Deformable registration techniques for thoracic CT images : an insight into medical image registration

This book focuses on novel approaches for thoracic computed tomography (CT) image registration and determination of respiratory motion models in a range of patient scenarios. It discusses the use of image registration processes to remove the inconsistencies between medical images acquired using different devices. In the context of comparative research and medical analysis, these methods are of immense value in image registration procedures, not just for thoracic CT images, but for all types of medical images in multiple modalities, and also in establishing a mean respiration motion model. Combined with advanced techniques, the methods proposed have the potential to advance the field of computer vision and help improve existing methods. The book is a valuable resource for those in the scientific community involved in modeling respiratory motion for a large number of people. -- Provided by publisher Contents 5 About the Authors 7 1 Introduction 8 1.1 Background 8 1.2 Motivation 14 1.3 Objective of the Book 17 1.4 Contributions 18 1.5 Organization of the Book 19 2 Theoretical Background 21 2.1 Introduction 21 2.2 Morphological Classification of Images 22 2.2.1 Rigid Images 22 2.2.2 Deformable Images 25 2.3 Geometric Deformation Models: A Survey 29 2.4 Classification of Registration Methodology Used 33 2.4.1 Feature-Based Registration 33 2.4.2 Intensity-Based Registration 36 2.5 Feature Detection/Description Methods 38 2.6 Database Employed 40 2.7 Accuracy and Similarity Measures Used 42 2.7.1 Target Registration Error 42 2.7.2 Signal-to-Noise Ratio (SNR) 44 2.7.3 Peak Signal-to-Noise Ratio (PSNR) 45 2.7.4 Structural Similarity Index (SSIM) 45 2.7.5 Normalized Cross-Correlation (NCC) 46 3 A Moving Least Square Based Framework for Thoracic CT Image Registration 47 3.1 Introduction 48 3.2 Background 48 3.3 Method 52 3.3.1 Preparation 52 3.3.2 Proposed Methodology 52 3.4 Results and Discussion 56 3.5 Conclusion 58 4 A Path Tracing and Deformity Estimation Methodology for Registration of Thoracic CT Image Sequences 60 4.1 Introduction 61 4.2 Background 63 4.3 Method 66 4.3.1 Preparation 66 4.3.2 Proposed Methodology 67 4.4 Results and Discussion 71 4.5 Conclusion 81 5 Deformable Thoracic CT Images Sequence Registration Using Strain Energy Minimization 82 5.1 Introduction 83 5.2 Background 84 5.3 Method 86 5.3.1 Preparation 86 5.3.2 Proposed Methodology 87 5.4 Results and Discussion 90 5.5 Conclusion 94 6 Conclusion and Future Work 96 6.1 Concluding Remarks 96 6.2 Scope for Future Work 97 Appendix A Geometrical Deformation Models for Elastic Images 98 Appendix B 111 Appendix C 115 Appendix D 120 Appendix E 124 References 134 Front Matter ....Pages i-vii Introduction (Ali Imam Abidi, S. K. Singh)....Pages 1-13 Theoretical Background (Ali Imam Abidi, S. K. Singh)....Pages 15-40 A Moving Least Square Based Framework for Thoracic CT Image Registration (Ali Imam Abidi, S. K. Singh)....Pages 41-53 A Path Tracing and Deformity Estimation Methodology for Registration of Thoracic CT Image Sequences (Ali Imam Abidi, S. K. Singh)....Pages 55-76 Deformable Thoracic CT Images Sequence Registration Using Strain Energy Minimization (Ali Imam Abidi, S. K. Singh)....Pages 77-90 Conclusion and Future Work (Ali Imam Abidi, S. K. Singh)....Pages 91-92 Back Matter ....Pages 93-135