Cover -- Contents -- List of Contributors -- Preface -- Acronyms -- Part I Introduction -- Chapter 1 Overview of Biomedical Image Understanding Methods -- 1.1 Segmentation and Object Detection -- 1.1.1 Methods Based on Image Processing Techniques -- 1.1.2 Methods Using Pattern Recognition and Machine Learning Algorithms -- 1.1.3 Model and Atlas-Based Segmentation -- 1.1.4 Multispectral Segmentation -- 1.1.5 User Interactions in Interactive Segmentation Methods -- 1.1.6 Frontiers of Biomedical Image Segmentation -- 1.2 Registration -- 1.2.1 Taxonomy of Registration Methods -- 1.2.2 Frontiers of Registration for Biomedical Image Understanding -- 1.3 Object Tracking -- 1.3.1 Object Representation -- 1.3.2 Feature Selection for Tracking -- 1.3.3 Object Tracking Technique -- 1.3.4 Frontiers of Object Tracking -- 1.4 Classification -- 1.4.1 Feature Extraction and Feature Selection -- 1.4.2 Classifiers -- 1.4.3 Unsupervised Classification -- 1.4.4 Classifier Combination -- 1.4.5 Frontiers of Pattern Classification for Biomedical Image Understanding -- 1.5 Knowledge-Based Systems -- 1.5.1 Semantic Interpretation and Knowledge-Based Systems -- 1.5.2 Knowledge-Based Vision Systems -- 1.5.3 Knowledge-Based Vision Systems in Biomedical Image Analysis -- 1.5.4 Frontiers of Knowledge-Based Systems -- References -- Part II Segmentation and Object Detection -- Chapter 2 Medical Image Segmentation and its Application in Cardiac MRI -- 2.1 Introduction -- 2.2 Background -- 2.2.1 Active Contour Models -- 2.2.2 Parametric and Nonparametric Contour Representation -- 2.2.3 Graph-Based Image Segmentation -- 2.2.4 Summary -- 2.3 Parametric Active Contours -- The Snakes -- 2.3.1 The Internal Spline Energy E int -- 2.3.2 The Image-Derived Energy E img -- 2.3.3 The External Control Energy E con.

2.3.4 Extension of Snakes and Summary of Parametric Active Contours -- 2.4 Geometric Active Contours -- The Level Sets -- 2.4.1 Variational Level Set Methods -- 2.4.2 Region-Based Variational Level Set Methods -- 2.4.3 Summary of Level Set Methods -- 2.5 Graph-Based Methods -- The Graph Cuts -- 2.5.1 Basic Graph Cuts Formulation -- 2.5.2 Patch-Based Graph Cuts -- 2.5.3 An Example of Graph Cuts -- 2.5.4 Summary of Graph Cut Methods -- 2.6 Case Study: Cardiac Image Segmentation Using A Dual Level Sets Model -- 2.6.1 Introduction -- 2.6.2 Method -- 2.6.3 Experimental Results -- 2.6.4 Conclusion of the Case Study -- 2.7 Conclusion and Near-Future Trends -- References -- Chapter 3 Morphometric Measurements of the Retinal Vasculature in Fundus Images With Vampire -- 3.1 Introduction -- 3.2 Assessing Vessel Width -- 3.2.1 Previous Work -- 3.2.2 Our Method -- 3.2.3 Results -- 3.2.4 Discussion -- 3.3 Artery or Vein? -- 3.3.1 Previous Work -- 3.3.2 Our Solution -- 3.3.3 Results -- 3.3.4 Discussion -- 3.4 Are My Program's Measurements Accurate? -- 3.4.1 Discussion -- References -- Chapter 4 Analyzing Cell and Tissue Morphologies Using Pattern Recognition Algorithms -- 4.1 Introduction -- 4.2 Texture Segmentation of Endometrial Images Using the Subspace Mumford--Shah Model -- 4.2.1 Subspace Mumford--Shah Segmentation Model -- 4.2.2 Feature Weights -- 4.2.3 Once-and-For-All Approach -- 4.2.4 Results -- 4.3 Spot Clustering for Detection of Mutants in Keratinocytes -- 4.3.1 Image Analysis Framework -- 4.3.2 Results -- 4.4 Cells and Nuclei Detection -- 4.4.1 Model -- 4.4.2 Neural Cells and Breast Cancer Cells Data -- 4.4.3 Performance Evaluation -- 4.4.4 Robustness Study -- 4.4.5 Results -- 4.5 Geometric Regional Graph Spectral Feature -- 4.5.1 Conversion of Image Patches into Region Signatures.

4.5.2 Comparing Region Signatures -- 4.5.3 Classification of Region Signatures -- 4.5.4 Random Masking and Object Detection -- 4.5.5 Results -- 4.6 Mitotic Cells in the H&E Histopathological Images of Breast Cancer Carcinoma -- 4.6.1 Mitotic Index Estimation -- 4.6.2 Mitotic Candidate Selection -- 4.6.3 Exclusive Independent Component Analysis (XICA) -- 4.6.4 Classification Using Sparse Representation -- 4.6.5 Training and Testing Over Channels -- 4.6.6 Results -- 4.7 Conclusions -- References -- Part III Registration and Matching -- Chapter 5 3D Nonrigid Image Registration by Parzen-Window-Based Normalized Mutual Information and its Application on Mr-Guided Microwave Thermocoagulation of Liver Tumors -- 5.1 Introduction -- 5.2 Parzen-Window-Based Normalized Mutual Information -- 5.2.1 Definition of Parzen-Window Method -- 5.2.2 Parzen-Window-Based Estimation of Joint Histogram -- 5.2.3 Normalized Mutual Information and its Derivative -- 5.3 Analysis of Kernel Selection -- 5.3.1 The Designed Kernel -- 5.3.2 Comparison in Theory -- 5.3.3 Comparison by Experiments -- 5.4 Application on MR-Guided Microwave Thermocoagulation of Liver Tumors -- 5.4.1 Introduction of MR-Guided Microwave Thermocoagulation of Liver Tumors -- 5.4.2 Nonrigid Registration by Parzen-Window-Based Mutual Information -- 5.4.3 Evaluation on Phantom Data -- 5.4.4 Evaluation on Clinical Cases -- 5.5 Conclusion -- Acknowledgements -- References -- Chapter 6 2D/3D Image Registration For Endovascular Abdominal Aortic Aneurysm (AAA) Repair -- 6.1 Introduction -- 6.2 Background -- 6.2.1 Image Modalities -- 6.2.2 2D/3D Registration Framework -- 6.2.3 Feature-Based Registration -- 6.2.4 Intensity-Based Registration -- 6.2.5 Number of Imaging Planes -- 6.2.6 2D/3D Registration for Endovascular AAA Repair.

6.3 Smart Utilization of Two X-Ray Images for Rigid-Body 2D/3D Registration -- 6.3.1 2D/3D Registration: Challenges in EVAR -- 6.3.2 3D Image Processing and DRR Generation -- 6.3.3 2D Image Processing -- 6.3.4 Similarity Measure -- 6.3.5 Optimization -- 6.3.6 Validation -- 6.4 Deformable 2D/3D Registration -- 6.4.1 Problem Formulation -- 6.4.2 Graph-Based Difference Measure -- 6.4.3 Length Preserving Term -- 6.4.4 Smoothness Term -- 6.4.5 Optimization -- 6.4.6 Validation -- 6.5 Visual Check of Patient Movement Using Pelvis Boundary Detection -- 6.6 Discussion and Conclusion -- References -- Part IV Object Tracking -- Chapter 7 Motion Tracking in Medical Images -- 7.1 Introduction -- 7.1.1 Point-Based Tracking -- 7.1.2 Silhouette-Based Tracking -- 7.1.3 Kernel-Based Tracking -- 7.2 Background -- 7.2.1 Point-Based Tracking -- 7.2.2 Silhouette-Based Tracking -- 7.2.3 Kernel-Based Tracking -- 7.2.4 Summary -- 7.3 Bayesian Tracking Methods -- 7.3.1 Kalman Filters -- 7.3.2 Particle Filters -- 7.3.3 Summary of Bayesian Tracking Methods -- 7.4 Deformable Models -- 7.4.1 Mathematical Foundations of Deformable Models -- 7.4.2 Energy-Minimizing Deformable Models -- 7.4.3 Probabilistic Deformable Models -- 7.4.4 Summary of Deformable Models -- 7.5 Motion Tracking Based on the Harmonic Phase Algorithm -- 7.5.1 HARP Imaging -- 7.5.2 HARP Tracking -- 7.5.3 Summary -- 7.6 Case Study: Pseudo Ground Truth-Based Nonrigid Registration of MRI for Tracking the Cardiac Motion -- 7.6.1 Data Fidelity Term -- 7.6.2 Spatial Smoothness Constraint -- 7.6.3 Temporal Smoothness Constraint -- 7.6.4 Energy Minimization -- 7.6.5 Preliminary Results -- 7.6.6 Nonrigid Registration of Myocardial Perfusion MRI -- 7.6.7 Experimental Results -- 7.7 Discussion -- 7.8 Conclusion and Near-Future Trends -- References -- Part V Classification.

Chapter 8 Blood Smear Analysis, Malaria Infection Detection, and Grading from Blood Cell Images -- 8.1 Introduction -- 8.2 Pattern Classification Techniques -- 8.2.1 Supervised and Nonsupervised Learning -- 8.2.2 Bayesian Decision Theory -- 8.2.3 Clustering -- 8.2.4 Support Vector Machines -- 8.3 GWA Detection -- 8.3.1 Image Analysis -- 8.3.2 Association between the Object Area and the Number of Cells Per Object -- 8.3.3 Clump Splitting -- 8.3.4 Clump Characterization -- 8.3.5 Classification -- 8.4 Dual-Model-Guided Image Segmentation and Recognition -- 8.4.1 Related Work -- 8.4.2 Strategies and Object Functions -- 8.4.3 Endpoint Adjacency Map Construction and Edge Linking -- 8.4.4 Parsing Contours and Their Convex Hulls -- 8.4.5 A Recursive and Greedy Splitting Approach -- 8.4.6 Incremental Model Updating and Bayesian Decision -- 8.5 Infection Detection and Staging -- 8.5.1 Related Work -- 8.5.2 Methodology -- 8.6 Experimental Results -- 8.6.1 GWA Classification -- 8.6.2 RBC Segmentation -- 8.6.3 RBC Classification -- 8.7 Summary -- References -- Chapter 9 Liver Tumor Segmentation Using SVM Framework and Pathology Characterization Using Content-Based Image Retrieval -- 9.1 Introduction -- 9.2 Liver Tumor Segmentation Under a Hybrid SVM Framework -- 9.2.1 Fundamentals of SVM for Classification -- 9.2.2 SVM Framework for Liver Tumor Segmentation and the Problems -- 9.2.3 A Three-Stage Hybrid SVM Scheme for Liver Tumor Segmentation -- 9.2.4 Experiment -- 9.2.5 Evaluation Metrics -- 9.2.6 Results -- 9.3 Liver Tumor Characterization by Content-Based Image Retrieval -- 9.3.1 Existing Work and the Rationale of Using CBIR -- 9.3.2 Methodology Overview and Preprocessing -- 9.3.3 Tumor Feature Representation -- 9.3.4 Similarity Query and Tumor Pathological Type Prediction -- 9.3.5 Experiment -- 9.3.6 Results -- 9.4 Discussion.

9.4.1 About Liver Tumor Segmentation Using Machine Learning.