SACB-Net: Spatial-awareness Convolutions for Medical Image Registration

2025-03-25

AI Summary

Existing deep learning-based image registration methods struggle to model spatially non-uniform deformations, particularly failing to capture non-local spatial variations in feature maps. To address this, we propose a spatially aware convolution mechanism centered on a novel 3D Spatially Aware Convolution Block (SACB): leveraging feature-similarity-driven regional clustering, SACB adaptively generates spatially variant convolutional kernels for fine-grained deformation modeling. We further design SACB-Net, a multi-scale pyramid flow estimation network capable of handling large deformations. Evaluated on the IXI and LPBA brain MRI datasets and an abdominal CT dataset, our method consistently outperforms state-of-the-art deep registration models in both deformation field accuracy—measured by target registration error—and anatomical consistency—assessed via landmark and structure overlap metrics. The implementation is publicly available.

Technology Category

• Computer Vision: 3D Computer Vision
• Machine Learning: Multimodal Learning
• Search and Optimization: Non-convex Optimization

Application Category

• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Web Mining and Content Analysis: Large pretrained models with web data

Abstract

Deep learning-based image registration methods have shown state-of-the-art performance and rapid inference speeds. Despite these advances, many existing approaches fall short in capturing spatially varying information in non-local regions of feature maps due to the reliance on spatially-shared convolution kernels. This limitation leads to suboptimal estimation of deformation fields. In this paper, we propose a 3D Spatial-Awareness Convolution Block (SACB) to enhance the spatial information within feature representations. Our SACB estimates the spatial clusters within feature maps by leveraging feature similarity and subsequently parameterizes the adaptive convolution kernels across diverse regions. This adaptive mechanism generates the convolution kernels (weights and biases) tailored to spatial variations, thereby enabling the network to effectively capture spatially varying information. Building on SACB, we introduce a pyramid flow estimator (named SACB-Net) that integrates SACBs to facilitate multi-scale flow composition, particularly addressing large deformations. Experimental results on the brain IXI and LPBA datasets as well as Abdomen CT datasets demonstrate the effectiveness of SACB and the superiority of SACB-Net over the state-of-the-art learning-based registration methods. The code is available at https://github.com/x-xc/SACB_Net .

Problem

• Improving medical image registration with spatial-awareness convolutions
• Addressing suboptimal deformation fields from spatially-shared kernels
• Enhancing multi-scale flow estimation for large deformations

Innovation

• 3D Spatial-Awareness Convolution Block (SACB)
• Adaptive convolution kernels for spatial variations
• Pyramid flow estimator for multi-scale deformations