CN108416792B - Medical computed tomography image segmentation method based on active contour model - Google Patents

Medical computed tomography image segmentation method based on active contour model Download PDF

Info

Publication number
CN108416792B
CN108416792B CN201810038957.3A CN201810038957A CN108416792B CN 108416792 B CN108416792 B CN 108416792B CN 201810038957 A CN201810038957 A CN 201810038957A CN 108416792 B CN108416792 B CN 108416792B
Authority
CN
China
Prior art keywords
function
equation
image
given
definition
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
Application number
CN201810038957.3A
Other languages
Chinese (zh)
Other versions
CN108416792A (en
Inventor
王相海
方玲玲
宋传鸣
孙杨
吕芳
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Liaoning Normal University
Original Assignee
Liaoning Normal University
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Liaoning Normal University filed Critical Liaoning Normal University
Priority to CN201810038957.3A priority Critical patent/CN108416792B/en
Publication of CN108416792A publication Critical patent/CN108416792A/en
Application granted granted Critical
Publication of CN108416792B publication Critical patent/CN108416792B/en
Expired - Fee Related legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/10—Segmentation; Edge detection
    • G06T7/149—Segmentation; Edge detection involving deformable models, e.g. active contour models
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/10—Segmentation; Edge detection
    • G06T7/136—Segmentation; Edge detection involving thresholding
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/10—Image acquisition modality
    • G06T2207/10072—Tomographic images
    • G06T2207/10081—Computed x-ray tomography [CT]
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/10—Image acquisition modality
    • G06T2207/10072—Tomographic images
    • G06T2207/10088—Magnetic resonance imaging [MRI]
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/10—Image acquisition modality
    • G06T2207/10072—Tomographic images
    • G06T2207/10104—Positron emission tomography [PET]

Landscapes

  • Engineering & Computer Science (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Software Systems (AREA)
  • Apparatus For Radiation Diagnosis (AREA)

Abstract

The invention discloses a medical computed tomography image segmentation method based on an active contour model, and belongs to the field of medical image processing. Firstly, a GAC model and an SBGFRLS model are combined by using a weight function, so that the possibility of zero-level set inspection of the multi-layer contour of the medical image is improved; secondly, in order to further simplify the level set algorithm in the model, a Gaussian filtering technology is introduced, and the convergence rate of the segmentation model is improved; and finally, the Hausdorff distance is introduced to measure the distance between the targets, so that the problem of poor segmentation effect of the traditional model on the convex region in the medical image is solved, the capture capability of the multi-target object edge in the medical computed tomography image is improved, and the segmentation precision is high.

Description

Medical computed tomography image segmentation method based on active contour model
Technical Field
The invention relates to the field of image processing, in particular to a medical computed tomography image segmentation method based on an active contour model, which has high convergence rate and high robustness to initial contour curve positions and noise.
Background
Image segmentation is one of the most basic and important problems in the field of image processing. Although many researchers have proposed various image segmentation methods such as a morphological watershed-based segmentation method, a grayscale threshold-based segmentation method, an edge detection-based segmentation method, a partial differential equation-based segmentation method, and the like, an image segmentation method suitable for all applications has not appeared at present.
With the continuous innovation and development of medical imaging equipment and examination technology, digital imaging enables computer-aided detection and computer-aided diagnosis, and imaging examination is increasingly applied to clinical disease diagnosis, and especially shows more and more important roles in tumor 'precise medical treatment' planning. In order to accurately judge the disease condition of a patient, a doctor needs to effectively segment a lesion site included in a medical image. Medical image segmentation is the basis for extracting a mutation region, measuring a specific tissue and realizing reconstruction of three dimensions, is an important step for distinguishing and sketching a focus region, is also the basis and the key of accurate medical treatment, and the quality of segmentation directly influences whether a clinician diagnoses the state of an illness of a patient. However, according to the introduction of the imaging department doctor, on one hand, due to the influence of many factors such as the physiology of the patient, the body position movement, the acquisition system and the like, the quality of the acquired medical image is sometimes low; on the other hand, due to the different tissues to be segmented and the different imaging mechanisms, the segmentation methods are often inconsistent. Therefore, when the clinical treatment is carried out, the doctor still needs to segment the focus and the peripheral tissues of the patient through manual operation, which wastes time and labor and affects the diagnosis efficiency of the clinician. Moreover, although Computed Tomography (CT) images have high resolution and are relatively accurate in locating lesion regions, doctors have certain difficulty in determining tumor positions and boundaries when there is no significant difference in the density between tissues with tumor infiltration and normal tissues. As such, there is an urgent need for a practical segmentation method for specific medical images.
Jehan-Besson et al propose a medical image segmentation method based on an extreme learning machine, which reduces segmentation time, but has low precision. The royal yog et al propose a multi-threshold medical image segmentation algorithm based on one-dimensional Otsu, and though the segmentation precision is improved to a certain extent, the segmentation effect on the deep convex region is poor. In view of the characteristics of uneven gray distribution, fuzzy edge, high noise intensity and the like of medical images, variation models have shown greater and greater application potential in recent years. The variation model is a model driven by 'internal force' constraint and 'external force', and has a good segmentation effect on medical images. The Mumford and J.Shah firstly propose a Mumford-Shah variation model, and firstly expand a partial differential model based on image recovery to the image segmentation field. Chan et al simplify the Mumford-Shah model and then propose a CV model, which has the advantages that the segmentation result is independent of the initial contour position, the boundary of the image can be accurately extracted, and the robustness to noise is high. However, since the level set function is introduced, the computation amount of the CV model is high, the level set function needs to be reinitialized, and the gradient descent is easy to fall into a local minimum value. In addition, since the CV model is based on global gray scale fitting, it is not suitable for medical image segmentation with uneven gray scale distribution. For this reason, Li et al propose a RSF (regional Scalable fixing) model containing a regional Scalable energy term, which is capable of segmenting images with non-uniform gray levels, but is sensitive to the initial position of the active contour and noise. Zhang et al further introduces global information of the CV model into a gac (geodesic Active content) model, proposes a Level Set segmentation model sbgfrls (selected and filtered regulated Level Set gaussian) model based on a region, considers region information and edge information of an image, reduces sensitivity of the model to an initial contour position, but cannot effectively segment heterogeneous images with uneven gray levels, such as medical images.
Disclosure of Invention
The invention aims to solve the technical problems in the prior art and provides a medical computed tomography image segmentation method based on an active contour model, which has high convergence rate and high robustness on the initial contour curve position and noise.
The technical solution of the invention is as follows: a medical computed tomography image segmentation method based on an active contour model is characterized by comprising the following steps:
step 1, establishing a moving contour segmentation model of the medical computed tomography image, wherein the definition of a level set evolution equation is given by a formula (1):
Figure 324737DEST_PATH_IMAGE001
(1)
the above-mentioned
Figure 307737DEST_PATH_IMAGE002
Representing the sign pressure function for controlling the direction of evolution of the curve,
Figure 640629DEST_PATH_IMAGE003
representing the image in coordinates
Figure 554534DEST_PATH_IMAGE004
The value of the pixel of (a) is,
Figure 580259DEST_PATH_IMAGE005
the function of the level set is represented,
Figure 784976DEST_PATH_IMAGE006
the gradient operator is represented by a gradient operator,
Figure 92198DEST_PATH_IMAGE007
representing information about an initial input image
Figure 698760DEST_PATH_IMAGE008
The weight function of (a) is determined,
Figure 895386DEST_PATH_IMAGE009
indicating an expectation of 0 and a standard deviation of
Figure 760967DEST_PATH_IMAGE010
The gaussian kernel function of (a) is,
Figure 701241DEST_PATH_IMAGE011
the divergence operator is represented by a vector of vectors,
Figure 162310DEST_PATH_IMAGE012
represents the normalized Hausdorff distance,
Figure 700476DEST_PATH_IMAGE013
and
Figure 879785DEST_PATH_IMAGE014
are all scale factors, wherein the weight function
Figure 358171DEST_PATH_IMAGE015
Is given by equation (2):
Figure 112893DEST_PATH_IMAGE016
(2)
the above-mentioned
Figure 651322DEST_PATH_IMAGE017
Indicating an expectation of 0 and a standard deviation of
Figure 488566DEST_PATH_IMAGE010
A gaussian kernel function of "
Figure DEST_PATH_IMAGE018
"represents a convolution operation, and the definition of the normalized Hausdorff distance is given by equation (3):
Figure 708326DEST_PATH_IMAGE019
(3)
the above-mentioned
Figure 583134DEST_PATH_IMAGE020
Representing the Hausdorff distance between the gray information in the inner and outer local regions of the active contour curve, the definition of which is given by equation (4):
Figure 26885DEST_PATH_IMAGE021
(4)
the above-mentioned
Figure 180786DEST_PATH_IMAGE022
Representing two sets of discrete points
Figure 437192DEST_PATH_IMAGE023
And
Figure 524097DEST_PATH_IMAGE024
is given by equation (5):
Figure 279695DEST_PATH_IMAGE025
(5)
the above-mentioned
Figure 168496DEST_PATH_IMAGE026
A set of pixels representing an inner region of the active contour curve,
Figure 730058DEST_PATH_IMAGE027
set of pixels, coordinates, representing the outer region of the active contour curve
Figure 874732DEST_PATH_IMAGE004
The pixel value of (b) is given by equation (6) -equation (7):
Figure 424400DEST_PATH_IMAGE028
(6)
Figure 224997DEST_PATH_IMAGE029
(7)
the above-mentioned
Figure 386988DEST_PATH_IMAGE030
Represents the Heaviside function, the definition of which is given by equation (8):
Figure 90895DEST_PATH_IMAGE031
(8)
step 2, inputting the medical computed tomography image to be segmented
Figure 312929DEST_PATH_IMAGE032
Setting a relevant Gaussian kernel function and calculating a weight function by using the formula (2)
Figure 833778DEST_PATH_IMAGE015
；
Step 3, if
Figure 799460DEST_PATH_IMAGE032
If the image is a CT image or an MRI image, the step 3.1 is carried out; if it is
Figure 918725DEST_PATH_IMAGE032
Turning to step 3.2 for the PET image;
step 3.1 calculation using equation (9)
Figure 750809DEST_PATH_IMAGE002
The value of the function:
Figure 588315DEST_PATH_IMAGE033
(9)
the above-mentioned
Figure DEST_PATH_IMAGE034
And
Figure 466010DEST_PATH_IMAGE035
the average pixel values representing the inner and outer regions of the active contour, respectively, are defined by equations (10) and (11), respectively:
Figure 439782DEST_PATH_IMAGE036
(10)
Figure 3618DEST_PATH_IMAGE037
(11)
turning to the step 4;
step 3.2 calculation using equation (12)
Figure 501989DEST_PATH_IMAGE002
The value of the function:
Figure 809474DEST_PATH_IMAGE038
(12)
the above-mentioned
Figure 903332DEST_PATH_IMAGE039
A gray threshold value representing a judgment tumor region, whose definition is given by equation (13):
Figure 74288DEST_PATH_IMAGE040
(13)
the above-mentioned
Figure 620807DEST_PATH_IMAGE041
The derivative function, representing the Weibull probability distribution function, is defined by equation (14):
Figure 466403DEST_PATH_IMAGE042
(14)
the above-mentioned
Figure 842197DEST_PATH_IMAGE043
Is a shape parameter of the probability distribution function,
Figure 747836DEST_PATH_IMAGE044
scale parameters of the probability distribution function;
step 4. initialize level set function
Figure 781651DEST_PATH_IMAGE045
And make an order
Figure 867156DEST_PATH_IMAGE046
，
Figure 935606DEST_PATH_IMAGE047
；
Step 5, calculating the normalized Hausdorff distance calculation of the pixel values of the inner and outer regions of the active contour by using a formula (3);
step 6, updating the level set function by using a finite difference method and a formula (1)
Figure 12147DEST_PATH_IMAGE048
；
Step 7, checking whether the evolution curve is stably converged, if so, stopping iteration, and ending the algorithm; otherwise, go to step 3.
Compared with the prior art, the invention has the following advantages: firstly, the convergence rate of the segmentation is improved by utilizing the normalized Hausdorff distance; secondly, the introduction of Gaussian filtering simplifies the evolution process of the level set; thirdly, the capability of detecting the multilayer outline of the object by adopting the weight function is improved, and the capturing capability of the deep convex region and the multi-target object edge in the computed tomography image is also improved.
Drawings
FIG. 1 is a graph comparing the CT or MRI image segmentation results with other methods according to the present invention.
FIG. 2 is a graph comparing the segmentation results of PET images according to the embodiment of the present invention and other methods.
Detailed Description
The invention relates to a medical computed tomography image segmentation method based on an active contour model, which is carried out according to the following steps:
step 1, establishing a moving contour segmentation model of the medical computed tomography image, wherein the definition of a level set evolution equation is given by a formula (1):
Figure 267679DEST_PATH_IMAGE001
(1)
the above-mentioned
Figure 159805DEST_PATH_IMAGE002
Representing the sign pressure function for controlling the direction of evolution of the curve,
Figure 82761DEST_PATH_IMAGE003
representing the image in coordinates
Figure 766421DEST_PATH_IMAGE004
The value of the pixel of (a) is,
Figure 509250DEST_PATH_IMAGE005
the function of the level set is represented,
Figure 500339DEST_PATH_IMAGE006
the gradient operator is represented by a gradient operator,
Figure 277802DEST_PATH_IMAGE007
representing information about an initial input image
Figure 135293DEST_PATH_IMAGE008
The weight function of (a) is determined,
Figure 630997DEST_PATH_IMAGE009
indicating an expectation of 0 and a standard deviation of
Figure 160198DEST_PATH_IMAGE010
The gaussian kernel function of (a) is,
Figure 493965DEST_PATH_IMAGE011
the divergence operator is represented by a vector of vectors,
Figure 83210DEST_PATH_IMAGE012
represents the normalized Hausdorff distance,
Figure 800630DEST_PATH_IMAGE013
and
Figure 431725DEST_PATH_IMAGE014
are all scale factors, wherein the weight function
Figure 324726DEST_PATH_IMAGE015
Is given by equation (2):
Figure 881609DEST_PATH_IMAGE016
(2)
the above-mentioned
Figure 788123DEST_PATH_IMAGE017
Indicating an expectation of 0 and a standard deviation of
Figure 659127DEST_PATH_IMAGE010
A gaussian kernel function of "
Figure 857501DEST_PATH_IMAGE018
"represents a convolution operation, and the definition of the normalized Hausdorff distance is given by equation (3):
Figure 428028DEST_PATH_IMAGE019
(3)
the above-mentioned
Figure 385620DEST_PATH_IMAGE020
Representing the Hausdorff distance between the gray information in the inner and outer local regions of the active contour curve, the definition of which is given by equation (4):
Figure 263577DEST_PATH_IMAGE021
(4)
the above-mentioned
Figure 226110DEST_PATH_IMAGE022
Representing two sets of discrete points
Figure 328059DEST_PATH_IMAGE023
And
Figure 445050DEST_PATH_IMAGE024
is given by equation (5):
Figure 421971DEST_PATH_IMAGE025
(5)
the above-mentioned
Figure 675229DEST_PATH_IMAGE026
A set of pixels representing an inner region of the active contour curve,
Figure 213658DEST_PATH_IMAGE027
set of pixels, coordinates, representing the outer region of the active contour curve
Figure 53831DEST_PATH_IMAGE004
The pixel value of (b) is given by equation (6) -equation (7):
Figure 335908DEST_PATH_IMAGE028
(6)
Figure 942208DEST_PATH_IMAGE029
(7)
the above-mentioned
Figure 651538DEST_PATH_IMAGE030
Represents the Heaviside function, the definition of which is given by equation (8):
Figure 477542DEST_PATH_IMAGE031
(8)
step 2, inputting the medical computed tomography image to be segmented
Figure 799196DEST_PATH_IMAGE032
Setting the associated Gaussian kernel functionAnd calculating a weight function using the formula (2)
Figure 27046DEST_PATH_IMAGE015
；
Step 3, if
Figure 907277DEST_PATH_IMAGE032
If the image is a CT image or an MRI image, the step 3.1 is carried out; if it is
Figure 719113DEST_PATH_IMAGE032
Turning to step 3.2 for the PET image;
step 3.1 calculation using equation (9)
Figure 342993DEST_PATH_IMAGE002
The value of the function:
Figure 487666DEST_PATH_IMAGE033
(9)
the above-mentioned
Figure 966228DEST_PATH_IMAGE034
And
Figure 766825DEST_PATH_IMAGE035
the average pixel values representing the inner and outer regions of the active contour, respectively, are defined by equations (10) and (11), respectively:
Figure 427351DEST_PATH_IMAGE036
(10)
Figure 692110DEST_PATH_IMAGE037
(11)
turning to the step 4;
step 3.2 calculation using equation (12)
Figure 586248DEST_PATH_IMAGE002
The value of the function:
Figure 437923DEST_PATH_IMAGE038
(12)
the above-mentioned
Figure 341288DEST_PATH_IMAGE039
A gray threshold value representing a judgment tumor region, whose definition is given by equation (13):
Figure 460554DEST_PATH_IMAGE040
(13)
the above-mentioned
Figure 289707DEST_PATH_IMAGE041
The derivative function, representing the Weibull probability distribution function, is defined by equation (14):
Figure 861634DEST_PATH_IMAGE042
(14)
the above-mentioned
Figure 804575DEST_PATH_IMAGE043
Is a shape parameter of the probability distribution function,
Figure 43927DEST_PATH_IMAGE044
scale parameters of the probability distribution function;
step 4. initialize level set function
Figure 607763DEST_PATH_IMAGE045
And make an order
Figure 103205DEST_PATH_IMAGE046
，
Figure 410689DEST_PATH_IMAGE047
；
Step 5, calculating the normalized Hausdorff distance calculation of the pixel values of the inner and outer regions of the active contour by using a formula (3);
step 6, updating the level set function by using a finite difference method and a formula (1)
Figure 442230DEST_PATH_IMAGE048
；
Step 7, checking whether the evolution curve is stably converged, if so, stopping iteration, and ending the algorithm; otherwise, go to step 3.
The embodiment of the invention and other methods compare the segmentation results of CT or MRI images as shown in figure 1: from left to right are (a) original images; (b) the embodiment of the invention divides the result; (c) a segmentation result based on an extreme learning machine proposed by Jehan-Besson; (c) the one-dimensional Otsu-based multi-threshold segmentation result proposed by Queen yog.
The embodiment of the invention and other methods compare the segmentation results of the PET image as shown in FIG. 2: from left to right are (a) original images; (b) the embodiment of the invention divides the result; (c) a segmentation result based on an extreme learning machine proposed by Jehan-Besson; (c) the one-dimensional Otsu-based multi-threshold segmentation result proposed by Queen yog.
The iteration number and time ratio required for segmenting 3 CT or MRI images (figure 1) by the embodiment of the invention and other methods are shown in table 1.
The error rate ratio of 3 CT or MRI images (FIG. 1) segmented by the embodiment of the invention and other methods is shown in Table 2.
The number of iterations and time ratio required to segment 3 PET images (fig. 2) for embodiments of the present invention and other methods are shown in table 3.
The error rate ratio of 3 PET images (FIG. 2) segmented by the embodiment of the present invention and other methods is shown in Table 4.
TABLE 1 iteration count vs. time (units: seconds) for the 3 images of FIG. 1
Figure 678433DEST_PATH_IMAGE049
TABLE 2 segmentation error Rate comparison of the 3 images of FIG. 1
Figure 162635DEST_PATH_IMAGE050
TABLE 3 iteration count vs. time (units: seconds) for the 3 images of FIG. 2
Figure 178870DEST_PATH_IMAGE051
TABLE 4 segmentation error Rate comparison of the 3 images of FIG. 2
Figure 127235DEST_PATH_IMAGE052
。
The comparison result shows that: the invention can more accurately segment the heterogeneous medical computed tomography image with low contrast and uneven gray scale in a short time.

Claims (1)

1. A medical computed tomography image segmentation method based on an active contour model is characterized by comprising the following steps:
step 1, establishing a moving contour segmentation model of a medical computed tomography image, wherein the definition of a level set evolution equation is given by a formula (1):
Figure FDA0003000373290000011
spf (-) denotes a symbolic pressure function for controlling the direction of curve evolution, i (x) denotes the pixel value of the image at coordinate x, phi denotes a level set function,
Figure FDA0003000373290000017
representing the gradient operator, v (u)0(x) Is shown with respect to the initial input image u0(x) Weight function of δσ(. represents desirably 0,Gaussian kernel function with standard deviation σ, div denotes the divergence operator, DHExpressing the normalized Hausdorff distance, α and μ are both scale factors, where the definition of the weight function v is given by equation (2):
Figure FDA0003000373290000012
the G isσRepresenting a gaussian kernel with a desired 0 and standard deviation σ, "# represents a convolution operation, and the definition of the normalized Hausdorff distance is given by equation (3):
Figure FDA0003000373290000013
d isHRepresenting the Hausdorff distance between the gray information in the inner and outer local regions of the active contour curve, the definition of which is given by equation (4):
dH＝max(d(Ω1，Ω2)，d(Ω2，Ω1)) (4)
the d represents the distance of two discrete point sets a and B, the definition of which is given by equation (5):
Figure FDA0003000373290000014
the omega1Set of pixels, Ω, representing the inner region of the active contour curve2A set of pixels representing the area outside the active contour curve, the pixel value at coordinate x being given by equation (6) -equation (7):
Figure FDA0003000373290000015
Figure FDA0003000373290000016
said Hε(φ) represents the Heaviside function, the definition of which is given by equation (8):
Figure FDA0003000373290000021
step 2, inputting a medical computed tomography image u to be segmented0(x) Setting a relevant Gaussian kernel function, and calculating a weight function v by using a formula (2);
step 3. if u0(x) If the image is a CT image or an MRI image, the step 3.1 is carried out; if u0(x) Turning to step 3.2 for the PET image;
step 3.1 calculates the value of the spf (-) function using equation (9):
Figure FDA0003000373290000022
c is mentioned1And c2The average pixel values representing the inner and outer regions of the active contour, respectively, are defined by equations (10) and (11), respectively:
Figure FDA0003000373290000023
Figure FDA0003000373290000024
turning to the step 4;
step 3.2 calculates the value of the spf (-) function using equation (12):
Figure FDA0003000373290000025
the th represents a threshold value of the gray scale for judging the tumor region, and the definition is given by equation (13):
Figure FDA0003000373290000026
the f' (a) represents the derivative function of the Weibull probability distribution function, the definition of which is given by equation (14):
Figure FDA0003000373290000027
k is a shape parameter of the probability distribution function, and lambda is a scale parameter of the probability distribution function;
step 4, initializing a level set function phi (x, t) to be 0, and making alpha to be-0.1 and mu to be 0.1;
step 5, calculating the normalized Hausdorff distance calculation of the pixel values of the inner and outer regions of the active contour by using a formula (3);
step 6, updating the level set function phi (x, t) by using a finite difference method and a formula (1);
step 7, checking whether the evolution curve is stably converged, if so, stopping iteration, and ending the algorithm; otherwise, go to step 3.
CN201810038957.3A 2018-01-16 2018-01-16 Medical computed tomography image segmentation method based on active contour model Expired - Fee Related CN108416792B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201810038957.3A CN108416792B (en) 2018-01-16 2018-01-16 Medical computed tomography image segmentation method based on active contour model

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201810038957.3A CN108416792B (en) 2018-01-16 2018-01-16 Medical computed tomography image segmentation method based on active contour model

Publications (2)

Publication Number Publication Date
CN108416792A CN108416792A (en) 2018-08-17
CN108416792B true CN108416792B (en) 2021-07-06

Family

ID=63125713

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201810038957.3A Expired - Fee Related CN108416792B (en) 2018-01-16 2018-01-16 Medical computed tomography image segmentation method based on active contour model

Country Status (1)

Country Link
CN (1) CN108416792B (en)

Families Citing this family (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110335287A (en) * 2019-07-15 2019-10-15 北华航天工业学院 The extracting method and device of Architectural drawing data
CN116740768B (en) * 2023-08-11 2023-10-20 南京诺源医疗器械有限公司 Navigation visualization method, system, equipment and storage medium based on nasoscope

Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102024260A (en) * 2010-12-30 2011-04-20 浙江工商大学 Local Gamma fitting-based active contour SAR image segmentation method
CN103065309A (en) * 2012-12-26 2013-04-24 辽宁师范大学 Image segmentation method based on simplified local binary fitting (LBF) model
CN105761274A (en) * 2016-03-21 2016-07-13 辽宁师范大学 Medical image segmentation method combining edge and area information
CN106570867A (en) * 2016-10-18 2017-04-19 浙江大学 ACM (Active Contour Model) image rapid segmentation method based on gray scale morphological energy method
CN106803260A (en) * 2016-12-28 2017-06-06 辽宁师范大学 Infrared ship activity of imagination contours segmentation method based on the convex optimization of local entropy
CN106997596A (en) * 2017-04-01 2017-08-01 太原理工大学 A kind of Lung neoplasm dividing method of the LBF movable contour models based on comentropy and joint vector
CN107154047A (en) * 2017-04-24 2017-09-12 天津大学 Multi-mode brain tumor image blend dividing method and device

Family Cites Families (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
KR101090375B1 (en) * 2011-03-14 2011-12-07 동국대학교 산학협력단 Ct image auto analysis method, recordable medium and apparatus for automatically calculating quantitative assessment index of chest-wall deformity based on automized initialization

Patent Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102024260A (en) * 2010-12-30 2011-04-20 浙江工商大学 Local Gamma fitting-based active contour SAR image segmentation method
CN103065309A (en) * 2012-12-26 2013-04-24 辽宁师范大学 Image segmentation method based on simplified local binary fitting (LBF) model
CN105761274A (en) * 2016-03-21 2016-07-13 辽宁师范大学 Medical image segmentation method combining edge and area information
CN106570867A (en) * 2016-10-18 2017-04-19 浙江大学 ACM (Active Contour Model) image rapid segmentation method based on gray scale morphological energy method
CN106803260A (en) * 2016-12-28 2017-06-06 辽宁师范大学 Infrared ship activity of imagination contours segmentation method based on the convex optimization of local entropy
CN106997596A (en) * 2017-04-01 2017-08-01 太原理工大学 A kind of Lung neoplasm dividing method of the LBF movable contour models based on comentropy and joint vector
CN107154047A (en) * 2017-04-24 2017-09-12 天津大学 Multi-mode brain tumor image blend dividing method and device

Non-Patent Citations (4)

* Cited by examiner, † Cited by third party
Title
ACM-Based Automatic Liver Segmentation From 3-D CT Images by Combining Multiple Atlases and Improved Mean-Shift Techniques;Hongwei Ji等;《IEEE Journal of Biomedical and Health Informatics》;20130124;第17卷(第3期);第690-698页 *
图像分割的活动轮廓模型研究;方玲玲等;《中国博士学位论文全文数据库 信息科技辑》;20130815(第8期);第I138-190页 *
基于改进GAC模型的二值水平集前列腺超声图像自动分割算法;杨振森等;《中国科学技术大学学报》;20100515;第40卷(第5期);第441-445页 *
活动轮廓模型的图像分割方法综述;王相海等;《模式识别与人工智能》;20130815;第26卷(第8期);第751-760页 *

Also Published As

Publication number Publication date
CN108416792A (en) 2018-08-17

Similar Documents

Publication Publication Date Title
Shen et al. An automated lung segmentation approach using bidirectional chain codes to improve nodule detection accuracy
US7876938B2 (en) System and method for whole body landmark detection, segmentation and change quantification in digital images
CN109074639B (en) Image registration system and method in medical imaging system
Reeves et al. On measuring the change in size of pulmonary nodules
US7760941B2 (en) Method and apparatus of segmenting an object in a data set and of determination of the volume of segmented object
Kumar et al. Automatic liver and lesion segmentation: a primary step in diagnosis of liver diseases
Aprovitola et al. Knee bone segmentation from MRI: A classification and literature review
Gao et al. Automatic segmentation of coronary tree in CT angiography images
Schreibmann et al. Multiatlas segmentation of thoracic and abdominal anatomy with level set‐based local search
Bardera et al. Semi-automated method for brain hematoma and edema quantification using computed tomography
CN104424647A (en) Method and apparatus for registering medical images
Liu et al. Cardiac magnetic resonance image segmentation based on convolutional neural network
Hong et al. Automatic lung nodule matching on sequential CT images
Camara et al. Explicit incorporation of prior anatomical information into a nonrigid registration of thoracic and abdominal CT and 18-FDG whole-body emission PET images
US20090041322A1 (en) Computer Assisted Detection of Polyps Within Lumen Using Enhancement of Concave Area
CN108416792B (en) Medical computed tomography image segmentation method based on active contour model
Liu et al. Robust cortical thickness morphometry of neonatal brain and systematic evaluation using multi-site MRI datasets
Yin et al. Automatic breast tissue segmentation in MRIs with morphology snake and deep denoiser training via extended Stein’s unbiased risk estimator
Farag et al. Appearance models for robust segmentation of pulmonary nodules in 3D LDCT chest images
Sun et al. Automatic segmentation and measurement on knee computerized tomography images for patellar dislocation diagnosis
KR101028798B1 (en) Method for detection of hepatic tumors using registration of multi-phase liver CT images
Dhalia Sweetlin et al. Patient-Specific Model Based Segmentation of Lung Computed Tomographic Images.
Hu et al. Image manifold revealing for breast lesion segmentation in DCE-MRI
Grigorios-Aris et al. Automatic segmentation of lungs in SPECT images using active shape model trained by meshes delineated in CT images
Tan et al. A segmentation method of lung parenchyma from chest CT images based on dual U-Net

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for 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

Granted publication date: 20210706

Termination date: 20220116

CF01 Termination of patent right due to non-payment of annual fee