Abstract:Medical image registration is vital for disease diagnosis and treatment with its ability to merge diverse information of images, which may be captured under different times, angles, or modalities. Although several surveys have reviewed the development of medical image registration, these surveys have not systematically summarized methodologies of existing medical image registration methods. To this end, we provide a comprehensive review of these methods from traditional and deep learning-based directions, aiming to help audiences understand the development of medical image registration quickly. In particular, we review recent advances in retinal image registration at the end of each section, which has not attracted much attention. Additionally, we also discuss the current challenges of retinal image registration and provide insights and prospects for future research.

What problem does this paper attempt to address?

This paper focuses on medical image registration and its application in retinal imaging. Medical image registration is a crucial step in fusing multiple images captured at different times, angles, or modalities to assist in disease diagnosis and treatment. Although previous studies have reviewed the development of medical image registration, there is a lack of systematic summaries of existing methods. Therefore, this paper provides a comprehensive review, discussing both traditional methods and deep learning-based methods, aiming to help readers quickly understand the progress of medical image registration. The paper specifically highlights the latest progress in retinal image registration, which is a relatively less studied area. The authors also discuss the current challenges in retinal image registration and provide insights and prospects for future research. Medical image registration methods can be mainly categorized into global coarse linear registration and local fine elastic registration. Additionally, methods can be divided into intensity-based and feature-based approaches based on how the matching images are conducted. In recent years, deep learning has played a significant role in medical image registration by learning features to provide more flexible and problem-specific representations, replacing manually designed features. These deep learning methods include convolutional neural networks (CNNs), generative adversarial networks (GANs), and Transformers, which can directly obtain registration results from input image pairs, improving efficiency and performance. As retinal images are a unique part of the human body that allows non-invasive observation of live blood vessels, their registration is crucial in integrating structural and functional information from different modalities. However, there is a relative scarcity of systematic reviews on retinal image registration, and they often focus on a single modality. This paper addresses this gap by analyzing the strengths and limitations of different methods and pointing out the current challenges, such as the lack of openly available datasets, the limitations of deep learning-based linear registration methods, and the inaccuracies in similarity measures. It also proposes future research directions.

Medical Image Registration and Its Application in Retinal Images: A Review

Medical image registration and its application in retinal images: a review

Review of Image Registration Methods for Medical Images

A Review of Deep Learning-Based Three-Dimensional Medical Image Registration Methods

Color Fundus Photograph Registration Based on Feature and Intensity for Longitudinal Evaluation of Diabetic Retinopathy Progression

A Review of Medical Image Registration for Different Modalities

A survey on deep learning in medical image registration: new technologies, uncertainty, evaluation metrics, and beyond

Deep learning in medical image registration: introduction and survey

Medical image identification methods: A review

Deep Learning in Medical Image Registration: A Review

Deep learning in medical image registration: a survey

Deep Learning for Medical Image Registration: A Comprehensive Review

Research of Medical Image Registration Technique

Medical image registration using deep neural networks: A comprehensive review

Salient Feature Region: A New Method for Retinal Image Registration

Retinal Image Registration Based on Salient Feature Regions

RetinaRegNet: A Zero-Shot Approach for Retinal Image Registration

An Overview of Medical Image Registration

Multimodal image registration techniques: a comprehensive survey

Deep learning-based lung image registration: A review