Adaptive Registration and Atlas Based Bayesian Segmentation

Over the last decade medical image analysis techniques have seen dramatic improvements. Medical image analysis is one of the fastest developing areas in medical research. Two main areas of research in medical image analysis are image registration and image segmentation. This book deals with both areas. For image registration, this book focuses on applying a non-linear warping registration especially suitable for soft tissue organs in human body. The degree of freedom (DOF) of warping transform is automatically determined by locally increasing the density of control points in regions where they are needed rather than globally increasing the density of control points. For image segmentation, this book focuses on constructing a probabilistic atlas of an abdomen consisting of 4 organs (i.e., liver, kidneys and spinal cord) and using the atlas information to perform segmentation of abdominal CT scans. The segmentation adopts Bayesian framework and incorporates Markov prior. This book will be useful to anyone, most likely an early year graduate student, who wants to get started in the area of medical image registration and segmentation.

Autorentext
Hyunjin Park received his BS in electrical eng (1997) from Seoul Natl Univ, Korea, MS in biomed eng (2000), and PhD in biomed eng (2003) from Univ of Michigan, Ann Arbor, USA. He was a research faculty with Univ of Michigan, Radiology (2004-2009). He is currently an asst prof with Dept of Biomed eng, Gachon Univ of Medicine and Science, Korea.