Deep learning in medical image registration: introduction and survey

Ahmad Hammoudeh,Stéphane Dupont

2024-01-10

Abstract:Image registration (IR) is a process that deforms images to align them with respect to a reference space, making it easier for medical practitioners to examine various medical images in a standardized reference frame, such as having the same rotation and scale. This document introduces image registration using a simple numeric example. It provides a definition of image registration along with a space-oriented symbolic representation. This review covers various aspects of image transformations, including affine, deformable, invertible, and bidirectional transformations, as well as medical image registration algorithms such as Voxelmorph, Demons, SyN, Iterative Closest Point, and SynthMorph. It also explores atlas-based registration and multistage image registration techniques, including coarse-fine and pyramid approaches. Furthermore, this survey paper discusses medical image registration taxonomies, datasets, evaluation measures, such as correlation-based metrics, segmentation-based metrics, processing time, and model size. It also explores applications in image-guided surgery, motion tracking, and tumor diagnosis. Finally, the document addresses future research directions, including the further development of transformers.

Image and Video Processing,Computer Vision and Pattern Recognition,Machine Learning

What problem does this paper attempt to address?

The main aim of this paper is to introduce and review the application of deep learning in medical image registration.

The paper first explains the historical origin and definition of the concept of "image registration" and provides a simple numerical example to introduce the process of image registration. It then discusses various constraints of image registration, related review literature, classification systems, and more. Additionally, it details various algorithms used in medical image registration, such as Voxelmorph, Demons, SyN, and explores atlas-based registration and multi-stage image registration techniques. The paper further discusses different metrics used to evaluate the effectiveness of medical image registration, including correlation metrics, segmentation metrics, processing time, model size, and clinical application evaluations.

Moreover, the paper lists some commonly used medical imaging datasets and discusses the applications of image registration in the medical field, such as image-guided surgery, tumor diagnosis and treatment, and motion tracking. Finally, the paper points out future research directions, including the development of deep learning models (such as transformers) and registration-free techniques.

In short, this paper provides a comprehensive understanding of the application of deep learning in medical image registration by reviewing existing research results and technologies, and it indicates possible directions for future researchers.