Image Registration: Fundamentals and Recent Advances Based on Deep Learning
Min Chen, Nicholas J. Tustison, Rohit Jena, James C. Gee
Abstract
Abstract Registration is the process of establishing spatial correspondences between images. It allows for the alignment and transfer of key information across subjects and atlases. Registration is thus a central technique in many medical imaging applications. This chapter first introduces the fundamental concepts underlying image registration. It then presents recent developments based on machine learning, specifically deep learning, which have advanced the three core components of traditional image registration methods—the similarity functions, transformation models, and cost optimization. Finally, it describes the key application of these techniques to brain disorders.
Topics & Concepts
Image registrationComputer scienceArtificial intelligenceKey (lock)Process (computing)Deep learningSimilarity (geometry)Computer visionTransformation (genetics)Image (mathematics)Computer securityChemistryGeneOperating systemBiochemistryMedical Image Segmentation TechniquesBrain Tumor Detection and ClassificationAdvanced Neural Network Applications