Dual-Attention Recurrent Networks for Affine Registration of Neuroimaging Data
Dual-Attention Recurrent Networks for Affine Registration of Neuroimaging Data
复制标题
DOI：
10.1137/1.9781611976236.43
复制
发表时间：
2020
期刊：
影响因子：
--
通讯作者：
Xin Dai;Xiangnan Kong;Xinyue Liu;J. B. Lee;C. Moore
中科院分区：
文献类型：
--
作者：
Xin Dai;Xiangnan Kong;Xinyue Liu;J. B. Lee;C. Moore
神经成像数据通常在进一步分析和挖掘之前经历几个预处理步骤。仿射图像配准是图像预处理的重要任务之一。近年来，基于卷积神经网络的图像配准方法被提出。然而，由于CNN的高计算和内存要求，这些方法不能实时用于fMRI等大型神经成像数据。在本文中，我们提出了一种双注意力循环网络（DRN），它使用硬注意力机制来允许模型关注输入图像中较小但与任务相关的部分，从而降低计算和内存成本。此外，DRN自然地支持原始输入图像（例如，功能性MRI）和我们想要将其对准的图像（例如，解剖MRI），因此它可以应用于更难的配准任务，如fMRI配准和归一化。在两个不同数据集上的大量实验表明，与其他基于神经网络的方法相比，DRN显着降低了计算和内存成本，而不会牺牲图像配准的质量。
Neuroimaging data typically undergoes several preprocessing steps before further analysis and mining can be done. Affine image registration is one of the important tasks during preprocessing. Recently, several image registration methods which are based on Convolutional Neural Networks have been proposed. However, due to the high computational and memory requirements of CNNs, these methods cannot be used in real-time for large neuroimaging data like fMRI. In this paper, we propose a Dual-Attention Recurrent Network (DRN) which uses a hard attention mechanism to allow the model to focus on small, but task-relevant, parts of the input image – thus reducing computational and memory costs. Furthermore, DRN naturally supports inhomogeneity between the raw input image (e.g., functional MRI) and the image we want to align it to (e.g., anatomical MRI) so it can be applied to harder registration tasks such as fMRI coregistration and normalization. Extensive experiments on two different datasets demonstrate that DRN significantly reduces the computational and memory costs compared with other neural network-based methods without sacrificing the quality of image registration.