Adaptive Convolutional Neural Networks for Accelerating Magnetic Resonance Imaging Via K -Space Data Interpolation

Tianming Du,Honggang Zhang,Yuemeng Li,Stephen Pickup,Mark Rosen,Rong Zhou,Hee Kwon Song,Yong Fan

DOI: https://doi.org/10.1016/j.media.2021.102098

2021-01-01

Abstract:Deep learning in k-space has demonstrated great potential for image reconstruction from undersam-pled k-space data in fast magnetic resonance imaging (MRI). However, existing deep learning-based im-age reconstruction methods typically apply weight-sharing convolutional neural networks (CNNs) to k-space data without taking into consideration the k-space data's spatial frequency properties, leading to ineffective learning of the image reconstruction models. Moreover, complementary information of spa-tially adjacent slices is often ignored in existing deep learning methods. To overcome such limitations, we have developed a deep learning algorithm, referred to as adaptive convolutional neural networks for k-space data interpolation (ACNN-k-Space), which adopts a residual Encoder-Decoder network ar-chitecture to interpolate the undersampled k-space data by integrating spatially contiguous slices as multi-channel input, along with k-space data from multiple coils if available. The network is enhanced by self-attention layers to adaptively focus on k-space data at different spatial frequencies and chan-nels. We have evaluated our method on two public datasets and compared it with state-of-the-art ex -isting methods. Ablation studies and experimental results demonstrate that our method effectively re-constructs images from undersampled k-space data and achieves significantly better image reconstruc-tion performance than current state-of-the-art techniques. Source code of the method is available at https://gitlab.com/qgpmztmf/acnn- k-space . (c) 2021 Elsevier B.V. All rights reserved.