Presentation on deformable model for medical image segmentation
-
Upload
subhash-basistha -
Category
Technology
-
view
448 -
download
1
description
Transcript of Presentation on deformable model for medical image segmentation
PRESENTATIONONDEFORMABLE MODEL FOR MEDICAL IMAGE SEGMENTATION
Presented By Subhash BasishthaAssam University Silchar
Presentation Outline Introduction to Image Processing Steps of Image Processing Types of Image Processing Introduction to Image Segmentation Introduction to Medical Image Segmentation Application of Image Segmentation Example of Image Segmentation Need for Deformable Model What is Deformable Model?? Types of Deformable Model
Introduction to Image Processing Image processing is a method to
convert an image into digital form and perform some operations on it, in order to get an enhanced image or to extract some useful information from it.
It is a type of signal dispensation in which input is image, like video frame or photograph and output may be image or characteristics associated with that image.
Steps of Image Processing Importing the image with optical scanner or
by digital photography. Analyzing and manipulating the image
which includes data compression and image enhancement and spotting patterns that are not to human eyes like tissue,satellite photographs.
Output is the last stage in which result can be altered image or report that is based on image analysis.
Phases of digital image processing
[13]
Contd.. Image processing Steps are:1. Image Acquisition2. Image Enhancement3. Image Restoration4. Color Image Processing5. Wavelets and Multiresolution Processing6. Compression7. Morphological Processing8. Segmentation9. Representation and Description10.Object recognition11.Knowledge Base
PurposeThe purpose of image processing is divided
into 5 groups. They are: 1. Visualization - Observe the objects that
are not visible. 2. Image sharpening and restoration - To
create a better image. 3. Image retrieval - Seek for the image of
interest. 4. Measurement of pattern – Measures
various objects in an image. 5. Image Recognition – Distinguish the
objects in an image.
Types
The two types of methods used for Image Processing are Analog and Digital Image Processing.
Analog or visual techniques of image processing can be used for the hard copies like printouts and photographs.
Digital Processing techniques help in manipulation of the digital images by using computers.
Introduction to Image Segmentation The purpose of image segmentation is to partition an image into
meaningful regions with respect to a particular application.
The goal of segmentation is to simplify and/or change the representation of an image into something that is more meaningful and easier to analyze
The segmentation is based on measurements taken from the image and might be grey level, colour, texture, depth or motion.
Usually image segmentation is an initial and vital step in a series of processes aimed at overall image understanding.
Introduction to Medical Image Segmentation An important goal of medical image processing is to
transform raw images into a numerically symbolic form for better representation, evaluation, and /or content based search and mining.
With the advent of medical image modalities that provide different measures of internal anatomical structure and function, physicians are now able to perform typical clinical tasks such as patient diagnosis and monitoring more safely than before such imaging technologies existed.
Computerized image segmentation has played an increasingly important role in medical imaging.
Application of Image Segmentation Segmented images are now used routinely in a
multitude of different applications, such as
- the quantification of tissue volumes
- diagnosis
- localization of pathology
- study of anatomical structure
- treatment planning and
- computer integrated surgery etc.
Application
[4]
Application
[5]
Example of Image Segmentation
Example 1 Segmentation based on greyscale Very simple ‘model’ of greyscale leads to inaccuracies
in object labelling
Example of Image Segmentation
Example 2 Segmentation based on texture Enables object surfaces with varying patterns of grey to be
segmented
Boundary Extraction example
Dilation: Dilation is an operation that ‘grows’ or ‘thickens’ objects in a binary image. Mathematically, dilation is defined in terms of set operations. The dilation of A and B is defined as
Erosion: Erosion is an operation that ‘Shrinks’ or ‘thins’ objects in a binary image. The mathematical definition of erosion of A by B is as
Dilation and Erosion
ABzBA z)(|
^
Cz ABzBA )(|
Algorithm
Add pixels to the boundaries of an image.
Numbers of pixels added from the objetcs in an image depends on the size of the structuring Element.
Function strel is used to generate the SE s.
Algorithm for Boundary Extraction
Erosion Algorithm: The boundary of a set A, denoted by β(A), can be obtained by first eroding A by B and then performing the set differences between A and its erosion. That is,
β(A)=A – (AΘB)
Dilation Algorithm: The boundary of a set A, denoted by β(A), can be obtained by first dilating A by B and then performing the set differences between A and its dilation. That is, β(A)= (A B) – A
Boundary Extraction with the help of Dilation:A=imread(‘a.jpg');
s=strel('disk',3);%Structuring element
F=imdilate(A,s); %Dialte the image by structuring element
figure,imshow(A);title('Original Image');
figure,imshow(F);title('Imdilate Image');
figure,imshow(F-A);title('Boundary extracted Image with using imdilate');
Boundary Extraction with the help of Erosion:A=imread('a.jpg');
s=strel('disk',3); %Structuring element
F=imerode(A,s); %Erode the image by structuring element
figure,imshow(A); title('Original Image');
figure,imshow(A-F); title('Boundary extracted Image with using imerode');
Matlab Practical
Need for Deformable Model
Due to both the tremendous variability of object shapes and the variation in image quality, Image Segmentation becomes a difficult task.
Problems do arise when medical images are corrupted with noise and the structure itself is not clearly or completely visible in the image.
This may result in detecting erroneous object regions or boundaries, or failing to detect true ones when applying classical segmentation techniques such as edge detection and thresholding.
To address these difficulties, deformable models have been extensively studied and widely used in medical image segmentation, with promising results.
What is Deformable Model??
Deformable models analyze those noisy images and provide a coherent representation for variable structure shapes.
Deformable models are curves or surfaces defined within an image domain that can move under the influence of internal forces and external forces.
Internal forces are defined within the curve or surface itself, and are designed to keep the model smooth during deformation.
External forces, are computed from the image data and are defined to move the model toward an object boundary or other desired features within an image.
What is Deformable Model?? They are designed to be attracted to external image features
(such as edges) while maintaining internal shape constraints (such as smoothness).
By constraining extracted boundaries to be smooth and incorporating other prior information about the object shape,
Deformable models offer:
- robustness to both image noise and boundary gaps.
Deformable models allow integrating boundary elements into a coherent and consistent mathematical description readily available for subsequent applications.
How Become Popular The popularity of deformable model is largely due to the
seminar paper “Snakes:ActiveContours” by Kass, Witkin, and Terzopoulos.
Since its publication, deformable models have grown to be one of the most active and successful research areas in image segmentation.
Image Segmentation Using Deformable Models
Fig: Variability of object shapes and imag equality. (a) A 2D MR image of the heartleft ventricle and (b) a 3D MR image of the brain.
Image Segmentation Using Deformable Models
Fig: Examples of using deformable models to extract object boundaries from medicalimages. (a) An example of using a deformable contour to extract the inner wall of the leftventricle of a human heart from a 2D MR image. (b) (b)An example of usinga deformable surface to reconstruct the brain cortical surface from a 3D MR image
Types of Deformable Model
There are basically two types of deformable models depending on how the model is defined in the shape domain.:
a) parametric deformable models Or active contours,
b) geometric deformable models Or implicit models.
Parametric Deformable Models Parametric deformable models or active contours were
first introduced in 1988, by Kass et al., under the name “snakes” .
Snakes, use parametric curves to represent the model shape.
During their evolution, the deformations are determined by geometry, kinematics, dynamics etc.
Parametric Deformable Models Parametric deformable models represent curves and surfaces
explicitly in their parametric forms during deformation.
This representation allows direct interaction with the model and can lead to a compact representation for fast real-time implementation.
The main advantage of parametric models is that they are usually very fast in their convergence, depending on the predetermined number of control points.
Problem: Adaptation of the model topology, such as splitting or merging parts during the deformation, can be difficult using parametric
models i.e it is topology dependent.
Parametric Deformable Models Two different types of formulations for parametric
deformable models:
- an energy minimizing formulation and
- a dynamic force formulation.
The first formulation has the advantage that its solution satisfies a minimum principle whereas the second formulation has the flexibility of allowing the use of more general types of external forces.
Internal Energy of Active Contours: The internal energy of an active contour can
be translated as the summation of forces applied along the curve to preserve its smoothness.
Mathematically
Where the individual energies eint represent the local state along the curve, and are
defined as,
External Energy of Active Contours Common active contours use primarily edge
(image gradient) information to derive external image forces that drive a shape based model.
External energy,
where is the image I after smoothing with a Gaussian kernel of standard deviation , and is the image gradient along the curve C.
Edge Based Active Contour
Edge Detection is a well-developed field on its own within image processing. Region boundaries and edges are closely related, since there is often a sharp adjustment in intensity at the region boundaries.
Region Based Active Contour
This method takes a set of seeds as input along with the image. The seeds mark each of the objects to be segmented.
Geometric Deformable Model Geometric deformable models, on the other hand, can
handle topological changes naturally. These model is based on the theory of curve evolution
and the level set method. These model represent curves and surfaces implicitly as
a level set of a higher-dimensional scalar function. Their parameterizations are computed only after
complete deformation, thereby allowing topological adaptivity to be easily accommodated.
Geometric Deformable Model They use a distance transformation to define the shape
from the n-dimensional to an n +1 dimensional domain, where n=1 for curves, n= 2 for surfaces on the image plane,etc.
Region Based & Segment Based
Region-based segmentation: Considers gray-levels from neighboring pixels , either by including similar neighboring pixels(region growing), split-and-merge, or watershed segmentation.
• Edge-based segmentation: Detects edge pixels and links them together to form contours.
Region vs. edge-based approaches Region based methods are robust because: – Regions cover more pixels than edges and thus you
have more information available in order to characterize your region
– When detecting a region you could for instance use texture which is not easy when dealing with edges
– Region growing techniques are generally better in noisy images where edges are difficult to detect
• The edge based method can be preferable because: – Algorithms are usually less complex – Edges are important features in a image to separate
regions. The edge of a region can often be hard to find because of noise or occlusions
Conclusion
In this paper, I have discussed medical image segmentation methods, particularly deformable model based methods, learning based classification methods . Understanding image content and extracting useful image features are critical to medical image search and mining.
Future Research We all are in midst of revolution ignited by
fast development in computer technology and imaging. Against common belief, computers are not able to match humans in calculation related to image processing and analysis. But with increasing sophistication and power of the modern computing, computation will go beyond conventional. Parallel and distributed computing paradigms are anticipated to improve responses for the image processing results.
References E L. Valverd, N . Guil, A DEFORMABLE MODEL FOR IMAGE SEGMENTATION IN NOISY
MEDICAL IMAGES,IEEE 2001. E L. Valverd, N .Guil, “A DEFORMABLE MODEL FOR IMAGE SEGMENTATION IN NOISY
MEDICAL IMAGES”,IEEE 2001. Xiaolei Huang “Medical Image Segmentation”, University of Miami Digital Image Segmentation
http://www.uio.no/studier/emner/matnat/ifi/INF4300/h10/undervisningsmateriale/INF4300-2010-f03-segmentation.pdf
Luis Ibfiez , Chafid Hamitouche , Christian Roux,” Structured Deformable Models Application of Metaballs to 3D Medical Image Segmentation”, , Annual EMBS International Conference
Lifeng Liu and Stan Sclaroff, “Medical Image Segmentation And Retrieval Via Deformable Models”, Boston University.
Blake A, Rother C, Brown M, Perez P, Torr P (2004) Interactive image segmentation using an adaptive GMMRF model. European Conference on Computer Vision, 2004
Xian Fan, Pierre-Louis Bazin, John Bogovic, Ying Bai and Jerry L. Prince ,“ A Multiple Geometric Deformable Model Framework for Homeomorphic 3D Medical Image Segmentation” International Journal of Advanced Research in Computer Science and Software Engineering.
Chou P, Brown C (1990) The theory and practice of bayesian image labeling. Int J Comput Vis 4:185–210
Y. S. Akgul and C. Kambhamettu, “A coarse-to-fine deformable contour
optimization framework,” IEEE Trans. on Pattern Analysis and Machine Intelligence, vol. 25, no. 2, pp. 174–186, 2003.
P. Etyngier, F. Segonne, and R. Keriven, “Shape priors using manifold
learning techniques,” in IEEE 11th Int. Conf. Comput. Vis., Oct. 2007,pp. 1–8.
http://www.engineersgarage.com/articles/image-processing-tutorial-
applications http://en.wikipedia.org/wiki/Dilation_(morphology) http://tanmayonrun.blogspot.in/2011/10/describe-fundamental-steps-of-
digital.html
THANK YOU