Harmonization in Magnetic Resonance Imaging: A Survey of Acquisition, Image-level, and Feature-level Methods

2025-07-22

AI Summary

Batch effects arising from cross-device and multi-site MRI acquisitions severely confound biological signal identification, undermining reproducibility in multicenter studies and degrading the generalizability of deep learning models. This paper systematically reviews methods for mitigating MRI image heterogeneity, uniquely integrating prospective acquisition protocols, retrospective image processing, feature disentanglement, and the traveling-subject paradigm. We comprehensively evaluate deep learning–based techniques—including image reconstruction, style transfer, adversarial training, and adaptive normalization—under a novel “biologically faithful correction” principle. We catalog widely used public datasets and standardized evaluation metrics. Our analysis identifies critical challenges in reproducibility validation, cross-modal generalization, and clinical deployment. Finally, we outline future research directions: interpretable correction frameworks, federation-compatible adaptation strategies, and real-world clinical validation.

Technology Category

• Computer Vision: Medical and Biological Imaging
• Machine Learning: Multimodal Learning
• Intelligent Robots: Multimodal Perception & Sensor Fusion

Application Category

• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• User Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking
• Graph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches for Web-related graphs

Abstract

Modern medical imaging technologies have greatly advanced neuroscience research and clinical diagnostics. However, imaging data collected across different scanners, acquisition protocols, or imaging sites often exhibit substantial heterogeneity, known as "batch effects" or "site effects". These non-biological sources of variability can obscure true biological signals, reduce reproducibility and statistical power, and severely impair the generalizability of learning-based models across datasets. Image harmonization aims to eliminate or mitigate such site-related biases while preserving meaningful biological information, thereby improving data comparability and consistency. This review provides a comprehensive overview of key concepts, methodological advances, publicly available datasets, current challenges, and future directions in the field of medical image harmonization, with a focus on magnetic resonance imaging (MRI). We systematically cover the full imaging pipeline, and categorize harmonization approaches into prospective acquisition and reconstruction strategies, retrospective image-level and feature-level methods, and traveling-subject-based techniques. Rather than providing an exhaustive survey, we focus on representative methods, with particular emphasis on deep learning-based approaches. Finally, we summarize the major challenges that remain and outline promising avenues for future research.

Problem

• Reducing heterogeneity in MRI data across scanners and protocols
• Preserving biological signals while removing site-related biases
• Improving reproducibility and generalizability of learning-based models

Innovation

• MRI harmonization via acquisition and reconstruction strategies
• Deep learning-based image-level and feature-level methods
• Traveling-subject techniques for bias reduction