Texture-based Deformable Snake Segmentation of the Liver Aaron Mintz Daniela Stan Raicu, PhD Jacob Furst, PhD.

Slides:



Advertisements
Similar presentations
IIIT Hyderabad ROBUST OPTIC DISK SEGMENTATION FROM COLOUR RETINAL IMAGES Gopal Datt Joshi, Rohit Gautam, Jayanthi Sivaswamy CVIT, IIIT Hyderabad, Hyderabad,
Advertisements

Empirical Evaluation of Dissimilarity Measures for Color and Texture
Computer vision: models, learning and inference Chapter 13 Image preprocessing and feature extraction.
嵌入式視覺 Feature Extraction
Computer Vision Lecture 16: Texture
Shaohui Huang, Boliang Wang, Xiaoyang Huang.  Traditional Active Contour (Snake)  Gradient Vector Flow Snake (GVF Snake)  SEGMENT CT IMAGES  Edge.
A Comprehensive Study on Third Order Statistical Features for Image Splicing Detection Xudong Zhao, Shilin Wang, Shenghong Li and Jianhua Li Shanghai Jiao.
Active Contour Models (Snakes)
NCIP SEGMENTATION OF MEDICAL IMAGES USING ACTIVE CONTOURS AND GRADIENT VECTOR FLOW B.Hemakumar M.Tech student, Biomedical signal processing and.
A Computer Aided Detection System For Digital Mammograms Based on Radial Basis Functions and Feature Extraction Techniques By Mohammed Jirari Shanghai,
Texture-Based Image Retrieval for Computerized Tomography Databases Winnie Tsang, Andrew Corboy, Ken Lee, Daniela Raicu and Jacob Furst.
Application of image processing techniques to tissue texture analysis and image compression Advisor : Dr. Albert Chi-Shing CHUNG Presented by Group ACH1.
Three-dimensional co-occurrence matrices & Gabor filters: Current progress Gray-level co-occurrence matrices Carl Philips Gabor filters Daniel Li Supervisor:
Active Contours Technique in Retinal Image Identification of the Optic Disk Boundary Soufyane El-Allali Stephen Brown Department of Computer Science and.
Texture Turk, 91.
MedIX – Summer 06 Lucia Dettori (room 745)
Motion Analysis (contd.) Slides are from RPI Registration Class.
CSci 6971: Image Registration Lecture 4: First Examples January 23, 2004 Prof. Chuck Stewart, RPI Dr. Luis Ibanez, Kitware Prof. Chuck Stewart, RPI Dr.
Texture Texture is a description of the spatial arrangement of color or intensities in an image or a selected region of an image. Structural approach:
Prénom Nom Document Analysis: Data Analysis and Clustering Prof. Rolf Ingold, University of Fribourg Master course, spring semester 2008.
Contrast Enhancement Crystal Logan Mentored by: Dr. Lucia Dettori Dr. Jacob Furst.
Face Recognition Based on 3D Shape Estimation
NSF MedIX REU Program Medical Imaging DePaul CDM Daniela S. Raicu, PhD Associate Professor Lab URL:
Binning Strategies for Tissue Texture Extraction in DICOM Images CTI Students: Bikash Bhattacharyya, Kriti Jauhar Advisors: Dr. Daniela Raicu, Dr. Jacob.
PROJECT 1: Voronoi Probability Maps for Seed Region Detection in Abdominal CT Images PROJECT 2: Kidney Seed Region Detection in Abdominal CT Images.
Texture Classification Based on Co-occurrence Matrices Presentation III Pattern Recognition Mohammed Jirari Spring 2003.
Texture Readings: Ch 7: all of it plus Carson paper
Feature Screening Concept: A greedy feature selection method. Rank features and discard those whose ranking criterions are below the threshold. Problem:
Pattern Recognition. Introduction. Definitions.. Recognition process. Recognition process relates input signal to the stored concepts about the object.
Heather Dunlop : Advanced Perception January 25, 2006
Face Recognition Using Neural Networks Presented By: Hadis Mohseni Leila Taghavi Atefeh Mirsafian.
Entropy and some applications in image processing Neucimar J. Leite Institute of Computing
MASKS © 2004 Invitation to 3D vision Lecture 3 Image Primitives andCorrespondence.
Copyright © 2012 Elsevier Inc. All rights reserved.
Spatial-based Enhancements Lecture 3 prepared by R. Lathrop 10/99 updated 10/03 ERDAS Field Guide 6th Ed. Ch 5: ;
2 Outline Introduction –Motivation and Goals –Grayscale Chromosome Images –Multi-spectral Chromosome Images Contributions Results Conclusions.
Deformable Models Segmentation methods until now (no knowledge of shape: Thresholding Edge based Region based Deformable models Knowledge of the shape.
Texture analysis Team 5 Alexandra Bulgaru Justyna Jastrzebska Ulrich Leischner Vjekoslav Levacic Güray Tonguç.
B. Krishna Mohan and Shamsuddin Ladha
Texture. Texture is an innate property of all surfaces (clouds, trees, bricks, hair etc…). It refers to visual patterns of homogeneity and does not result.
AUTOMATIZATION OF COMPUTED TOMOGRAPHY PATHOLOGY DETECTION Semyon Medvedik Elena Kozakevich.
Automatic Minirhizotron Root Image Analysis Using Two-Dimensional Matched Filtering and Local Entropy Thresholding Presented by Guang Zeng.
Course 9 Texture. Definition: Texture is repeating patterns of local variations in image intensity, which is too fine to be distinguished. Texture evokes.
BING: Binarized Normed Gradients for Objectness Estimation at 300fps
Feature based deformable registration of neuroimages using interest point and feature selection Leonid Teverovskiy Center for Automated Learning and Discovery.
Conclusions The success rate of proposed method is higher than that of traditional MI MI based on GVFI is robust to noise GVFI based on f1 performs better.
MedIX – Summer 07 Lucia Dettori (room 745)
Prostate Cancer CAD Michael Feldman, MD, PhD Assistant Professor Pathology University Pennsylvania.
DATA CLUSTERING WITH KERNAL K-MEANS++ PROJECT OBJECTIVES o PROJECT GOAL  Experimentally demonstrate the application of Kernel K-Means to non-linearly.
Levels of Image Data Representation 4.2. Traditional Image Data Structures 4.3. Hierarchical Data Structures Chapter 4 – Data structures for.
By Brian Lam and Vic Ciesielski RMIT University
CS 641 Term project Level-set based segmentation algorithms Presented by- Karthik Alavala (under the guidance of Dr. Jundong Liu)
Colour and Texture. Extract 3-D information Using Vision Extract 3-D information for performing certain tasks such as manipulation, navigation, and recognition.
A New Method for Crater Detection Heather Dunlop November 2, 2006.
Slides from Dr. Shahera Hossain
Course 5 Edge Detection. Image Features: local, meaningful, detectable parts of an image. edge corner texture … Edges: Edges points, or simply edges,
Image Quality Measures Omar Javed, Sohaib Khan Dr. Mubarak Shah.
MASKS © 2004 Invitation to 3D vision Lecture 3 Image Primitives andCorrespondence.
By Brian Lam and Vic Ciesielski RMIT University
Recognition of biological cells – development
Image Primitives and Correspondence
Texture Classification of Normal Tissues in Computed Tomography
Multi-modality image registration using mutual information based on gradient vector flow Yujun Guo May 1,2006.
Texture Analysis for Pulmonary Nodules Interpretation and Retrieval
Texture Classification of Normal Tissues in Computed Tomography
Visual Computing CTI, DePaul University
Outline Texture modeling - continued Julesz ensemble.
Blobworld Texture Features
Multiple Organ detection in CT Volumes - Week 3
Presentation transcript:

Texture-based Deformable Snake Segmentation of the Liver Aaron Mintz Daniela Stan Raicu, PhD Jacob Furst, PhD

Overview Objectives and Incentives Objectives and Incentives Tested Texture Methods Tested Texture Methods Tested Snake Deformations Tested Snake Deformations Numerical Evaluation Numerical Evaluation Future Work Future Work

Motivations Important Diagnostic Aid to Radiologist Important Diagnostic Aid to Radiologist Liver Cancer: Extremely Deadly Liver Cancer: Extremely Deadly Hypothesis: Texture vs. Intensity-based snake deformation Hypothesis: Texture vs. Intensity-based snake deformation Pixel-to-Pixel area information Pixel-to-Pixel area information Results Show up to 48% Increase in Segmentation Accuracy (Gabor) Results Show up to 48% Increase in Segmentation Accuracy (Gabor)

Process and Methods

Data Archive Original Computed- Tomography Scans Original Computed- Tomography Scans 25 Individual Patients 25 Individual Patients Greatly Varying Patient Sets Greatly Varying Patient Sets DICOM Format DICOM Format Binary Ground Truth Binary Ground Truth 2916 Image-Ground Pairs 2916 Image-Ground Pairs

Image Pre-processing: Gabor Filter Gabor Filter Gabor Filter Gaussian x Sinusoid Gaussian x Sinusoid Various Parameters Various Parameters Aspect Ratio Aspect Ratio Standard Deviation Standard Deviation Wavelength Wavelength Orientation Orientation

Image Pre-processing: Haralick Feature Extraction Locally-Calculated Process Locally-Calculated Process Bin Large Range of Intensity Values Bin Large Range of Intensity Values Window-Based Quantification of Intensity-Value Co-Occurrence Window-Based Quantification of Intensity-Value Co-Occurrence Numerical Analysis of Each Corresponding Matrix to Derive Features Numerical Analysis of Each Corresponding Matrix to Derive Features 9 Features Calculated 9 Features Calculated

Image Pre-processing: Markov Random Fields Also Locally-Calculated Also Locally-Calculated Estimate “Markovianity” of Windowed Regions Estimate “Markovianity” of Windowed Regions Orientation-based Texture Model Orientation-based Texture Model

Snake Constraints Limited Input Limited Input Too Many Corresponding Filters/Features per Image Pixel Too Many Corresponding Filters/Features per Image Pixel Principle Components Analysis Principle Components Analysis Equivalent Number of Principle Components Returned Equivalent Number of Principle Components Returned

Snake Input All Principle Components Evaluated Individually All Principle Components Evaluated Individually Gradient Value Edge Map Gradient Value Edge Map Second Gradient Edge Map Second Gradient Edge Map Automatic Initial Curve Point Selection Automatic Initial Curve Point Selection

Snake Segmentation Methods Traditional Vector Field Model Traditional Vector Field Model Gradient Vector Flow (GVF) Gradient Vector Flow (GVF) Level-Set Evolution Level-Set Evolution

Snake Segmentation Methods (cont.) Balance of Energy Equation Balance of Energy Equation Disadvantages of GVF, Level-Set Disadvantages of GVF, Level-Set

Metrics and Results Computationally Difficult to Evaluate Meaningfully Computationally Difficult to Evaluate Meaningfully Straightforward Measurement of Accuracy Straightforward Measurement of Accuracy 3-Dimensional Analysis 3-Dimensional Analysis Volumetric Overlap Volumetric Overlap Average Distance Average Distance Root-Mean-Square Distance Root-Mean-Square Distance Hausdorff Distance Hausdorff Distance

Results Effectiveness of Texture Heavily Dependent on Region of Liver Depicted Effectiveness of Texture Heavily Dependent on Region of Liver Depicted Gabor Statistics Across 20-Patient Dataset

Future Work Expanding Base of Co-Occurrence and Markov Comparison Expanding Base of Co-Occurrence and Markov Comparison Attempt Combined Principle Components Analysis Attempt Combined Principle Components Analysis Combined Approach – New Automatic Initial Point Selection Combined Approach – New Automatic Initial Point Selection

Credits Carl Philips Carl Philips Dr. Raicu, Dr. Furst Dr. Raicu, Dr. Furst Chenyang Xu, Jerry L. Prince, Chunming Li Chenyang Xu, Jerry L. Prince, Chunming Li

Questions?

Haralick Features Entropy: Entropy: Energy: Energy: Contrast: Contrast: Sum Average: Sum Average: Variance: Variance: Correlation: Correlation: Maximum Probability: Maximum Probability: Inverse Difference Moment: Inverse Difference Moment: Cluster Tendency: Cluster Tendency: