CN108596887B - Automatic segmentation method for liver tumor region image in abdominal CT sequence image - Google Patents

Automatic segmentation method for liver tumor region image in abdominal CT sequence image Download PDF

Info

Publication number
CN108596887B
CN108596887B CN201810341254.8A CN201810341254A CN108596887B CN 108596887 B CN108596887 B CN 108596887B CN 201810341254 A CN201810341254 A CN 201810341254A CN 108596887 B CN108596887 B CN 108596887B
Authority
CN
China
Prior art keywords
liver
image
abdominal
region
sequence
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
Application number
CN201810341254.8A
Other languages
Chinese (zh)
Other versions
CN108596887A (en
Inventor
廖苗
赵于前
刘毅志
方志雄
欧阳军林
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.)
Hunan University of Science and Technology
Original Assignee
Hunan University of Science and Technology
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 Hunan University of Science and Technology filed Critical Hunan University of Science and Technology
Priority to CN201810341254.8A priority Critical patent/CN108596887B/en
Publication of CN108596887A publication Critical patent/CN108596887A/en
Application granted granted Critical
Publication of CN108596887B publication Critical patent/CN108596887B/en
Active 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
    • G06T7/0002—Inspection of images, e.g. flaw detection
    • G06T7/0012—Biomedical image inspection
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T5/00—Image enhancement or restoration
    • G06T5/20—Image enhancement or restoration using local operators
    • G06T5/30—Erosion or dilatation, e.g. thinning
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T5/00—Image enhancement or restoration
    • G06T5/40—Image enhancement or restoration using histogram techniques
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T5/00—Image enhancement or restoration
    • G06T5/70—Denoising; Smoothing
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/10—Segmentation; Edge detection
    • G06T7/11—Region-based segmentation
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/10—Segmentation; Edge detection
    • G06T7/194—Segmentation; Edge detection involving foreground-background segmentation
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/10—Image acquisition modality
    • G06T2207/10072—Tomographic images
    • G06T2207/10081—Computed x-ray tomography [CT]
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/30—Subject of image; Context of image processing
    • G06T2207/30004—Biomedical image processing
    • G06T2207/30056—Liver; Hepatic
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/30—Subject of image; Context of image processing
    • G06T2207/30004—Biomedical image processing
    • G06T2207/30096—Tumor; Lesion

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • General Health & Medical Sciences (AREA)
  • Health & Medical Sciences (AREA)
  • Medical Informatics (AREA)
  • Nuclear Medicine, Radiotherapy & Molecular Imaging (AREA)
  • Radiology & Medical Imaging (AREA)
  • Quality & Reliability (AREA)
  • Apparatus For Radiation Diagnosis (AREA)
  • Image Processing (AREA)

Abstract

The invention discloses an automatic segmentation method of liver tumor region images in abdominal CT sequence images. The method comprises the following steps: preprocessing, namely preprocessing the abdominal CT sequence image to acquire a liver region in the abdominal CT sequence image; a liver enhancement step, namely adopting piecewise nonlinear enhancement and iterative convolution operation to improve the contrast of the normal liver parenchyma and the tumor tissue according to the gray level distribution characteristic of the liver region; an automatic segmentation step, namely constructing a multi-target segmented graph cutting energy function by utilizing the enhanced result and combining image boundary information, and minimizing the energy function by adopting an optimization algorithm to obtain a primary automatic segmentation result of the liver tumor; and a post-processing step, namely optimizing the primary segmentation result by adopting three-dimensional mathematical morphology opening operation, removing a mistaken segmentation area in the initial segmentation result, and improving the segmentation precision. The invention is helpful for radiologists and surgeons to effectively acquire the whole information and three-dimensional display of liver tumors in time, and provides technical support for computer-aided diagnosis and treatment of liver diseases.

Description

Automatic segmentation method for liver tumor region image in abdominal CT sequence image
Technical Field
The invention belongs to the technical field of image processing, relates to target segmentation in medical images, in particular to automatic segmentation of liver tumor region images in abdominal CT sequence images, and can be used for medical image auxiliary diagnosis and treatment.
Background
More than 50% of new onset and death liver cancer patients occur in China worldwide, and about 30 ten thousand people die of liver cancer in China every year. Because the symptoms of early liver cancer are not obvious, about 60 percent of patients do not see a doctor until the body is untimely, and the patients often enter the middle and late stages, so that the chance of radical treatment is lost. Statistical data show that the survival rate of the liver cancer patients in the late stage is only about 7 percent within 5 years.
Liver tumor burden analysis is commonly used to monitor disease progression in liver cancer patients, develop treatment regimens, make comparisons between different treatment regimens, predict and assess treatment efficacy, assess the effectiveness of anti-cancer drugs, and the like. Segmentation of liver tumor tissues in Computed Tomography (CT) sequence images is an important premise for liver tumor burden analysis and also an important basis for computer-aided diagnosis of liver diseases and formulation of surgical schemes. According to the segmentation result of liver tumor tissue of abdominal CT sequence image, the clinician can obtain the information of number, size, shape, position, lesion degree, infiltration depth, metastasis and the like of the lesion, diagnose the disease and formulate a proper treatment scheme. Because of the large number of image slices used in CT imaging (if the layer thickness is 1.5mm, about 120 slices are available for an abdominal CT sequence completely including the liver of a patient), the manual segmentation of each slice is labor-intensive and time-consuming, and the accuracy and validity of the segmentation result depend heavily on the experience, skill and subjective judgment of the radiologist. Therefore, developing and designing an automatic robust segmentation method for liver tumor of abdominal CT sequence image has important significance for improving the accuracy and efficiency of liver tumor load analysis and computer-aided diagnosis.
Disclosure of Invention
The invention aims to provide an automatic segmentation method of a liver tumor region image in an abdominal CT sequence image, which aims to solve the problem of inaccurate automatic segmentation of the liver tumor region image caused by fuzzy liver tumor boundary, low contrast with normal tissues, complex structure, various gray levels and the like in a CT image and improve the precision and efficiency of computer-aided diagnosis.
A method for automatically segmenting liver tumor region images in abdominal CT sequence images comprises the following steps: (1) preprocessing an abdomen CT sequence image f by adopting a sparse shape combination model to obtain the liver in the abdomen CT sequence image f
An area;
(2) fitting the gray level histogram of the whole liver region in the sequence by adopting a Gaussian function according to the probability of Gaussian distributionTheoretical and anatomical prior knowledge to obtain the approximate gray scale range of normal liver tissue [ Imin,Imax]In which IminRepresenting the minimum value of the gray scale, ImaxExpressing a maximum value of the gray scale;
(3) using grey value IminAnd ImaxThe liver region is subjected to piecewise nonlinear enhancement, the contrast between the tumor and the normal liver tissue is improved, and the enhanced result is recorded as zeta.
(4) To remove noise and smooth the image in the enhancement result, N is performed by checking the enhancement result ζ with a convolution of (2s +1) × (2s +1)iterPerforming sub-iteration convolution operation to obtain result zetaconvWherein s, NiterAll are natural numbers larger than 0, preferably s is a natural number of 1-5, and N isiterIs a natural number of 30 to 130;
(5) and constructing a multi-target segmented graph cut energy function by utilizing the enhancement result and combining image boundary information:
Figure GDA0002176841850000021
where P represents the set of all pixels in the image f, fpAnd fqRespectively representing pixel points p and q, N in an image fpNeighborhood set of pixels, R (f), representing pixel point pp) And B (f)p,fq) respectively a gray level penalty term and a boundary penalty term, respectively obtained by the enhancement result and the image gradient calculation, and respectively used for the label distribution of the background, the normal liver parenchyma and the tumor tissue in the image segmentation algorithm and the smoothness control of the boundary of the segmentation region, and the weight α is used for adjusting the gray level penalty term R (f)p) And a boundary penalty term B (f)p,fq) the proportion occupied in the graph cutting algorithm is in a value range of 0-1, and α is preferably a normal number of 0.5-1;
(6) minimizing an energy function E (f) by adopting an optimization algorithm to obtain a primary liver tumor segmentation result;
(7) and (4) performing post-processing on the primary segmentation result by adopting three-dimensional mathematical morphology opening operation, and removing the mistakenly segmented region to obtain accurate liver tumor tissue.
In the step (3), the piecewise nonlinear enhancement formula is as follows:
Figure GDA0002176841850000022
wherein I is the image pixel gray scale,
Figure GDA0002176841850000023
and theta is a penalty factor for regulating the enhancement degree of the normal liver region and the tumor region respectively,
Figure GDA0002176841850000024
the sum theta is a normal number, and when the gray level I of the image pixel falls in the interval [ Imin,Imax]In the above process, the probability that the pixel belongs to the normal liver is high, and a penalty factor for enhancing the pixel is obtained
Figure GDA0002176841850000031
The setting is relatively small, and when the gray I is less than IminOr greater than ImaxThe probability that the pixel point belongs to the liver tumor is high, the punishment factor theta value is set to be relatively high, and the preference is given
Figure GDA0002176841850000032
Is a normal number of 0.1 to 1, and theta is a normal number of 1 to 3.
In the step (5), the gray level penalty term R (f) of the energy function is mapped and cutp) The method relates to gray punishment that image pixel points belong to background, normal liver parenchyma and tumor tissues respectively, and comprises the following specific calculation formula:
Figure GDA0002176841850000033
wherein f ispAnd fqRespectively representing pixel points p and q, I in an image fpAnd IqRepresenting the gray values of pixels p and q, mask being a liver mask obtained by preprocessing an abdomen CT sequence image using the prior art, pixel markers belonging to the liver regionIs 1, the pixels belonging to the background are marked as 0, i.e.
Figure GDA0002176841850000034
Boundary penalty term B (f)p,fq) Punishment is carried out on the inconsistency of the gray levels of the adjacent pixels, and the calculation formula is as follows:
Figure GDA0002176841850000035
wherein
Figure GDA0002176841850000036
Figure GDA0002176841850000037
d (p, q) represents the Euclidean distance of pixels p and q, TPRepresenting the total number of pixels of the set P of pixels of the image f.
In the step (7), a spherical structure with the radius of r is preferably used as a structural element for three-dimensional morphological open operation, wherein r is preferably a natural number of 2-25.
Drawings
FIG. 1 4 randomly selected original images in a certain abdomen CT sequence
FIG. 2 illustrates an example of liver region segmentation result according to an embodiment of the present invention
FIG. 3 is a Gaussian fitting result example of a liver region gray histogram according to an embodiment of the present invention
FIG. 4 convolution kernel of an embodiment of the present invention
FIG. 5 exemplary liver region enhancement results for an embodiment of the invention
FIG. 6 is a diagram of a basic principle of a multi-objective segmentation-based graph cutting algorithm according to an embodiment of the present invention
FIG. 7 is a background region, normal liver parenchyma and liver tumor classification result example according to an embodiment of the present invention
FIG. 8 two-dimensional display of liver tumor segmentation results according to embodiments of the present invention
FIG. 9 three-dimensional display of liver tumor segmentation results according to embodiments of the present invention
Detailed Description
Example 1
In order to acquire the liver region in the abdomen CT sequence image, the abdomen CT sequence liver automatic segmentation method disclosed in the document "a longitudinal local region-based sparse shape composition for liver segmentation CT scans" (pattern recognition, pp.88-106,2016) is adopted to pre-process the original CT sequence image, and the liver region in the sequence is acquired. Fig. 1(a) - (d) show 4 original images randomly selected from a CT sequence, and fig. 2(a) - (d) show liver segmentation results, i.e. liver region masks, obtained by the method of this embodiment.
Example 2
A liver region enhancement method for an abdominal CT sequence image comprises the following specific implementation steps:
(1) the liver region in the abdominal CT sequence image f was acquired using example 1;
(2) in order to obtain the gray distribution range of the liver region, a Gaussian function is adopted to fit the gray histogram of the whole liver region in the sequence:
Figure GDA0002176841850000041
where c is the peak of the gaussian distribution and μ and σ are the center and standard deviation of the gaussian distribution, respectively. Fig. 3 is a result of performing gaussian fitting on the gray level histogram of the liver region in the sequence image shown in fig. 1, and it can be seen that the liver gray level can better conform to gaussian distribution. According to the probability theory of Gaussian distribution, [ mu-sigma, [ mu + sigma ]]、[μ-2σ,μ+2σ]And [ mu-3 sigma, [ mu +3 sigma ]]The gray scale ranges of (1) account for 68%, 95%, 99% of the pixels in the liver region, respectively. Considering noise and tumor tissue that may occur in the liver region, the present embodiment preferably has the gray scale minimum and maximum values of normal liver parenchyma as Iminμ -0.8 σ and Imax＝μ+0.8σ。
(3) Using grey value IminAnd ImaxPerforming piecewise nonlinear enhancement on liver region to improve tumor and normal liver tissueThe contrast of (2). The piecewise nonlinear enhancement formula is as follows:
Figure GDA0002176841850000051
wherein I is the image pixel gray scale,
Figure GDA0002176841850000052
and theta is a penalty factor for regulating the enhancement degree of the normal liver region and the tumor region respectively,
Figure GDA0002176841850000053
the sum theta is a normal number, and when the gray level I of the image pixel falls in the interval [ Imin,Imax]In the above process, the probability that the pixel belongs to the normal liver is high, and a penalty factor for enhancing the pixel is obtained
Figure GDA0002176841850000054
The setting is relatively small, and when the gray I is less than IminOr greater than ImaxIn the time, the probability that the pixel belongs to the liver tumor is high, the punishment factor theta value is set to be relatively high, and the optimization is preferably performed in the embodiment
Figure GDA0002176841850000055
θ＝2。
(4) Carrying out iterative convolution operation on the enhanced result zeta to obtain a result zetaconvThe preferred size of the convolution kernel in this embodiment is 3 x 3 as shown in fig. 4, and the preferred number of iterations is 60. The convolution operation can effectively remove noise and smooth the image, and meanwhile, the image boundary information is kept.
FIGS. 5(a) - (d) show the results of the enhancement of the liver region of FIGS. 1(a) - (d) using this example, showing that the contrast between normal liver parenchyma and tumor tissue is significantly improved.
Example 3
The embodiment 2 is adopted to obtain the liver enhancement result, and the image boundary information is combined to construct the multi-target segmentation image segmentation energy function:
Figure GDA0002176841850000056
wherein α is a normal number of 0-1, P represents all pixel sets in the abdomen CT sequence image f, and N ispNeighborhood set of pixels, R (f), representing pixel point pp) And B (f)p,fq) Respectively are gray level punishment items and boundary punishment items, and are respectively obtained by adopting the following formulas:
Figure GDA0002176841850000057
Figure GDA0002176841850000058
wherein
Figure GDA0002176841850000061
Figure GDA0002176841850000062
fpAnd fqRespectively representing pixel points p and q, I in an image fpAnd IqRepresenting the gray values of pixels p and q, d (p, q) representing the Euclidean distance of pixels p and q, TPThe total number of pixels in the set of pixels P of the image f is represented by the liver mask obtained in example 1, the pixels belonging to the liver region are labeled as 1, and the pixels belonging to the background are labeled as 0, i.e., the total number of pixels in the set of pixels P is represented by
Figure GDA0002176841850000063
FIG. 6 is a diagram of a basic principle of a graph cut algorithm of multi-target segmentation. In the multi-target segmentation-based graph cut algorithm, a gray level penalty term R (f)p) Label assignment for background, normal liver parenchyma and tumor tissue, corresponding to t-junctions in fig. 6, when the probability that a pixel belongs to a certain class is higher, the penalty for it is smaller, the corresponding t-junction value will be larger, i.e. the corresponding edge in the undirected graph is thicker, and vice versa. Boundary penalty term B (f)p,fq) the smoothness control for the segmentation region boundary, corresponding to the n-junction in fig. 6, will be larger for the smaller penalty as the gray scale between adjacent pixels is closer, i.e. the thicker the corresponding edge in the undirected graph, and vice versap) And a boundary penalty term B (f)p,fq) the proportion occupied in the graph cut algorithm is in a value range of 0-1, in the embodiment, alpha is preferably 0.6, the maximum flow minimum cut algorithm is adopted to minimize the energy function E (f), the CT image can be divided into three types of background, normal liver tissue and tumor, as shown in fig. 7(a) - (d), and the type of the tumor is extracted, so that the liver tumor preliminary segmentation result can be obtained.
Example 4
After the preliminary segmentation result of the liver tumor is obtained in embodiment 3, a three-dimensional morphological opening operation is performed on the preliminary segmentation result to remove noise and erroneous segmentation regions that may occur therein, so as to obtain a final segmentation result of the liver tumor, and a spherical structure with a radius of 8 is preferably used as a structural element of the morphological opening operation in this embodiment. Fig. 8(a) - (d) are two-dimensional displays of the liver tumor segmentation results obtained by the method of the present embodiment, wherein the tumor regions are all segmented completely and effectively. Fig. 9 is a three-dimensional display of liver tumor segmentation results, and it can be seen that the method of the present invention can effectively segment liver tumors with different sizes and shapes in abdominal CT sequence images.

Claims (6)

1. A method for automatically segmenting an image of a liver tumor region in an abdominal CT sequence image is characterized by comprising the following steps:
(1) preprocessing an abdomen CT sequence image f by adopting a sparse shape combination model to obtain a liver region;
(2) adopting a Gaussian function to fit a gray level histogram of the whole liver region in the sequence, and obtaining an approximate gray level range [ I ] of normal liver tissues according to probability theory of Gaussian distribution and prior knowledge of anatomymin,Imax]In which IminRepresenting the minimum value of the gray scale, ImaxExpressing a maximum value of the gray scale;
(3) using grey value IminAnd ImaxCarrying out piecewise nonlinear enhancement on the liver region, improving the contrast of the tumor and the normal liver parenchyma, and recording an enhancement result as zeta;
(4) to remove noise and smooth the image in the enhancement result, N is performed by checking the enhancement result ζ with a convolution of (2s +1) × (2s +1)iterPerforming sub-iteration convolution operation to obtain result zetaconvWherein s, NiterAre all natural numbers greater than 0;
(5) and constructing a multi-target segmented graph cut energy function by utilizing the enhancement result and combining image boundary information:
Figure FDA0002409012100000011
where P represents the set of all pixels in image f; f. ofpAnd fqRespectively representing pixel points p and q in the image f; n is a radical ofpA neighborhood pixel set representing a pixel point p; r (f)p) Punishment is carried out on image pixel points belonging to the background, the normal liver parenchyma and the tumor tissue respectively for a gray punishment item, and the calculation formula is as follows:
Figure FDA0002409012100000012
wherein, IpAnd IqRepresenting the gray values of the pixels p and q, wherein the mask is a liver mask obtained by preprocessing an abdomen CT sequence image by adopting a sparse shape combination model, the pixel belonging to the liver region is marked as 1, the pixel belonging to the background is marked as 0, namely
Figure FDA0002409012100000013
B(fp,fq) And punishing the inconsistency of the gray levels of the adjacent pixels as a boundary punishment item, wherein the calculation formula is as follows:
Figure FDA0002409012100000021
wherein
Figure FDA0002409012100000022
Figure FDA0002409012100000023
d (p, q) represents the Euclidean distance of pixels p and q, TPRepresents the total number of pixels of the set of pixels P of the image f;
the weight α is used for adjusting the gray penalty term R (f)p) And a boundary penalty term B (f)p,fq) The proportion occupied in the graph cutting algorithm is in a value range of 0-1;
(6) minimizing an energy function E (f) by adopting an optimization algorithm to obtain a primary liver tumor segmentation result;
(7) and (4) performing post-processing on the primary segmentation result by adopting three-dimensional mathematical morphology opening operation, and removing the mistakenly segmented region to obtain accurate liver tumor tissue.
2. The method for automatic segmentation of images of liver tumor regions in abdominal CT sequence images as set forth in claim 1, wherein: s is preferably a natural number of 1-5, Niterpreferably a natural number of 30 to 130, and alpha is preferably a normal number of 0.5 to 1.
3. The method for automatic segmentation of images of liver tumor regions in abdominal CT sequence images as set forth in claim 1, wherein: in the step (3), the piecewise nonlinear enhancement formula is as follows:
Figure FDA0002409012100000024
wherein I is the image pixel gray scale,
Figure FDA0002409012100000025
and theta is a penalty factor for regulating the enhancement degree of the normal liver region and the tumor region respectively,
Figure FDA0002409012100000026
the sum theta is a normal number, and when the gray level I of the image pixel falls in the interval [ Imin,Imax]In the above process, the probability that the pixel belongs to the normal liver is high, and a penalty factor for enhancing the pixel is obtained
Figure FDA0002409012100000027
The setting is relatively small, and when the gray I is less than IminOr greater than ImaxAnd then, the probability that the pixel belongs to the liver tumor is high, and the punishment factor theta value is set to be relatively high.
4. A method for automatic segmentation of images of liver tumor regions in abdominal CT series images as claimed in claim 3, characterized by: the above-mentioned
Figure FDA0002409012100000031
The number of theta is preferably 0.1 to 1, and the number of theta is preferably 1 to 3.
5. The method for automatic segmentation of images of liver tumor regions in abdominal CT sequence images as set forth in claim 1, wherein: in the above-mentioned step (7), a spherical structure having a radius r is preferable as a structural element for three-dimensional morphological opening operation.
6. The method of automatic segmentation of images of liver tumor regions in abdominal CT sequence images as set forth in claim 5, wherein: the r is preferably a natural number of 2-25.
CN201810341254.8A 2018-04-17 2018-04-17 Automatic segmentation method for liver tumor region image in abdominal CT sequence image Active CN108596887B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201810341254.8A CN108596887B (en) 2018-04-17 2018-04-17 Automatic segmentation method for liver tumor region image in abdominal CT sequence image

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201810341254.8A CN108596887B (en) 2018-04-17 2018-04-17 Automatic segmentation method for liver tumor region image in abdominal CT sequence image

Publications (2)

Publication Number Publication Date
CN108596887A CN108596887A (en) 2018-09-28
CN108596887B true CN108596887B (en) 2020-06-02

Family

ID=63622730

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201810341254.8A Active CN108596887B (en) 2018-04-17 2018-04-17 Automatic segmentation method for liver tumor region image in abdominal CT sequence image

Country Status (1)

Country Link
CN (1) CN108596887B (en)

Families Citing this family (18)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN109493351A (en) * 2018-11-12 2019-03-19 哈尔滨理工大学 The system that liver segmentation is carried out using probability map and level set to CT image
CN109753997B (en) * 2018-12-19 2022-11-22 湖南科技大学 Automatic accurate robust segmentation method for liver tumor in CT image
CN109741359B (en) * 2019-01-13 2022-05-31 中南大学 Method for segmenting lesion liver of abdominal CT sequence image
CN109934235B (en) * 2019-03-20 2021-04-20 中南大学 Unsupervised abdominal CT sequence image multi-organ simultaneous automatic segmentation method
CN109961449B (en) * 2019-04-15 2023-06-02 上海电气集团股份有限公司 Image segmentation method and device, and three-dimensional image reconstruction method and system
CN110223289A (en) * 2019-06-17 2019-09-10 上海联影医疗科技有限公司 A kind of image processing method and system
CN110276407A (en) * 2019-06-26 2019-09-24 哈尔滨理工大学 A kind of Hepatic CT staging system and classification method
CN110473196B (en) * 2019-08-14 2021-06-04 中南大学 Abdomen CT image target organ registration method based on deep learning
CN110610491B (en) * 2019-09-17 2021-11-19 湖南科技大学 Liver tumor region segmentation method of abdominal CT image
CN111047567A (en) * 2019-12-05 2020-04-21 电子科技大学 Kidney tumor picture determination method and related device
CN111161256A (en) * 2019-12-31 2020-05-15 北京推想科技有限公司 Image segmentation method, image segmentation device, storage medium, and electronic apparatus
CN111489434A (en) * 2020-03-18 2020-08-04 创业慧康科技股份有限公司 Medical image three-dimensional reconstruction method based on three-dimensional graph cut
CN111476881A (en) * 2020-03-18 2020-07-31 创业慧康科技股份有限公司 Human tissue organ three-dimensional reconstruction method based on structural similarity level set algorithm
CN112308082B (en) * 2020-11-05 2023-04-07 湖南科技大学 Dynamic video image segmentation method based on dual-channel convolution kernel and multi-frame feature fusion
CN112734790B (en) * 2020-12-30 2023-07-11 武汉联影生命科学仪器有限公司 Tumor region labeling method, system, device and readable storage medium
CN114820663B (en) * 2022-06-28 2022-09-09 日照天一生物医疗科技有限公司 Assistant positioning method for determining radio frequency ablation therapy
CN115294122B (en) * 2022-10-08 2023-04-07 江苏诺阳家居科技有限公司 Lung tumor feature extraction method and system based on CT image
CN116993764B (en) * 2023-09-26 2023-12-08 江南大学附属医院 Stomach CT intelligent segmentation extraction method

Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102385751A (en) * 2011-07-19 2012-03-21 中国科学院自动化研究所 Liver tumor region segmentation method based on watershed transform and classification through support vector machine
CN105139377A (en) * 2015-07-24 2015-12-09 中南大学 Rapid robustness auto-partitioning method for abdomen computed tomography (CT) sequence image of liver

Family Cites Families (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US9940711B2 (en) * 2015-11-25 2018-04-10 Zebra Medical Vision Ltd. Systems and methods for detecting a fatty liver from a computed tomography (CT) scan

Patent Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102385751A (en) * 2011-07-19 2012-03-21 中国科学院自动化研究所 Liver tumor region segmentation method based on watershed transform and classification through support vector machine
CN105139377A (en) * 2015-07-24 2015-12-09 中南大学 Rapid robustness auto-partitioning method for abdomen computed tomography (CT) sequence image of liver

Also Published As

Publication number Publication date
CN108596887A (en) 2018-09-28

Similar Documents

Publication Publication Date Title
CN108596887B (en) Automatic segmentation method for liver tumor region image in abdominal CT sequence image
EP3432784B1 (en) Deep-learning-based cancer classification using a hierarchical classification framework
Wu et al. Automatic liver segmentation on volumetric CT images using supervoxel‐based graph cuts
Haas et al. Automatic segmentation of thoracic and pelvic CT images for radiotherapy planning using implicit anatomic knowledge and organ-specific segmentation strategies
Elizabeth et al. Computer-aided diagnosis of lung cancer based on analysis of the significant slice of chest computed tomography image
CN109753997B (en) Automatic accurate robust segmentation method for liver tumor in CT image
Abd-Elaziz et al. Liver tumors segmentation from abdominal CT images using region growing and morphological processing
CN106846317B (en) Medical image retrieval method based on feature extraction and similarity matching
Akram et al. Artificial neural network based classification of lungs nodule using hybrid features from computerized tomographic images
Pulagam et al. Automated lung segmentation from HRCT scans with diffuse parenchymal lung diseases
CN106056596B (en) Full-automatic three-dimensional liver segmentation method based on local prior information and convex optimization
Chen et al. Pathological lung segmentation in chest CT images based on improved random walker
CN112767407A (en) CT image kidney tumor segmentation method based on cascade gating 3DUnet model
Tseng et al. An adaptive thresholding method for automatic lung segmentation in CT images
Lim et al. Segmentation of the liver using the deformable contour method on CT images
Hamad et al. Segmentation and measurement of lung pathological changes for COVID-19 diagnosis based on computed tomography
Ali et al. Diagnosis of liver tumor from CT images using digital image processing
Jaffery et al. Performance analysis of image segmentation methods for the detection of masses in mammograms
Farag et al. Variational approach for segmentation of lung nodules
Jalab et al. Fractional Renyi entropy image enhancement for deep segmentation of kidney MRI
CN110533667B (en) Lung tumor CT image 3D segmentation method based on image pyramid fusion
Likhitkar et al. Automated detection of cancerous lung nodule from the computed tomography images
Ali et al. Comparative analysis of learning algorithms for lung cancer identification
Mini et al. A neural network method for mammogram analysis based on statistical features
CN111145353B (en) Method for generating 3D point cloud through image segmentation and grid feature point extraction algorithm

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