardial border detected by the Boundary Extraction module 110 of Fig. 1 of example；
Figure 10 is the Vascular Ultrasonography figure that the cluster module 144 through Fig. 1 of example has carried out the 2 of cluster calculation
The schematic diagram of the cluster areas as being formed；
The cluster areas of Figure 10 that Figure 11 is example is formed after the rectification module 108 of Fig. 1 is corrected
Image；
Figure 12 is the vessel lumen border detected by the Boundary Extraction module 110 of Fig. 1 of example；
Figure 13 is the blood detected in epicardial border 902 and Figure 12 in the blood vessel that detected in Fig. 9 of example
Pipe tube chamber border 1202 identifies the blood vessel at rectangular coordinate after rectangular coordinates transformation block transforms is rectangular coordinate
Schematic diagram on image；
Figure 14 is the flow chart of a kind of embodiment of ultrasonic blood vessel boundary detection method of the present invention.
Detailed description of the invention
Fig. 1 show the function structure chart of a kind of embodiment of ultrasonic blood vessel border detection system of the present invention.
According to this embodiment, ultrasonic blood vessel border detection system 10 includes that image zooming-out module 102, polar coordinate become
Die change block 104, filtering and noise reduction module 106, middle epicardial border detection subsystem 12 and tube chamber border detection
System 14, rectification module 108, Boundary Extraction module 110, rectangular coordinates transformation module 112.
Image zooming-out module 102 is for extracting blood vessel characteristic image region from ultrasonoscopy.This blood vessel feature
Image-region typically has a characteristics of image bigger with background image difference, thus image zooming-out module 102
Extracting method can include the methods such as rectangular histogram, Fourier transform, least square, is not limited thereto.Show
Example ground, Fig. 2 show image in ultrasonic blood vessel, and the image 302 shown in Fig. 3 is figure in Fig. 2 ultrasonic blood vessel
As gained blood vessel characteristic image region after present embodiment image zooming-out module 102 is extracted.
Polar coordinate transform module 104 is for carrying out polar coordinate transform by the blood vessel characteristic image extracted, to obtain
Polar coordinate transform image to this blood vessel characteristic image.Specifically, ultrasonoscopy the former blood vessel extracted
Characteristic image is rectangular coordinate image.This polar coordinate transform module 104 takes in the rectangular coordinate of this ultrasonoscopy
Coordinate isPixel as co-ordinate zero point, this blood vessel characteristic image is carried out polar coordinate change
Change, wherein: h be the height of ultrasonoscopy, w be this ultrasonoscopy width,Represent downward to this value
Round numbers.Illustratively, Fig. 4 show the blood vessel characteristic image of Fig. 3 gained pole seat after polar coordinate transform
Logo image.
Filtering and noise reduction module 106 for the blood vessel feature polar coordinate image after transformed is filtered and denoising,
To remove the noise in former ultrasonoscopy.Illustratively, in alternative embodiments, filtering and noise reduction module 106
Described blood vessel feature polar coordinate image is carried out medium filtering to remove noise.Illustratively, the mould of medium filtering
Plate window is the rectangular window of predefined size, in addition can to described transformed after blood vessel feature polar diagram
As carrying out repeatedly Filtering Processing, repeatedly the rectangular window size used in filtering also can differ.Such as:
Can successively with the rectangular window template that size is 3 × 3,4 × 4,5 × 5,6 × 6 successively to described transformed after
Blood vessel feature polar diagram carries out medium filtering.As it is shown in figure 5, it is that filtering and noise reduction module 106 is to Fig. 4
Blood vessel feature polar diagram make use of the rectangular window module of 3 × 3,4 × 4,5 × 5,6 × 6 to carry out successively
The filtering image of gained after 4 medium filterings.
Described middle epicardial border detection subsystem 12 is for by detecting tunica media, outer in described filtering image
The border of film；Described tube chamber border detection subsystem 14 is for being detected subsystem 12 by described middle epicardial border
Detected middle film interior zone detects tube chamber border further.
Middle epicardial border detection subsystem 12 includes region growing module 122, incremental modular 124, labelling mould
Block 126.
Region growing module 122 is used for setting up seed point set, and based on initial pixel point by described filter profile
In the pixel that conforms to a predetermined condition add described seed point set to.Specifically, in alternative embodiments,
Described initial pixel point is that on described filtering image, coordinate is the pixel of (1,1).Described predetermined condition is:
1, for seed point set be neighborhood 8 connect pixel；
2, in this pixel and filtering image coordinate be gray scale difference between the initial pixel point of (1,1) less than or etc.
In predetermined threshold value T.
Thus, region growing module 122 can obtain corresponding seed point set, this kind based on predetermined threshold value T
The region S that pixel corresponding to sub-point set is formed includes N number of pixel.
Further, described predetermined threshold value T is the multiple values in given range, region growing module 122
Set up according to the plurality of predetermined threshold value T and obtain corresponding multiple region S and corresponding pixel number N.
Illustratively, in alternative embodiments, threshold value T is the continuous integral number between 1 to 255, is designated as respectively
T1=1, T2=2 ... T255=255, obtain region S the most accordingly1、S2、……S255And pixel number N1、
N2、……N255。
Incremental modular 124 for the seed points of zoning pop-in upgrades 122 gained concentrate pixel number according to
The increment of predetermined condition.Specifically, incremental modular 124 calculates according to pixel N during threshold value T monotone increasing
Increment A, incremental modular 124 calculates increment A such as following formula:
A1=N2-N1、A2=N3-N2、……A254=N255-N254；
Thus obtain the corresponding set of increment A corresponding to threshold value T.Illustratively, Fig. 6 show with monotone increasing
Threshold value T added is abscissa, increment A is vertical coordinate a group calculated by incremental modular 124 accordingly
The relation curve of threshold value-increment.
Mark module 126 is for the pixel region corresponding to labelling maximal increment.Specifically, mark module
126 for selecting the increment A of maximum from one group of increment A that incremental modular 124 is calculatedj, and labelling should
Maximal increment AjCorresponding seed points concentrates the region S that each pixel is formedj.Illustratively, institute in Fig. 7
The region 702 shown carries out the life of seed point set for the filter profile shown in Fig. 5 through region growing module 122
After long, incremental modular 124 carries out incremental computations, the region marked by mark module 126.
Rectification module 108 is for correcting the region of 126 labellings of mark module.Specifically, correct
Module 108 uses the Mathematical Morphology Method region S to 126 labellings of mark modulejCorrect.Example
Ground, rectification module 108 is to marked region SjCarry out a series of morphology operations, including:
1, to region SjCarrying out morphology and close (Morphology Closing) computing, fortune is closed in described morphology
Calculating and using radius is R1Round die block enter computing, radius R1Calculating formula be:Wherein Nj
For region SjThe number of included pixel.
2, closing operation of mathematical morphology rear region S is deletedjHole；Illustratively, region 702 during described hole is Fig. 7
White portion in included black part；
3, to region SjCarrying out morphology and close (Morphology Closing) computing, fortune is closed in described morphology
Calculating and using radius is R2Round die block enter computing, radius R2Calculating formula be:Wherein
NjFor region SjThe number of included pixel.
Through the image calculating gained of rectification module 108, can clearly reflect blood-vessel image medium vessels China and foreign countries
The image difference of membrane boundary both sides.Illustratively, Fig. 8 show the marked region of Fig. 7 gained through overcorrection mould
Block 108 correct after the regional graphics of gained, in this figure, outer membrane edge in the blood vessel of region 802 reflection
Boundary's interior zone, it has compared with the middle epicardial border perimeter that region 804 is reflected significantly distinguishes,
It is prone to differentiate.
The Boundary Extraction module 110 image after correcting according to rectification module 108 detects described blood-vessel image
Blood vessel in epicardial border.As it is shown in figure 9, after the rectification of rectification module 108, described blood-vessel image
In middle epicardial border both sides image there is significantly difference, it is easy to detect middle epicardial border 902.
Described tube chamber border detection subsystem 14 is for being detected from described middle epicardial border detection subsystem
Detecting the border of vessel lumen in epicardial border interior zone further, this tube chamber border detection subsystem 14 wraps
Include internal selection module 142, cluster module 144.
Internal select module 142 in the blood vessel detected according to Boundary Extraction module 110 epicardial border by
Filtering image after described filtering and noise reduction module 106 has carried out filtering and noise reduction selects described blood vessel China and foreign countries
The image in the region within membrane boundary.Specifically, as a example by the middle epicardial border 902 shown in Fig. 9, with limit
Boundary 902 is that boundary selects the region above border 902 to be epicardial border inner area in blood vessel in filtering image
The image in territory.
Cluster module 144 is for obtaining each pixel in the internal interior zone selected selected by module 142
Specific characteristic vector, and carry out clustering computing, with by described tube chamber border two according to described specific characteristic vector
The pixel of side region image clusters respectively, so that the some set of the cluster of described cluster is formed cluster areas respectively.
Specifically, cluster module 144 in obtaining described interior zone the gray value of each pixel lead with each pixel 8
The standard deviation of gray value in the range of territory, using the specific characteristic vector as this pixel.Cluster module 144 points
The specific characteristic vector C that coordinate on described filtering image is the pixel of (1,1) is not set1And foregoing boundary 902
Characteristic vector C of upper arbitrary pixel2For cluster centre, utilize K means clustering method to each in interior zone
The characteristic vector of pixel calculates and clusters, thus obtains on filtering image coordinate for the pixel of (1,1)
Pixel for cluster centre clusters GxGather with arbitrary pixel pixel as cluster centre on border 902
Class Gy.Pixel cluster GxPixel cluster GyRepresentative pixel forms cluster areas S respectivelyxWith poly-
Class region Sy.Illustratively, Figure 10 show the blood of Fig. 1 having carried out cluster calculation through cluster module 144
The schematic diagram of the cluster areas that pipe ultrasonoscopy is formed, in figure, region 1002 represents pixel cluster GxInstitute
Cluster areas S that the pixel represented is formedx, region 1004 represents pixel cluster GyRepresentative pixel
Cluster areas S that point is formedy。
Rectification module 108 is additionally operable to the cluster areas of formation calculated to cluster module and corrects.Specifically,
Rectification module 108 uses Mathematical Morphology Method to cluster areas SxCorrect.Illustratively, optionally
In embodiment, rectification module 108 carries out following rectification computing to cluster areas:
1, cluster areas carrying out morphology and close (Morphology Closing) computing, fortune is closed in described morphology
Calculating and using radius is R3Round die block computing, radius R3Calculating formula be: R3=0.02 × polarwidth,
Wherein polarwidth is the horizon rule of the polar coordinate image of gained after described polar coordinate transform module 104 converts
Very little；
2, closing operation of mathematical morphology rear region S is deletedxHole；
3, to region SxCarrying out morphology and close (Morphology Closing) computing, fortune is closed in described morphology
Calculating and using radius is R4Round die block computing, radius R4Calculating formula be: R4=0.03 × polarwidth,
Wherein polarwidth is the horizon rule of the polar coordinate image of gained after described polar coordinate transform module 104 converts
Very little.
Illustratively, Figure 11 is the institute after the rectification computing of rectification module 108 of the cluster areas shown in Figure 10
The image formed, by Figure 11 it will be seen that after overcorrection computing, the vessel lumen limit of described blood-vessel image
Both sides, boundary image has significantly difference, it is easy to differentiate.
Boundary Extraction module 110 is additionally operable to the cluster areas image after correcting according to rectification module 108 and detects institute
State the tube chamber border of blood-vessel image.As shown in figure 12, after the rectification of rectification module 108, described blood vessel
Middle epicardial border both sides image in image has significantly difference, it is easy to detect middle epicardial border 1202.
Rectangular coordinates transformation module 112 is for the polar coordinate system that will detect through described Boundary Extraction module 110
Under blood vessel in epicardial border and vessel lumen border be transformed to rectangular coordinate, and then can be at rectangular coordinate
Middle epicardial border and the tube chamber border of blood vessel is demonstrated under blood-vessel image.Illustratively, shown in Figure 13 for Fig. 9
The vessel lumen border 1202 detected in epicardial border 902 and Figure 12 in middle detected blood vessel is through right angle
Coordinate transformation module identifies the schematic diagram on the blood-vessel image of rectangular coordinate after being transformed to rectangular coordinate.
Image in ultrasonic blood vessel is made by the way of region growing by the ultrasonic blood vessel border detection system of the present invention
Obtain the image of epicardial border both sides in blood-vessel image to be prone to differentiate and make blood-vessel image is managed by cluster mode
Chamber boundaries on either side image is prone to differentiate, thus the middle epicardial border detected and tube chamber boundary coordinate are accurate.
Figure 14 show a kind of embodiment of the ultrasonic blood vessel boundary detection method of the present invention, below in conjunction with figure
This embodiment is illustrated by the ultrasonic blood vessel border detection system shown in 1.
Step 1402, extracts blood vessel characteristic image region from ultrasonoscopy.This blood vessel characteristic image region one
As there is the characteristics of image bigger with background image difference, thus the method for image zooming-out can include rectangular histogram,
The method such as Fourier transform, least square, is not limited thereto.Illustratively, Fig. 2 show ultrasonic blood vessel
Interior image, image 302 shown in Fig. 3 is that image walks through the image zooming-out of present embodiment in Fig. 2 ultrasonic blood vessel
Gained blood vessel characteristic image region after rapid extraction.
Step 1404, carries out polar coordinate transform by the blood vessel characteristic image extracted, to obtain this blood vessel feature
The polar coordinate transform image of image.Specifically, ultrasonoscopy the former blood vessel characteristic image extracted is straight
Angular coordinate image.Taking coordinate in the rectangular coordinate of this ultrasonoscopy isPixel as seat
Mark zero point, carries out polar coordinate transform to this blood vessel characteristic image, wherein: h is that the height of ultrasonoscopy, w are
The width of this ultrasonoscopy,Represent round numbers downward to this value.Illustratively, Fig. 4 show the blood of Fig. 3
Pipe characteristic image gained polar coordinate image after polar coordinate transform.
Step 1406, is filtered and denoising the blood vessel feature polar coordinate image after transformed, former to remove
Noise in ultrasonoscopy.Illustratively, in alternative embodiments, to described blood vessel feature polar diagram
As carrying out medium filtering to remove noise.Illustratively, the template window of medium filtering is the rectangle of predefined size
Window, in addition can to described transformed after blood vessel feature polar coordinate image carry out repeatedly Filtering Processing, many
Rectangular window size used in secondary filtering also can differ.Such as: can be 3 × 3,4 by size successively
The rectangular window template of × 4,5 × 5,6 × 6 successively to described transformed after blood vessel feature polar diagram carry out
Value filtering.As it is shown in figure 5, its to be the blood vessel feature polar diagram to Fig. 4 make use of successively 3 × 3,4 × 4,
The rectangular window module of 5 × 5,6 × 6 has carried out the filtering image of gained after 4 medium filterings.
Step 1408, sets up seed point set, and predetermined by meeting in described filter profile based on initial pixel point
The pixel of condition adds described seed point set to.Specifically, in alternative embodiments, described initially
Pixel is that on described filtering image, coordinate is the pixel of (1,1).Described predetermined condition is:
1, for seed point set be neighborhood 8 connect pixel；
2, in this pixel and filtering image coordinate be gray scale difference between the initial pixel point of (1,1) less than or etc.
In predetermined threshold value T.
Thus, corresponding seed point set can be obtained based on predetermined threshold value T, this picture corresponding to seed point set
The region S that vegetarian refreshments is formed includes N number of pixel.
Further, described predetermined threshold value T is the multiple values in given range, region growing module 122
Set up according to the plurality of predetermined threshold value T and obtain corresponding multiple region S and corresponding pixel number N.
Illustratively, in alternative embodiments, threshold value T is the continuous integral number between 1 to 255, is designated as respectively
T1=1, T2=2 ... T255=255, obtain region S the most accordingly1、S2、……S255And pixel number N1、
N2、……N255。
Step 1410, calculates seed points and concentrates pixel number according to the increment of predetermined condition.Specifically, calculate
According to the increment A of pixel N during threshold value T monotone increasing, calculate increment A such as following formula:
A1=N2-N1、A2=N3-N2、……A254=N255-N254；
Thus obtain the corresponding set of increment A corresponding to threshold value T.Illustratively, Fig. 6 show with monotone increasing
Threshold value T added is abscissa, correspondingly calculates the pass of increment A is vertical coordinate one group of threshold value-increment of gained
It it is curve.
Step 1412, the pixel region corresponding to labelling maximal increment.Specifically, from aforementioned calculated
One group of increment A selects the increment A of maximumj, and labelling this maximal increment AjCorresponding seed points is concentrated each
The region S that pixel is formedj.Illustratively, the region 702 shown in Fig. 7 is the filtering figure shown in Fig. 5
Shape after the growth, incremental computations of seed point set, the region marked by this step.
Step 1414, corrects the region of foregoing tags.Specifically, Mathematical Morphology Method pair is used
The region S of institute's labellingjCorrect.Illustratively, can be to marked region SjCarry out a series of morphology operations,
Including:
1, to region SjCarrying out morphology and close (Morphology Closing) computing, fortune is closed in described morphology
Calculating and using radius is R1Round die block enter computing, radius R1Calculating formula be:Wherein Nj
For region SjThe number of included pixel.
2, closing operation of mathematical morphology rear region S is deletedjHole；Illustratively, region 702 during described hole is Fig. 7
White portion in included black part；
3, to region SjCarrying out morphology and close (Morphology Closing) computing, fortune is closed in described morphology
Calculating and using radius is R2Round die block computing, radius R2Calculating formula be:Wherein Nj
For region SjThe number of included pixel.
Through the image that this step house of correction obtains, can clearly reflect epicardial border two in blood-vessel image medium vessels
The image difference of side.Illustratively, Fig. 8 show the marked region of Fig. 7 gained district of gained after overcorrection
Territory figure, in this figure, epicardial border interior zone in the blood vessel of region 802 reflection, itself and region 804
The middle epicardial border perimeter reflected is compared has significantly difference, it is easy to differentiate.
Step 1416, detects epicardial border in the blood vessel of described blood-vessel image according to the image after correcting.Such as figure
Shown in 9, after overcorrection, the middle epicardial border both sides image in described blood-vessel image has significantly difference,
It is prone to detect middle epicardial border 902.
Step 1418, according to epicardial border in the blood vessel detected by having carried out the filtering image after filtering and noise reduction
The image in region within epicardial border in the described blood vessel of middle selection.Specifically, with the middle adventitia shown in Fig. 9
As a example by border 902, the region above border 902 is selected to be blood in filtering image with border 902 for boundary
The image of epicardial border interior zone in pipe.
Step 1420, the specific characteristic vector of each pixel in the interior zone selected by acquisition, and according to institute
State specific characteristic vector to carry out clustering computing, with by the pixel of described tube chamber boundaries on either side area image respectively
Cluster, to form cluster areas respectively by the some set of the cluster of described cluster.Specifically, described inside is obtained
The gray value of each pixel and the standard deviation of gray value in each pixel 8 territory in region, using as this
The specific characteristic vector of pixel.Be respectively provided with feature that coordinate on described filtering image is the pixel of (1,1) to
Amount C1And characteristic vector C of arbitrary pixel on foregoing boundary 9022For cluster centre, K average is utilized to gather
The characteristic vector of pixel each in interior zone is calculated and clusters by class method, thus obtains with filtering figure
As the pixel that pixel the is cluster centre cluster G that upper coordinate is (1,1)xWith with any pixel on border 902
The pixel cluster G that point is cluster centrey.Pixel cluster GxPixel cluster GyRepresentative pixel
Form cluster areas S respectivelyxWith cluster areas Sy.Illustratively, Figure 10 show the figure that have passed through cluster calculation
The schematic diagram of the cluster areas that the Vascular Ultrasonography image of 1 is formed, in figure, region 1002 represents pixel cluster
GxCluster areas S that representative pixel is formedx, region 1004 represents pixel cluster GyRepresentative
Cluster areas S that pixel is formedy。
Step 1422, the cluster areas of formation calculated to cluster module is corrected.Specifically, can make
By Mathematical Morphology Method to cluster areas SxCorrect.Illustratively, in alternative embodiments, right
Cluster areas carries out following rectification computing:
1, cluster areas carrying out morphology and close (Morphology Closing) computing, fortune is closed in described morphology
Calculating and using radius is R3Round die block enter computing, radius R3Calculating formula be: R3=0.02 × polarwidth,
Wherein polarwidth is the horizontal size of the polar coordinate image of gained after described polar coordinate transform；
2, closing operation of mathematical morphology rear region S is deletedxHole；
3, to region SxCarrying out morphology and close (Morphology Closing) computing, fortune is closed in described morphology
Calculating and using radius is R4Round die block computing, radius R4Calculating formula be: R4=0.03 × polarwidth,
Wherein polarwidth is the horizontal size of the polar coordinate image of gained after described polar coordinate transform.
Illustratively, the image that Figure 11 is formed after overcorrection computing by the cluster areas shown in Figure 10, by
Figure 11 is it will be seen that after overcorrection computing, the vessel lumen boundaries on either side image of described blood-vessel image has
Significantly difference, it is easy to differentiate.
Step 1424, detects the tube chamber border of described blood-vessel image according to the cluster areas image after correcting.As
Shown in Figure 12, after overcorrection, the middle epicardial border both sides image in described blood-vessel image has obvious district
Not, it is easy to detect middle epicardial border 1202.
Step 1426, is transformed to epicardial border in the blood vessel under the polar coordinate system of detection and vessel lumen border
Rectangular coordinate, and then middle epicardial border and the tube chamber of blood vessel can be demonstrated under the blood-vessel image of rectangular coordinate
Border.Illustratively, shown in Figure 13 by epicardial border 902 and blood vessel pipe in the blood vessel that detected in Figure 12
Border, chamber 1202 identifies the blood-vessel image at rectangular coordinate after rectangular coordinates transformation block transforms is rectangular coordinate
On schematic diagram.
Image in ultrasonic blood vessel is made by the way of region growing by the ultrasonic blood vessel boundary detection method of the present invention
Obtain the image of epicardial border both sides in blood-vessel image to be prone to differentiate and make blood-vessel image is managed by cluster mode
Chamber boundaries on either side image is prone to differentiate, thus the middle epicardial border detected and tube chamber boundary coordinate are accurate.
Embodiment described above only have expressed the several embodiments of the present invention, and it describes more concrete and detailed,
But therefore can not be interpreted as the restriction to the scope of the claims of the present invention.It should be pointed out that, for this area
Those of ordinary skill for, without departing from the inventive concept of the premise, it is also possible to make some deformation and
Improving, these broadly fall into protection scope of the present invention.Therefore, the protection domain of patent of the present invention should be with appended
Claim is as the criterion.
Claims (21)
1. a ultrasonic blood vessel border detection system, including:
Image zooming-out module, for extracting blood vessel characteristic image from ultrasonoscopy；
Middle epicardial border detection subsystem, for being detected by region growing by described blood vessel characteristic image
Go out the border of tunica media, adventitia；
Tube chamber border detection subsystem, for the China and foreign countries detected from described middle epicardial border detection subsystem
Further by the border of cluster detection vessel lumen in membrane boundary interior zone；
Described middle epicardial border detection subsystem includes:
Region growing module, is used for setting up seed point set, and based on initial pixel point by described blood vessel feature
The pixel conformed to a predetermined condition in image adds described seed point set to, and described initial pixel point is described blood
On pipe characteristic image, coordinate is the pixel of (1,1)；
Incremental modular, concentrates pixel number according to the increment of predetermined condition, institute for calculating described seed points
Stating predetermined condition is: be the pixel of neighborhood 8 connection for seed point set, this pixel and filtering figure
In Xiang, coordinate is that the gray scale difference between the initial pixel point of (1,1) is less than or equal to predetermined threshold value T, described predetermined
Threshold value T be the multiple values in given range, described predetermined threshold value T is the most whole between 1 to 255
Number, described region growing module is set up according to multiple predetermined threshold values T and is obtained corresponding multiple region S and phase
The pixel number N answered；
Mark module, for the pixel region corresponding to the described increment that labelling is maximum；
Described region growing module sets up the multiple seed point sets of formation, described increment according to multiple predetermined conditions
Module calculates pixel number that the plurality of seed points concentrates according to the monotone increasing of the plurality of predetermined condition
The increment produced；
In the plurality of seed point set, each pixel forms corresponding image-region, and described mark module is used for
The region that the pixel corresponding to described increment of labelling maximum is formed；
Described tube chamber border detection subsystem includes:
Internal selection module, for the blood vessel China and foreign countries detected according to described middle epicardial border detection subsystem
Membrane boundary is by the image in the region selected in described blood vessel characteristic image in described blood vessel within epicardial border；
Cluster module, for obtaining the specific characteristic vector of each pixel in described images of interior regions, and
Carry out clustering computing according to described specific characteristic vector；
Described specific characteristic vector includes the gray value of each pixel and each pixel 8 in described interior zone
The standard deviation of gray value in territory；
The pixel of described tube chamber boundaries on either side area image is clustered by described cluster module respectively, with by institute
The cluster point set stating cluster forms cluster areas respectively；
Described cluster computing with coordinate on described blood vessel characteristic image be the pixel of (1,1) specific characteristic to
In amount and described blood vessel, the take up an official post characteristic vector of a pixel of epicardial border is cluster centre；
Described cluster module utilizes the appointment to pixel each in described interior zone of the K means clustering method
Characteristic vector calculates and clusters, and each pixel cluster forms cluster areas respectively.
2. ultrasonic blood vessel border detection system as claimed in claim 1, farther includes polar coordinate transform mould
Block, for carrying out polar coordinate transform by the blood vessel characteristic image extracted, to obtain described blood vessel characteristic image
Polar coordinate transform image, described middle epicardial border detection subsystem, described tube chamber border detection subsystem divide
Not according to described polar coordinate transform image detects the middle epicardial border of described blood vessel and the limit of described vessel lumen
Boundary.
3. ultrasonic blood vessel border detection system as claimed in claim 2, it is characterised in that: described polar coordinate
Conversion module takes coordinate in the rectangular coordinate of this blood vessel characteristic imagePixel conduct
Co-ordinate zero point, carries out polar coordinate transform to this blood vessel characteristic image, wherein: h is the height of ultrasonoscopy, w
For this ultrasonoscopy width,Represent downward round numbers.
4. ultrasonic blood vessel border detection system as claimed in claim 2, farther includes rectangular coordinates transformation
Module, for being transformed to right angle by epicardial border in the blood vessel under the polar coordinate system of detection and vessel lumen border
Coordinate.
5. ultrasonic blood vessel border detection system as claimed in claim 1, farther includes filtering and noise reduction module,
For described blood vessel characteristic image is filtered and denoising.
6. ultrasonic blood vessel border detection system as claimed in claim 5, it is characterised in that: described filtering is gone
Module of making an uproar carries out medium filtering to remove noise to described blood vessel characteristic image.
7. ultrasonic blood vessel border detection system as claimed in claim 6, it is characterised in that: described filtering is gone
Module of making an uproar carries out the rectangular window that template window is predefined size of described medium filtering.
8. ultrasonic blood vessel border detection system as claimed in claim 6, it is characterised in that: described filtering is gone
Module of making an uproar carries out the most described medium filtering to described blood vessel characteristic image.
9. ultrasonic blood vessel border detection system as claimed in claim 8, it is characterised in that: described filtering is gone
It is 3 × 3,4 × 4,5 that module of making an uproar carries out the template window of the most described medium filtering to described blood vessel characteristic image
The rectangular window template of × 5,6 × 6.
10. ultrasonic blood vessel border detection system as claimed in claim 1, farther includes rectification module,
For the region of described mark module institute labelling is corrected.
11. ultrasonic blood vessel border detection systems as claimed in claim 10, it is characterised in that: described rectification
Module uses Mathematical Morphology Method to correct the region of institute's labelling.
12. ultrasonic blood vessel border detection systems as claimed in claim 11, it is characterised in that: described rectification
Module carries out closing operation of mathematical morphology to the region of described institute labelling, and described closing operation of mathematical morphology uses radius to be R1
Round die block computing, radius R1Calculating formula be:Wherein NjDistrict for described institute labelling
The number of the pixel included by territory.
13. ultrasonic blood vessel border detection systems as claimed in claim 12, it is characterised in that: described rectification
The hole in described region after the module described closing operation of mathematical morphology of deletion.
14. ultrasonic blood vessel border detection systems as claimed in claim 11, it is characterised in that: described rectification
Module carries out closing operation of mathematical morphology to the region of described institute labelling, and described closing operation of mathematical morphology uses radius to be R2
Round die block computing, radius R2Calculating formula be:Wherein NjDistrict for described institute labelling
The number of the pixel included by territory.
15. ultrasonic blood vessel border detection systems as claimed in claim 1, farther include rectification module,
For described cluster module being clustered vessel lumen interior zone described in the representative that each pixel of gained is formed
Cluster areas is corrected.
16. ultrasonic blood vessel border detection systems as claimed in claim 15, it is characterised in that: described rectification
Module uses Mathematical Morphology Method to correct described cluster areas.
17. ultrasonic blood vessel border detection systems as claimed in claim 16, it is characterised in that: described rectification
Module carries out closing operation of mathematical morphology to described cluster areas, and described closing operation of mathematical morphology uses radius to be R3Circle
Shape module arithmetic, described radius R3Calculating formula be: R3=0.02 × polarwidth, wherein polarwidth
It it is described blood vessel characteristic image horizontal size of the polar coordinate image of gained after polar coordinate transform converts.
18. ultrasonic blood vessel border detection systems as claimed in claim 17, it is characterised in that: described rectification
The hole of described cluster areas after module deletion closing operation of mathematical morphology.
19. ultrasonic blood vessel border detection systems as claimed in claim 16, it is characterised in that: described rectification
Module carries out closing operation of mathematical morphology to described cluster areas, and described closing operation of mathematical morphology uses radius to be R4Circle
Shape module arithmetic, radius R4Calculating formula be: R4=0.03 × polarwidth, wherein polarwidth is through institute
State blood vessel characteristic image horizontal size of the polar coordinate image of gained after polar coordinate transform converts.
20. ultrasonic blood vessel border detection systems as claimed in claim 1, farther include Boundary Extraction mould
Block, epicardial border and the border of described vessel lumen in the blood vessel detecting described blood vessel characteristic image.
21. 1 kinds of ultrasonic blood vessel boundary detection methods, it is characterised in that comprise the steps: from ultrasonic figure
Blood vessel characteristic image region is extracted in Xiang；
The blood vessel characteristic image extracted is carried out polar coordinate transform, to obtain the pole seat of this blood vessel characteristic image
Mark changing image；
Blood vessel feature polar coordinate image after transformed is filtered and denoising, to remove in former ultrasonoscopy
Noise；
Set up seed point set, and the pixel that will conform to a predetermined condition in filtered image based on initial pixel point
Point adds described seed point set to；
Pixel region corresponding to labelling maximal increment；
The region of labelling is corrected；
According to epicardial border in the blood vessel of the image detection blood vessel characteristic image after correcting；
Described by the filtering image after having carried out filtering and noise reduction selects according to epicardial border in the blood vessel detected
The image in region within epicardial border in blood vessel；
The specific characteristic vector of each pixel in interior zone selected by acquisition, and according to described specific characteristic
Vector carries out clustering computing, clusters respectively with the pixel by tube chamber boundaries on either side area image, with by described
The cluster point set of cluster forms cluster areas respectively；
Cluster areas is corrected；
Tube chamber border according to the cluster areas image detection blood vessel characteristic image after correcting；
Epicardial border in blood vessel under the polar coordinate system of detection and vessel lumen border are transformed to rectangular coordinate,
And then middle epicardial border and the tube chamber border of blood vessel can be demonstrated under the blood-vessel image of rectangular coordinate.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201210584625.8A CN103093457B (en) | 2012-12-28 | 2012-12-28 | Ultrasonic blood vessel border detection System and method for |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201210584625.8A CN103093457B (en) | 2012-12-28 | 2012-12-28 | Ultrasonic blood vessel border detection System and method for |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN103093457A CN103093457A (en) | 2013-05-08 |
| CN103093457B true CN103093457B (en) | 2016-08-24 |
Family
ID=48205988
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201210584625.8A Active CN103093457B (en) | 2012-12-28 | 2012-12-28 | Ultrasonic blood vessel border detection System and method for |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN103093457B (en) |
Families Citing this family (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104182984B (en) * | 2014-09-01 | 2017-02-01 | 云南大学 | Method and system for rapidly and automatically collecting blood vessel edge forms in dynamic ultrasonic image |
| CN105708495B (en) * | 2016-01-26 | 2018-08-17 | 飞依诺科技(苏州)有限公司 | Ultrasonic color blood flow imaging boundary processing method and system |
| CN106920245B (en) * | 2017-03-13 | 2020-01-14 | 深圳怡化电脑股份有限公司 | Boundary detection method and device |
| CN107346545A (en) * | 2017-05-22 | 2017-11-14 | 沈阳工业大学 | Improved confinement growing method for the segmentation of optic cup image |
| CN109410191B (en) * | 2018-10-18 | 2022-03-25 | 中南大学 | OCT (optical coherence tomography) image-based fundus blood vessel positioning method and anemia screening method thereof |
| CN109978790A (en) * | 2019-03-26 | 2019-07-05 | 数坤（北京）网络科技有限公司 | A kind of calcification artifact eliminating method, equipment and computer storage medium |
| CN110490846B (en) * | 2019-07-30 | 2023-01-06 | 武汉睿芯特种光纤有限责任公司 | Method and device for testing geometric dimension of optical fiber image |
| CN115100696A (en) * | 2022-08-29 | 2022-09-23 | 山东圣点世纪科技有限公司 | Connected domain rapid marking and extracting method and system in palm vein recognition |
Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101833757A (en) * | 2009-03-11 | 2010-09-15 | 深圳迈瑞生物医疗电子股份有限公司 | Method and system for detection edge of blood vessel graphic tissue structure and blood vessel endangium |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US7831078B2 (en) * | 2006-07-24 | 2010-11-09 | Siemens Medical Solutions Usa, Inc. | System and method for statistical shape model based segmentation of intravascular ultrasound and optical coherence tomography images |
-
2012
- 2012-12-28 CN CN201210584625.8A patent/CN103093457B/en active Active
Patent Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101833757A (en) * | 2009-03-11 | 2010-09-15 | 深圳迈瑞生物医疗电子股份有限公司 | Method and system for detection edge of blood vessel graphic tissue structure and blood vessel endangium |
Non-Patent Citations (3)
| Title |
|---|
| 新型血管内超声图像处理关键技术研究;董海艳;《万方学位论文数据库》;20101013;正文第8-9,50-58页 * |
| 结合图像增强的心血管内超声中-外膜边缘检测;邱璇 等;《中国图象图形学报》;20120627;第17卷(第4期);第537-545页 * |
| 结合硬斑块特征的心血管内超声中-外膜边缘检测;邢栋 等;《中国生物医学工程学报》;20120420;第31卷(第1期);第25-31页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN103093457A (en) | 2013-05-08 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN103093457B (en) | Ultrasonic blood vessel border detection System and method for | |
| Mendonça et al. | Automatic localization of the optic disc by combining vascular and intensity information | |
| JP5498299B2 (en) | System and method for providing 2D CT images corresponding to 2D ultrasound images | |
| CN106651827B (en) | A kind of ocular fundus image registration method based on SIFT feature | |
| CN111292338B (en) | Method and system for segmenting choroidal neovascularization from fundus OCT image | |
| Molinari et al. | Completely automated multiresolution edge snapper—a new technique for an accurate carotid ultrasound IMT measurement: clinical validation and benchmarking on a multi-institutional database | |
| WO2022052303A1 (en) | Method, apparatus and device for registering ultrasound image and ct image | |
| CN102982542B (en) | Fundus image vascular segmentation method based on phase congruency | |
| US20180110491A1 (en) | Dynamic image processor | |
| CN103810709A (en) | Vessel-based registration method for eye fundus image and SD-OCT projection image | |
| CN109685770A (en) | Retinal vessel curvature determines method | |
| CN105139377A (en) | Rapid robustness auto-partitioning method for abdomen computed tomography (CT) sequence image of liver | |
| CN102034115B (en) | Markov random field model and non-local prior based image registration method | |
| Zhang et al. | Optic disc localization by projection with vessel distribution and appearance characteristics | |
| CN105021196A (en) | Crop row detection method based minimum tangent circle and morphological principle | |
| CN106778499B (en) | Method for rapidly positioning human iris in iris acquisition process | |
| CN102393956B (en) | Retinal image filter method based on radial symmetry transformation | |
| Rocha et al. | Automatic detection of the carotid lumen axis in B-mode ultrasound images | |
| CN104915989B (en) | Blood vessel three-dimensional dividing method based on CT images | |
| CN109166127A (en) | A kind of wearable plant phenotype sensory perceptual system | |
| CN108665474B (en) | B-COSFIRE-based retinal vessel segmentation method for fundus image | |
| CN112183518A (en) | Vein target point automatic determination method, device and equipment | |
| CN102663749A (en) | Ultrasonic image lesion segmentation technology based on phase information | |
| US10062164B2 (en) | Method for the analysis of image data representing a three-dimensional volume of biological tissue | |
| Molinari et al. | Completely automated multiresolution edge snapper (CAMES): A new technique for an accurate carotid ultrasound IMT measurement and its validation on a multi-institutional database |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant | ||
| TR01 | Transfer of patent right | ||
| TR01 | Transfer of patent right |
Effective date of registration: 20200623 Address after: Office building of Shenzhen Institute of advanced technology A-207 518000 in Guangdong city of Shenzhen province Nanshan District City Road No. 1068 Chinese Academy of Shenzhen University Academy of Sciences Patentee after: Shenzhen advanced science and technology Cci Capital Ltd. Address before: 1068 No. 518055 Guangdong city in Shenzhen Province, Nanshan District City Xili University School Avenue Patentee before: SHENZHEN INSTITUTES OF ADVANCED TECHNOLOGY CHINESE ACADEMY OF SCIENCES |
|
| TR01 | Transfer of patent right | ||
| TR01 | Transfer of patent right |
Effective date of registration: 20201028 Address after: 519000 the 12 floor of A District, No. 1, Port Road, Tangjia Bay, Zhuhai High-tech Zone, Guangdong. Patentee after: ZHUHAI INSTITUTE OF ADVANCED TECHNOLOGY CHINESE ACADEMY OF SCIENCES Co.,Ltd. Address before: Office building of Shenzhen Institute of advanced technology A-207 518000 in Guangdong city of Shenzhen province Nanshan District City Road No. 1068 Chinese Academy of Shenzhen University Academy of Sciences Patentee before: Shenzhen advanced science and technology Cci Capital Ltd. |
|
| TR01 | Transfer of patent right | ||
| TR01 | Transfer of patent right |
Effective date of registration: 20210420 Address after: Room 1102, building 5, Longyuan intelligent industrial park, No.2, hagongda Road, Tangjiawan Town, high tech Zone, Zhuhai City, Guangdong Province, 519000 Patentee after: Zhuhai Zhongke advanced technology industry Co.,Ltd. Address before: 519000 the 12 floor of A District, No. 1, Port Road, Tangjia Bay, Zhuhai High-tech Zone, Guangdong. Patentee before: ZHUHAI INSTITUTE OF ADVANCED TECHNOLOGY CHINESE ACADEMY OF SCIENCES Co.,Ltd. |