Search-MIND: Training-Free Multi-Modal Medical Image Registration

2026-04-10

AI Summary

This work addresses the challenges of multimodal medical image registration, including nonlinear intensity discrepancies, susceptibility to local optima, and poor generalization of deep learning models. The authors propose a training-free, end-to-end iterative optimization framework that follows a coarse-to-fine strategy: first performing hierarchical coarse alignment, followed by refined deformation using a deformable transformation. Key innovations include a variance-weighted mutual information (VWMI) loss that emphasizes high-information regions, and a Search-MIND (S-MIND) loss that extends the convergence basin of structural descriptors, substantially enhancing robustness. Evaluated on the CARE Liver 2025 and CHAOS Challenge datasets, the method outperforms conventional algorithms such as ANTs and foundation-model-based approaches like DINO-reg, demonstrating superior accuracy and stability across diverse imaging modalities.

Technology Category

• Computer Vision: Multi-modal Vision
• Machine Learning: Multimodal Learning
• Search and Optimization: Learning to Search

Application Category

• Search and Retrieval-Augmented AI: Retrieval-Augmented Generation (RAG) and multi-modal RAG
• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs
• Systems and Infrastructure for Web, Mobile and WoT: Experiences and lessons learnt from Web-based algorithms and system deployments

Abstract

Multi-modal image registration plays a critical role in precision medicine but faces challenges from non-linear intensity relationships and local optima. While deep learning models enable rapid inference, they often suffer from generalization collapse on unseen modalities. To address this, we propose Search-MIND, a training-free, iterative optimization framework for instance-specific registration. Our pipeline utilizes a coarse-to-fine strategy: a hierarchical coarse alignment stage followed by deformable refinement. We introduce two novel loss functions: Variance-Weighted Mutual Information (VWMI), which prioritizes informative tissue regions to shield global alignment from background noise and uniform regions, and Search-MIND (S-MIND), which broadens the convergence basin of structural descriptors by considering larger local search range. Evaluations on CARE Liver 2025 and CHAOS Challenge datasets show that Search-MIND consistently outperforms classical baselines like ANTs and foundation model-based approaches like DINO-reg, offering superior stability across diverse modalities.

Problem

• multi-modal image registration
• non-linear intensity relationships
• local optima
• generalization collapse

Innovation

• training-free registration
• multi-modal image registration
• Variance-Weighted Mutual Information
• coarse-to-fine optimization
• Search-MIND