Biomechanical Constraints Assimilation in Deep-Learning Image Registration: Application to sliding and locally rigid deformations

2025-04-07

AI Summary

Traditional regularization strategies in medical image registration impose global, uniform constraints that fail to capture tissue heterogeneity-induced non-uniform deformations—particularly violating biomechanical priors in low-contrast regions. To address this, we propose the first spatially adaptive solid-mechanics-based regularization framework for deep registration, embedding tissue-specific physical constraints—including local rigidity, interfacial sliding, and pseudo-elasticity—directly into CNN/U-Net architectures. Our method employs a learnable, physics-guided loss to enable anatomy-aware deformation estimation. Evaluated end-to-end on synthetic and real 3D thoracoabdominal datasets, it reduces deformation error in hard tissues by 32%, improves interfacial sliding localization accuracy by 27%, and significantly enhances both biomechanical plausibility and cross-domain generalizability of predicted deformations. The implementation is publicly available.

Technology Category

• Computer Vision: Low Level & Physics-based Vision
• Machine Learning: Adversarial Learning & Robustness
• Search and Optimization: Sampling/Simulation-based Search

Application Category

• User Modeling, Personalization and Recommendation: On-Device user modeling, personalization, and recommendation
• Graph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches for Web-related graphs
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking

Abstract

Regularization strategies in medical image registration often take a one-size-fits-all approach by imposing uniform constraints across the entire image domain. Yet biological structures are anything but regular. Lacking structural awareness, these strategies may fail to consider a panoply of spatially inhomogeneous deformation properties, which would faithfully account for the biomechanics of soft and hard tissues, especially in poorly contrasted structures. To bridge this gap, we propose a learning-based image registration approach in which the inferred deformation properties can locally adapt themselves to trained biomechanical characteristics. Specifically, we first enforce in the training process local rigid displacements, shearing motions or pseudo-elastic deformations using regularization losses inspired from the field of solid-mechanics. We then show on synthetic and real 3D thoracic and abdominal images that these mechanical properties of different nature are well generalized when inferring the deformations between new image pairs. Our approach enables neural-networks to infer tissue-specific deformation patterns directly from input images, ensuring mechanically plausible motion. These networks preserve rigidity within hard tissues while allowing controlled sliding in regions where tissues naturally separate, more faithfully capturing physiological motion. The code is publicly available at https://github.com/Kheil-Z/biomechanical_DLIR .

Problem

• Adapting deformation properties to biomechanical characteristics locally
• Ensuring mechanically plausible motion in image registration
• Preserving rigidity in hard tissues while allowing sliding

Innovation

• Adaptive local deformation properties learning
• Biomechanical regularization losses training
• Tissue-specific motion inference networks