This book provides a comprehensive introduction to current state-of-the-art auto-segmentation approaches used in radiation oncology for auto-delineation of organs-of-risk for thoracic radiation treatment planning. Containing the latest, cutting edge technologies and treatments, it explores deep-learning methods, multi-atlas-based methods, and model-based methods that are currently being developed for clinical radiation oncology applications. Each chapter focuses on a specific aspect of algorithm choices and discusses the impact of the different algorithm modules to the algorithm performance as well as the implementation issues for clinical use (including data curation challenges and auto-contour evaluations).
This book is an ideal guide for radiation oncology centers looking to learn more about potential auto-segmentation tools for their clinic in addition to medical physicists commissioning auto-segmentation for clinical use.
Features:
Up-to-date with the latest technologies in the field
Edited by leading authorities in the area, with chapter contributions from subject area specialists
All approaches presented in this book are validated using a standard benchmark dataset established by the Thoracic Auto-segmentation Challenge held as an event of the 2017 Annual Meeting of American Association of Physicists in Medicine
Recenzijas
"This textbook provides a comprehensive overview of multi-atlas and deep learning approaches to auto-contouring. Furthermore, key questions on clinical implementation are considered. The first introductory chapter describes the main focus of this book being the Thoracic Auto-segmentation Challenge held as an event of the 2017 Annual Meeting of the American Association of Physicists in Medicine (AAPM). Several challenge participants contributed a chapter to this book, addressing a specific strength of their segmentation algorithms. The lack of broad clinical introduction of auto-segmentation, which according to the editors is partly due to the lack of commissioning guidelines, made them dedicate the third part of the book to clinical implementation concerns. The book is written for everyone working in the field of auto-segmentation in radiotherapy. The experienced editors are from academia, clinical physics, and industry; their broad experience gives excellent perspective to this bookThis book was useful toward improving my understanding of deep learning-based methods in medical image segmentation. To the best of my knowledge, this is the only textbook available on auto-segmentation dedicated to radiation oncology. Practical concerns and recommendations for implementation make this textbook a must-have for every radiation oncology department."
Charlotte Brouwer, M.Sc. in Medical Physics (December, 2021)
| Foreword I |
|
ix | |
| Foreword II |
|
xi | |
| Editors |
|
xiii | |
| Contributors |
|
xv | |
|
Chapter 1 Introduction to Auto-Segmentation in Radiation Oncology |
|
|
1 | (12) |
|
|
|
|
|
|
|
PART I Multi-Atlas for Auto-Segmentation |
|
|
|
Chapter 2 Introduction to Multi-Atlas Auto-Segmentation |
|
|
13 | (6) |
|
|
|
Chapter 3 Evaluation of Atlas Selection: How Close Are We to Optimal Selection? |
|
|
19 | (20) |
|
|
|
Chapter 4 Deformable Registration Choices for Multi-Atlas Segmentation |
|
|
39 | (10) |
|
|
|
|
|
|
|
|
|
Chapter 5 Evaluation of a Multi-Atlas Segmentation System |
|
|
49 | (22) |
|
|
|
|
|
|
|
PART II Deep Learning for Auto-Segmentation |
|
|
|
Chapter 6 Introduction to Deep Learning-Based Auto-Contouring for Radiotherapy |
|
|
71 | (10) |
|
|
|
Chapter 7 Deep Learning Architecture Design for Multi-Organ Segmentation |
|
|
81 | (32) |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Chapter 8 Comparison of 2D and 3D U-Nets for Organ Segmentation |
|
|
113 | (12) |
|
|
|
|
|
Chapter 9 Organ-Specific Segmentation Versus Multi-Class Segmentation Using U-Net |
|
|
125 | (8) |
|
|
|
|
|
Chapter 10 Effect of loss functions in deep Learning-bsasea segmentation |
|
|
133 | (18) |
|
|
|
|
|
|
|
|
|
|
|
|
|
Chapter 11 Data Augmentation for Training Deep Neural Networks |
|
|
151 | (14) |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Chapter 12 Identifying Possible Scenarios Where a Deep Learning Auto-Segmentation Model Could Fail |
|
|
165 | (24) |
|
|
|
PART III Clinical Implementation Concerns |
|
|
|
Chapter 13 Clinical Commissioning Guidelines |
|
|
189 | (12) |
|
|
|
Chapter 14 Data Curation Challenges for Artificial Intelligence |
|
|
201 | (16) |
|
|
|
|
|
|
|
|
|
Jayashree Kalpathy-Cramer |
|
|
Chapter 15 On the Evaluation of Auto-Contouring in Radiotherapy |
|
|
217 | (36) |
|
|
| Index |
|
253 | |
Jinzhong Yang earned his BS and MS degrees in Electrical Engineering from the University of
Science and Technology of China, in 1998 and 2001, and his PhD degree in Electrical Engineering
from Lehigh University in 2006. In July 2008, Dr Yang joined the University of Texas MD Anderson
Cancer Center as a Senior Computational Scientist, and since January 2015 he has been an Assistant
Professor of Radiation Physics. Dr Yang is a board-certified medical physicist. His research interest
focuses on deformable image registration and image segmentation for radiation treatment planning
and image-guided adaptive radiotherapy, radiomics for radiation treatment outcome modeling and
prediction, and novel imaging methodologies and applications in radiotherapy.
Greg Sharp earned a PhD in Computer Science and Engineering from the University of Michigan
and is currently Associate Professor in Radiation Oncology at Massachusetts General Hospital
and Harvard Medical School. His primary research interests are in medical image processing and
image-guided radiation therapy, where he is active in the open source software community.
Mark Gooding earned his MEng in Engineering Science in 2000 and DPhil in Medical Imaging
in 2004, both from the University of Oxford. He was employed as a postdoctoral researcher both
in university and hospital settings, where his focus was largely around the use of 3D ultrasound
segmentation in womens health. In 2009, he joined Mirada Medical Ltd, motivated by a desire to
see technical innovation translated into clinical practice. While there, he has worked on a broad
spectrum of clinical applications, developing algorithms and products for both diagnostic and therapeutic
purposes. If given a free choice of research topic, his passion is for improving image segmentation,
but in practice he is keen to address any technical challenge. Dr Gooding now leads the
research team at Mirada, where in addition to the commercial work he continues to collaborate both
clinically and academically.
