Symmetric transformer-based network for unsupervised image registration

摘要

Medical image registration is a fundamental and critical task in medical image analysis. With the rapid development of deep learning, convolutional neural networks (CNNs) have dominated the medical image registration field. Due to the disadvantage of the local receptive field of CNNs, some recent registration methods have focused on using transformers for nonlocal registration. However, the standard transformer has a vast number of parameters and high computational complexity, which means that it can only be applied at the bottom of registration models. As a result, only coarse information is available at the lowest resolution, limiting the contribution of the transformer in these models. To address these challenges, we propose a convolution-based efficient multihead self-attention (CEMSA) block, which reduces the number of parameters of the traditional transformer and captures local spatial context information to reduce semantic ambiguity in the attention mechanism. Based on the proposed CEMSA, we present a novel symmetric transformer-based model (SymTrans). SymTrans employs the transformer blocks in the encoder and the decoder to model the long-range spatial cross-image relevance. We apply SymTrans to the displacement field and diffeomorphic registration. Experimental results show that our proposed method achieves state-of-the-art performance in image registration. Our code is publicly available at https://github.com/MingR-Ma/SymTrans.