ShiftMorph: A Fast and Robust Convolutional Neural Network for 3D Deformable Medical Image Registration

Lijian Yang, Weisheng Li, Yucheng Shu, Jianxun Mi, Yuping Huang, Bin Xiao

Abstract: Deformable image registration (DIR) is crucial for many medical image applications. In recent years, learning-based methods utilizing the convolutional neural network (CNN) or the Transformer have demonstrated their superiority in image registration, dominating a new era for DIR. However, very few of these methods can satisfy the demands of real-time applications due to the high spatial resolution of 3D volumes and the high complexity of 3D operators. To tackle this, we propose losslessly downsampling by shifting the strided convolution. A grouping strategy is then used to reduce redundant computations and support self-consistency learning. As an inherent regularizer of the network design, self-consistency learning improves the deformation quality and enables halving the proposed network after training. Furthermore, the proposed shifted connection converts the decoding operations into a lower-dimensional space, significantly reducing decoding overhead. Extensive experimental results on medical image registration demonstrate that our method is competitive with state-of-the-art methods in terms of registration performance, and additionally, it achieves over $3\times$ the speed of most of them.

Primary Subject Area: [Experience] Multimedia Applications

Secondary Subject Area: [Content] Vision and Language

Relevance To Conference: Image registration, a fundamental technique in biomedical imaging, plays a critical role in a wide range of applications involving the integration of multimedia data. It is widely used in areas such as surgical guidance, histological imaging, and neurosurgery to align and fuse multiple modalities of images for enhanced visualization and analysis. In surgical navigation systems, image registration is particularly valued for its ability to seamlessly integrate different imaging modalities, ultimately providing surgeons with high-quality intra-operative images that support accurate and efficient decision-making during procedures. However, most current deformable image registration methods cannot meet the requirements of real-time applications. This paper presents a fast and robust CNN method for medical image registration. Extensive experimental results on brain MRI and lung CT images demonstrate the effectiveness of our method across multiple modalities.

Submission Number: 541