A comprehensive guide to understanding and interpreting
digital images in medical and functional applications

Biomedical Image Understanding focuses on image
understanding and semantic interpretation, with clear introductions
to related concepts, in-depth theoretical analysis, and detailed
descriptions of important biomedical applications. It covers image
processing, image filtering, enhancement, de-noising, restoration,
and reconstruction; image segmentation and feature extraction;
registration; clustering, pattern classification, and data
fusion.

With contributions from experts in China, France, Italy, Japan,
Singapore, the United Kingdom, and the United States, Biomedical
Image Understanding:

* Addresses motion tracking and knowledge-based systems, two
areas which are not covered extensively elsewhere in a biomedical
context

* Describes important clinical applications, such as virtual
colonoscopy, ocular disease diagnosis, and liver tumor
detection

* Contains twelve self-contained chapters, each with an
introduction to basic concepts, principles, and methods, and a case
study or application

With over 150 diagrams and illustrations, this bookis an
essential resource for the reader interested in rapidly advancing
research and applications in biomedical image understanding.

Joo-Hwee Lim is the Head of the Visual Computing
Department at the Institute for Infocomm Research (I²R),
A*STAR, Singapore, and an Adjunct Associate Professor at the School
of Computer Engineering, Nanyang Technological University,
Singapore. He is the co-Director of IPAL (Image & Pervasive
Access Laboratory), a French-Singapore Joint Lab. He established
the medical image analysis group at I²R in 2006,
collaborating with clinicians closely, resulting in strong
competency in ocular imaging, brain image analysis, cell image
understanding etc at the institute. He has published over 200
journal and conference papers and owns 17 patents in the areas of
computer vision, cognitive vision, pattern recognition, and medical
image analysis.

Sim-Heng Ong is an Associate Professor in the Departments of
Electrical Engineering and Bioengineering at the National
University of Singapore. He received his PhD from the University of
Sydney, Australia. His major research areas are computer vision and
medical image analysis and visualization. He has worked extensively
with clinicians in developing algorithms for a variety of medical
applications, and has publications in many highly respected
journals and conferences.

Wei Xiong is a Research Scientist at the Institute for
Infocomm Research (I²R), A*STAR, Singapore. He obtained
his PhD degree from the National University of Singapore. His
research interest is in computer vision, image processing, pattern
classification and acoustic imaging. Dr. Xiong has published over
60 technical papers.

A comprehensive guide to understanding and interpreting digital images in medical and functional applications

Biomedical Image Understanding focuses on image understanding and semantic interpretation, with clear introductions to related concepts, in-depth theoretical analysis, and detailed descriptions of important biomedical applications. It covers image processing, image filtering, enhancement, de-noising, restoration, and reconstruction; image segmentation and feature extraction; registration; clustering, pattern classification, and data fusion.

With contributions from experts in China, France, Italy, Japan, Singapore, the United Kingdom, and the United States, Biomedical Image Understanding:

• Addresses motion tracking and knowledge-based systems, two areas which are not covered extensively elsewhere in a biomedical context
• Describes important clinical applications, such as virtual colonoscopy, ocular disease diagnosis, and liver tumor detection
• Contains twelve self-contained chapters, each with an introduction to basic concepts, principles, and methods, and a case study or application

With over 150 diagrams and illustrations, this bookis an essential resource for the reader interested in rapidly advancing research and applications in biomedical image understanding.

List of Contributors xv

Preface xix

Acronyms xxiii

PART I INTRODUCTION 1

1 Overview of Biomedical Image Understanding Methods 3
Wei Xiong, Jierong Cheng, Ying Gu, Shimiao Li and Joo Hwee Lim

1.1 Segmentation and Object Detection 5

1.1.1 Methods Based on Image Processing Techniques 6

1.1.2 Methods Using Pattern Recognition and Machine Learning Algorithms 7

1.1.3 Model and Atlas-Based Segmentation 8

1.1.4 Multispectral Segmentation 9

1.1.5 User Interactions in Interactive Segmentation Methods 10

1.1.6 Frontiers of Biomedical Image Segmentation 11

1.2 Registration 11

1.2.1 Taxonomy of Registration Methods 12

1.2.2 Frontiers of Registration for Biomedical Image Understanding 15

1.3 Object Tracking 16

1.3.1 Object Representation 17

1.3.2 Feature Selection for Tracking 18

1.3.3 Object Tracking Technique 19

1.3.4 Frontiers of Object Tracking 19

1.4 Classification 20

1.4.1 Feature Extraction and Feature Selection 21

1.4.2 Classifiers 22

1.4.3 Unsupervised Classification 23

1.4.4 Classifier Combination 24

1.4.5 Frontiers of Pattern Classification for Biomedical Image Understanding 25

1.5 Knowledge-Based Systems 26

1.5.1 Semantic Interpretation and Knowledge-Based Systems 26

1.5.2 Knowledge-Based Vision Systems 27

1.5.3 Knowledge-Based Vision Systems in Biomedical Image Analysis 28

1.5.4 Frontiers of Knowledge-Based Systems 29

References 29

PARTII SEGMENTATION AND OBJECT DETECTION 47

2 Medical Image Segmentation and its Application in Cardiac MRI 49
Dong Wei, Chao Li, and Ying Sun

2.1 Introduction 50

2.2 Background 51

2.2.1 Active Contour Models 51

2.2.2 Parametric and Nonparametric Contour Representation 52

2.2.3 Graph-Based Image Segmentation 53

2.2.4 Summary 54

2.3 Parametric Active Contours The Snakes 54

2.3.1 The Internal Spline Energy Eint 54

2.3.2 The Image-Derived Energy Eimg 55

2.3.3 The External Control Energy Econ 55

2.3.4 Extension of Snakes and Summary of Parametric Active Contours 57

2.4 Geometric Active Contours The Level Sets 58

2.4.1 Variational Level Set Methods 58

2.4.2 Region-Based Variational Level Set Methods 60

2.4.3 Summary of Level Set Methods 64

2.5 Graph-Based Methods The Graph Cuts 65

2.5.1 Basic Graph Cuts Formulation 65

2.5.2 Patch-Based Graph Cuts 66

2.5.3 An Example of Graph Cuts 68

2.5.4 Summary of Graph Cut Methods 72

2.6 Case Study: Cardiac Image Segmentation Using A Dual Level Sets Model 73

2.6.1 Introduction 73

2.6.2 Method 74

2.6.3 Experimental Results 79

2.6.4 Conclusion of the Case Study 81

2.7 Conclusion and Near-Future Trends 81

References 83

3 Morphometric Measurements of the Retinal Vasculature in Fundus Images With Vampire 91
Emanuele Trucco, Andrea Giachetti, Lucia Ballerini, Devanjali Relan, Alessandro Cavinato, and Tom Macgillivray

3.1 Introduction 92

3.2 Assessing Vessel Width 93

3.2.1 Previous Work 93

3.2.2 Our Method 94

3.2.3 Results 95

3.2.4 Discussion 96

3.3 Artery or Vein? 98

3.3.1 Previous Work 98