CN110136124A - A kind of pleura contact Lung neoplasm dividing method based on Robust Speed function - Google Patents
A kind of pleura contact Lung neoplasm dividing method based on Robust Speed function Download PDFInfo
- Publication number
- CN110136124A CN110136124A CN201910412792.6A CN201910412792A CN110136124A CN 110136124 A CN110136124 A CN 110136124A CN 201910412792 A CN201910412792 A CN 201910412792A CN 110136124 A CN110136124 A CN 110136124A
- Authority
- CN
- China
- Prior art keywords
- speed function
- lung neoplasm
- lung
- robust
- robust speed
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
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
- 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
- G06T7/149—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
- 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
- G06T2207/30061—Lung
- G06T2207/30064—Lung nodule
Abstract
The present invention provides a kind of pleura contact Lung neoplasm dividing method based on Robust Speed function, the following steps are included: S1, will by simply dealt lung CT image combination Gray-level co-occurrence and feature F-KNN sorting algorithm, calculating Robust Speed function in probability score；Robust Speed function in S2, tectonic activity skeleton pattern；S3, calculate movable contour model energy functional and carried out minimum calculation processing；S4, evaluation is compared by three indexs to segmentation precision, the present invention, which uses, combines Gray-level co-occurrence and the fuzzy F-KNN algorithm model of local binary patterns, that is, LBP feature, calculate the probability score in Robust Speed function, for reinforcing the differentiation of Lung neoplasm and lung wall and its ambient background, then Robust Speed function is introduced into movable contour model, so that Robust Speed function approaches zero in the boundary of Lung neoplasm, active contour curve stops developing, to improve the segmentation precision of lung wall adhesive type Lung neoplasm.
Description
Technical field
The invention belongs to Lung neoplasm cutting method technical fields, and in particular to a kind of pleura based on Robust Speed function connects
Touching property Lung neoplasm dividing method.
Background technique
Lung cancer is one of highest malignant tumour of morbidity and mortality in worldwide, and the early stage of lung cancer is with Lung neoplasm shape
Formula performance.When Lung neoplasm and lung wall stick together, cancer cell is easily invaded to pleura, once there is a situation where invading,
Cancer cell can be transferred to its hetero-organization of human body by pleura, and patient is made to be more difficult to cure, and the Accurate Segmentation of Lung neoplasm is lung
The important preprocessing step of nodule detection.
The non-reality Lung neoplasm of lung wall adhesive type and ambient background have low contrast, lung wall adhesive type reality Lung neoplasm and lung
Wall has similar gray value, and all has the characteristics that obscurity boundary and Lung neoplasm internal brightness are non-uniform, to the essence of Lung neoplasm
Really segmentation brings great challenge.
In recent years, lot of domestic and foreign scholar has been proposed largely dividing Lung neoplasm method, is based on movable contour model
The dividing method of (Active Contour Models, ACM) divides the image into problem and is converted into solution and three-dimensional level set function
Related partial differential equation numerical problem, and variation that can obtain the curve being smoothly closed and realization curve topological structure etc. is excellent
Point has attracted the extensive research of domestic and foreign scholars.Awad etc. has chosen initial surface using Otus multi-threshold method, then uses
Level set based on sparse face is split Lung neoplasm.The combined datas such as Farhangi information and prior shape information drive
Dynamic level set movements, realize the segmentation to various types Lung neoplasm.Farag etc. proposes a kind of using implicit space as symbol
Distance function adaptively divides the Level Set Method of Lung neoplasm.Li et al. using based on probability density function similarity distance and
The adaptive local movable contour model of multiple features dynamic clustering is split Lung neoplasm.Also by movable contour model and mould
Paste clustering algorithm, which combines, to be proposed a kind of movable contour model based on fuzzy velocity function and divides ground glass shadow Lung neoplasm
It cuts.
Movable contour model of some propositions based on small echo indicates the brightness of image, using wavelet energy to reinforce lung
The differentiation of tubercle and ambient background tissue.Keshani etc. rotates Lung neoplasm image and converts stand alone for non-orphaned type Lung neoplasm
Lung neoplasm recycles movable contour model to be split.Using bezier surface by the smooth lung skeleton pattern through affine transformation
It is fitted on target lung, then be split with movable contour model.
But have between Lung neoplasm and lung wall in cutting procedure due to the boundary in lung wall adhesion reality Lung neoplasm
Similar contrast has lower comparison in the boundary of the non-reality Lung neoplasm of lung wall adhesive type between Lung neoplasm and background
Degree, and Lung neoplasm boundary has the characteristics that obscurity boundary and brightness irregularities, so that the boundary based on gradient information driving
Movable contour model can not accurately identify boundary, the regional activity profile die based on characteristic statistics property driving inside and outside contour curve
The energy functional of type can not reach minimum value, influence the segmentation precision of contact Lung neoplasm.
Summary of the invention
It is a kind of based on robust speed technical problem to be solved by the present invention lies in view of the above shortcomings of the prior art, providing
The pleura contact Lung neoplasm dividing method of function is spent, to solve to have between Lung neoplasm mentioned above in the background art and background
There is lower contrast, and Lung neoplasm boundary has the characteristics that obscurity boundary and brightness irregularities, so that believing based on gradient
The boundary movable contour model of breath driving can not accurately identify boundary, the poor problem of segmentation precision.
In order to solve the above technical problems, the technical solution adopted by the present invention is that: a kind of pleura based on Robust Speed function
Contact Lung neoplasm dividing method, comprising the following steps:
S1, the F-KNN sorting algorithm that will pass through simply dealt lung CT image combination Gray-level co-occurrence and feature, calculate
Probability score in Robust Speed function；
Robust Speed function in S2, tectonic activity skeleton pattern；
S3, calculate movable contour model energy functional and carried out minimum calculation processing；
S4, evaluation is compared by three indexs to segmentation precision.
Preferably, it is specially first to combine wavelet energy and latent structure eigenmatrix in S1, is then classified using F-KNN
The probability score matrix of algorithm construction Lung neoplasm image.
Preferably, three indexs in S4 are specially three positive rate, false positive rate and similarity indexs, and comparison is
It is compared with doctor's manual segmentation result.
Compared with the prior art, the present invention has the following advantages:
The present invention, which uses, combines Gray-level co-occurrence and the fuzzy F-KNN algorithm model of local binary patterns, that is, LBP feature,
The probability score in Robust Speed function is calculated, for reinforcing the differentiation of Lung neoplasm and lung wall and its ambient background, then by Shandong
Rod speed function is introduced into movable contour model, so that Robust Speed function approaches zero in the boundary of Lung neoplasm, castor
Wide curve stops developing, to complete the accurate segmentation of Lung neoplasm in the case of lung wall adhesion, improves lung wall adhesive type Lung neoplasm
Segmentation precision.
Detailed description of the invention
Fig. 1 is flow chart of steps of the present invention.
Specific embodiment
Following will be combined with the drawings in the embodiments of the present invention, and technical solution in the embodiment of the present invention carries out clear, complete
Site preparation description, it is clear that described embodiments are only a part of the embodiments of the present invention, instead of all the embodiments.It is based on
Embodiment in the present invention, it is obtained by those of ordinary skill in the art without making creative efforts every other
Embodiment shall fall within the protection scope of the present invention.
As shown in Figure 1, the present invention provides a kind of technical solution: a kind of pleura contact lung knot based on Robust Speed function
Section dividing method, including the following steps are included:
S1, the F-KNN sorting algorithm that will pass through simply dealt lung CT image combination Gray-level co-occurrence and feature, calculate
Probability score in Robust Speed function；
Since Lung neoplasm has the characteristics that obscurity boundary and internal brightness is non-uniform, and the part of wavelet energy characterization image
Feature, and there is multiple dimensioned characteristic and empty frequency domain characteristic, the differentiation of Lung neoplasm and ambient background can be enhanced using wavelet energy,
Define pixel (x, y) wavelet energy be
W (x, y)=EA(x,y)+EH(x,y)+EV(x,y)+ED(x,y)
Wherein EA(x,y)、EH(x,y)、EV(x, y) and ED(x, y) is the approximate detail wavelet energy of image, level respectively
Detail wavelet energy, vertical detail wavelet energy and diagonal detail wavelet energy, are defined as follows:
Wherein DA(x, y) is low-frequency wavelet coefficients, DH(x,y)、DV(x, y) and DD(x, y) is high-frequency wavelet coefficient, and Z is
The set of integer, r and s are the abscissa and ordinate of neighborhood territory pixel respectively, and K () is gaussian kernel function.
LBP is a kind of characterization image Local textural feature, is able to reflect the correlation with neighborhood territory pixel point, is that Lung neoplasm is known
Other validity feature, the LBP characterizing definition of pixel (x, y) are as follows:
Wherein, LBP (x, y)P,RIt indicates to choose P sampling in the round field pixel space that pixel (x, y) radius is R
Point carries out LBP calculating, gcFor the pixel value of pixel (x, y), gpFor neighborhood territory pixel value, s (h) indicates a two-valued function, definition
It is as follows:
In order to obtain the not LBP feature by Effect of Rotation, definition:
Wherein, ROR (x, i) is that logarithm x is moved i times by turn backward around circle, and x is formed by P.
In conjunction with Gray-level co-occurrence and LBP feature construction feature vector, X, then by Gray-level co-occurrence and LBP feature knot
Composite character matrix calculates the probability score of pixel, using F-KNN algorithm to reinforce the differentiation of Lung neoplasm and ambient background.
F-KNN is to determine sample category to be sorted by the class degree of membership and distance weighting of k neighbours of sample to be sorted
In the degree of membership of every one kind, its output is a possibility that sample belongs to each classification.X is calculated first with feature vector, Xtrain
Euclidean distance between all training samples searches for an immediate sample, after obtaining nearest-neighbors, in training set
Categorical data be blurred after obtain classification degree of membership:
Wherein, nlIt indicatesBelong to neighbours' number of l class.
It then calculates and calculates X using Euclidean distancetestThe distance between test sample obtains k nearest neighbours；Most
Whole then classification degree of membership and distance weighting by this k neighbour determines the class degree of membership of test sample, that is, belongs to lung knot
The probability score of section.
Wherein, l indicates classification,Indicate sample X to be sorted(x,y)J-th of neighbour, m be fuzzy weighted values regulatory factor,
According to ul(X(x,y)) the probability score u (x, y) that pixel (x, y) belongs to Lung neoplasm can be obtained.
Robust Speed function in S2, tectonic activity skeleton pattern；
In ideal conditions, the probability score of Lung neoplasm is greater than 0.5, is approximately equal in Lung neoplasm boundary probability score
0.5, and the probability score of blood vessel or background is less than 0.5.In order to enable contour curve stops developing in Lung neoplasm boundary, profile
The Robust Speed function V that curve inwardly develops1The Robust Speed function V that (x, y), contour curve develop outward2(x, y) and reflection
The Robust Speed function V of profile and border3(x, y) must satisfy:
(1) in Lung neoplasm boundary, Robust Speed function is approximately equal to zero；
(2) far from Lung neoplasm boundary, Robust Speed function is increasing；
(3) in approach Lung neoplasm boundary, Robust Speed function is smaller and smaller.
The V in model proposed in this paper1(x,y)、V2(x, y) and V3(x, y) is respectively defined as:
Wherein, t1, t2And t3For normal number.
S3, calculate movable contour model energy functional and carried out minimum calculation processing；
In movable contour model, usually assumes that C is a closed contour curve, image-region Ω can be divided
Are as follows: curvilinear inner region Ω1With curved exterior region Ω2.It is inwardly drilled according to contour curve and becomes blurred velocity function V1And profile
Curve is drilled outward becomes blurred velocity function V2It can define the area item of movable contour model
E1(φ)=∫ ∫Ω[V1(x,y)|I(x,y)-c1|2H(φ)+V2(x,y)|I(x,y)-c2|2(1-H(φ))]dxdy
Wherein φ is level set function, and I (x, y) is the gray value at pixel (x, y), c1And c2Respectively indicate profile song
Average gray outside line interior intensity average value and contour curve, H () is Heaviside function.
According to the fuzzy velocity function V of reflection profile and border3It can define the border detection item of movable contour model:
Wherein,δ (φ) is the derivative of H ().
In conjunction with both the above formula, the energy functional of movable contour model be may be defined as:
Wherein, the right Section 3 is the canonical for making level set function φ remain distance function in evolutionary process
, i.e. level set function φ meets always in evolutionary processContour curve is avoided in evolution process to level set
Function phi is reinitialized.λ, μ and γ are the weight parameter of control area, border detection item and regular terms, c respectively1
And c2Respectively indicate average gray outside contour curve interior intensity average value and contour curve.
It is general using Euler-Lagrange's variation method, then by the gradient descent method solution movable contour model energy of standard
The minimization problem of letter.Firstly, asking energy functional E (φ) about c respectively under conditions of fixed level set function φ1And c2
Variation, can be obtained
Then, fixed c1And c2, the variation about φ is asked about E (φ) to energy functional.If F () is active contour mould
The integrand of the energy functional formula of type omits variable x and y in writing, then
Wherein, e1=V1(x,y)|I(x,y)-c1|2, e2=V2(x,y)|I(x,y)-c2|2, and the minimum that above formula is corresponding
Change to correspond to and solve following partial differential equation (PDE):
I.e. respectively to φ, φxAnd φySeek partial derivative:
It can further obtain:
Again
Above formula, which is carried out integrating operation, can be obtained
Above formula corresponds to final Euler-Lagrange equation of the energy function about level set function φ.
The gradient descent flow of φ can be obtained by variation principle:
S4, evaluation is compared by three indexs to segmentation precision, positive rate True Positive is respectively adopted
Ratio, RTP, false positive rate False Positive Ratio, RFP and similarity Similarity Degree, DS tri- fingers
Mark is to be evaluated；
Wherein, it is the doctor region that segmentation obtains by hand, is the region that dividing method obtains.RTPReflection is dividing method
To the coverage rate in doctor's manual segmentation region, value is bigger in the region of acquisition, and the region for illustrating that dividing method obtains is got over comprising target
It is more.RFPWhat is reflected is the accounting of background and doctor's manual segmentation region that the region that dividing method obtains includes, is worth smaller, explanation
The region that dividing method obtains includes less background.DSReflection is the region of dividing method acquisition to doctor's manual segmentation area
The similarity in domain, value is bigger, illustrates that the region that dividing method obtains and the region that doctor's craft segmentation obtains are similar to very much.
It should be noted that, in this document, relational terms such as first and second and the like are used merely to a reality
Body or operation are distinguished with another entity or operation, are deposited without necessarily requiring or implying between these entities or operation
In any actual relationship or order or sequence.Moreover, the terms "include", "comprise" or its any other variant are intended to
Non-exclusive inclusion, so that the process, method, article or equipment including a series of elements is not only wanted including those
Element, but also including other elements that are not explicitly listed, or further include for this process, method, article or equipment
Intrinsic element.
It although an embodiment of the present invention has been shown and described, for the ordinary skill in the art, can be with
A variety of variations, modification, replacement can be carried out to these embodiments without departing from the principles and spirit of the present invention by understanding
And modification, the scope of the present invention is defined by the appended.
Claims (3)
1. a kind of pleura contact Lung neoplasm dividing method based on Robust Speed function, it is characterised in that: the following steps are included:
S1, the F-KNN sorting algorithm that will pass through simply dealt lung CT image combination Gray-level co-occurrence and feature, calculate robust
Probability score in velocity function；
Robust Speed function in S2, tectonic activity skeleton pattern；
S3, calculate movable contour model energy functional and carried out minimum calculation processing；
S4, evaluation is compared by three indexs to segmentation precision.
2. a kind of pleura contact Lung neoplasm dividing method based on Robust Speed function according to claim 1, special
Sign is, is specially first to combine wavelet energy and latent structure eigenmatrix in S1, is then constructed using F-KNN sorting algorithm
The probability score matrix of Lung neoplasm image.
3. a kind of pleura contact Lung neoplasm dividing method based on Robust Speed function according to claim 1, special
Sign is that three in S4 index is specially three positive rate, false positive rate and similarity indexs, and comparison is and doctor's hand
Dynamic segmentation result compares.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910412792.6A CN110136124A (en) | 2019-05-17 | 2019-05-17 | A kind of pleura contact Lung neoplasm dividing method based on Robust Speed function |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910412792.6A CN110136124A (en) | 2019-05-17 | 2019-05-17 | A kind of pleura contact Lung neoplasm dividing method based on Robust Speed function |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN110136124A true CN110136124A (en) | 2019-08-16 |
Family
ID=67574945
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910412792.6A Pending CN110136124A (en) | 2019-05-17 | 2019-05-17 | A kind of pleura contact Lung neoplasm dividing method based on Robust Speed function |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110136124A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112607363A (en) * | 2020-12-16 | 2021-04-06 | 长沙聿鸿环保科技有限公司 | Intelligent bulk material conveying system and application thereof |
Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101620670A (en) * | 2008-07-01 | 2010-01-06 | 邹采荣 | Fuzzy discriminant analysis method applied to facial expression recognition |
| CN102999593A (en) * | 2012-11-16 | 2013-03-27 | 江苏大学 | Breast cancer distinguishing method based on gaussian kernel function fuzzy noncorrelation distinguishing conversion |
| CN103268635A (en) * | 2013-05-15 | 2013-08-28 | 北京交通大学 | Segmentation and semantic annotation method of geometry grid scene model |
| CN103632144A (en) * | 2013-12-05 | 2014-03-12 | 江苏大学 | Fuzzy two-dimensional k-nearest neighbor based face recognition method |
| CN106203528A (en) * | 2016-07-19 | 2016-12-07 | 华侨大学 | A kind of feature based merges and the 3D of KNN draws intelligent classification algorithm |
| 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 |
| CN107392142A (en) * | 2017-07-19 | 2017-11-24 | 广东工业大学 | A kind of true and false face identification method and its device |
| CN108876589A (en) * | 2018-05-03 | 2018-11-23 | 深圳信息职业技术学院 | Methods of risk assessment, device, equipment and storage medium |
| CN109741349A (en) * | 2019-01-24 | 2019-05-10 | 桂林航天工业学院 | A kind of method of cerebral arterial thrombosis image segmentation |
-
2019
- 2019-05-17 CN CN201910412792.6A patent/CN110136124A/en active Pending
Patent Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101620670A (en) * | 2008-07-01 | 2010-01-06 | 邹采荣 | Fuzzy discriminant analysis method applied to facial expression recognition |
| CN102999593A (en) * | 2012-11-16 | 2013-03-27 | 江苏大学 | Breast cancer distinguishing method based on gaussian kernel function fuzzy noncorrelation distinguishing conversion |
| CN103268635A (en) * | 2013-05-15 | 2013-08-28 | 北京交通大学 | Segmentation and semantic annotation method of geometry grid scene model |
| CN103632144A (en) * | 2013-12-05 | 2014-03-12 | 江苏大学 | Fuzzy two-dimensional k-nearest neighbor based face recognition method |
| CN106203528A (en) * | 2016-07-19 | 2016-12-07 | 华侨大学 | A kind of feature based merges and the 3D of KNN draws intelligent classification algorithm |
| 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 |
| CN107392142A (en) * | 2017-07-19 | 2017-11-24 | 广东工业大学 | A kind of true and false face identification method and its device |
| CN108876589A (en) * | 2018-05-03 | 2018-11-23 | 深圳信息职业技术学院 | Methods of risk assessment, device, equipment and storage medium |
| CN109741349A (en) * | 2019-01-24 | 2019-05-10 | 桂林航天工业学院 | A kind of method of cerebral arterial thrombosis image segmentation |
Non-Patent Citations (7)
| Title |
|---|
| SARAH SOLTANINEJAD等: "Lung nodule detection by KNN classifier and active contour modelling and 3D visualization", 《THE 16TH CSI INTERNATIONAL SYMPOSIUM ON ARTIFICIAL INTELLIGENCE AND SIGNAL PROCESSING (AISP 2012)》 * |
| 乐昂霖: "基于SVM的肺结节分割与识别", 《中国优秀硕士学位论文全文数据库 医药卫生科技辑》 * |
| 冯宝 等: "小波能量引导下基于活动轮廓模型的部分实性肺结节分割", 《华南理工大学学报( 自然科学版)》 * |
| 冯宝 等: "结合小波能量和汉森形状指数的肺结节分割", 《仪器仪表学报》 * |
| 廖晓磊: "薄扫CT序列图像的肺实质分割和肺结节检测方法研究", 《中国优秀硕士学位论文全文数据库 医药卫生科技辑》 * |
| 李祥霞: "恶性肺结节计算机辅助诊断关键技术研究", 《中国博士学位论文全文数据库 医药卫生科技辑》 * |
| 陈侃: "基于活动轮廓模型的肺结节分割方法研究", 《中国博士学位论文全文数据库 信息科技辑》 * |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112607363A (en) * | 2020-12-16 | 2021-04-06 | 长沙聿鸿环保科技有限公司 | Intelligent bulk material conveying system and application thereof |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Yin et al. | Large scale remote sensing image segmentation based on fuzzy region competition and Gaussian mixture model | |
| Huang et al. | A robust graph-based segmentation method for breast tumors in ultrasound images | |
| Liu et al. | Probability density difference-based active contour for ultrasound image segmentation | |
| Paik et al. | Surface normal overlap: a computer-aided detection algorithm with application to colonic polyps and lung nodules in helical CT | |
| CN109146854B (en) | Analysis method for association relationship between pulmonary nodule and pulmonary blood vessel | |
| CN110766051A (en) | Lung nodule morphological classification method based on neural network | |
| JP2020508143A (en) | Prostate Cancer Detection in Multiparametric MRI Using Random Forest with Instance Weighting and MR Prostate Segmentation by Deep Learning Using Holistic Nested Networks | |
| CN108257135A (en) | The assistant diagnosis system of medical image features is understood based on deep learning method | |
| CN107590797A (en) | A kind of CT images pulmonary nodule detection method based on three-dimensional residual error neutral net | |
| Feng et al. | Multiphase SAR image segmentation with $ G^{0} $-statistical-model-based active contours | |
| El-Baz et al. | Three-dimensional shape analysis using spherical harmonics for early assessment of detected lung nodules | |
| US20080118136A1 (en) | Propagating Shell for Segmenting Objects with Fuzzy Boundaries, Automatic Volume Determination and Tumor Detection Using Computer Tomography | |
| US20080002870A1 (en) | Automatic detection and monitoring of nodules and shaped targets in image data | |
| CN105139015B (en) | A kind of remote sensing images Clean water withdraw method | |
| CN103745227A (en) | Method for identifying benign and malignant lung nodules based on multi-dimensional information | |
| CN109978882A (en) | A kind of medical imaging object detection method based on multi-modal fusion | |
| CN110706225B (en) | Tumor identification system based on artificial intelligence | |
| Zhang et al. | Level set evolution driven by optimized area energy term for image segmentation | |
| CN106960226B (en) | A kind of soft subspace MR image clustering methods based on fireworks algorithm | |
| CN106504239B (en) | A kind of method of liver area in extraction ultrasound image | |
| Yue et al. | Multiscale roughness measure for color image segmentation | |
| CN109461163A (en) | A kind of edge detection extraction algorithm for magnetic resonance standard water mould | |
| Li et al. | Segmentation of pulmonary nodules using adaptive local region energy with probability density function-based similarity distance and multi-features clustering | |
| Li et al. | Detection of pulmonary nodules in CT images based on fuzzy integrated active contour model and hybrid parametric mixture model | |
| CN110136124A (en) | A kind of pleura contact Lung neoplasm dividing method based on Robust Speed function |
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 | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20190816 |
|
| RJ01 | Rejection of invention patent application after publication |