Complex Fully Convolutional Neural Networks for MR Image Reconstruction
This work addresses faster and more accurate MRI reconstruction for medical imaging, representing an incremental improvement by adapting existing deep learning techniques to complex-valued data.
The paper tackles the problem of reconstructing MRI images from undersampled k-space data by proposing a complex-valued fully convolutional neural network (CDFNet) that directly processes complex inputs, resulting in improved perceptual quality and anatomical structure recovery compared to real-valued methods.
Undersampling the k-space data is widely adopted for acceleration of Magnetic Resonance Imaging (MRI). Current deep learning based approaches for supervised learning of MRI image reconstruction employ real-valued operations and representations by treating complex valued k-space/spatial-space as real values. In this paper, we propose complex dense fully convolutional neural network ($\mathbb{C}$DFNet) for learning to de-alias the reconstruction artifacts within undersampled MRI images. We fashioned a densely-connected fully convolutional block tailored for complex-valued inputs by introducing dedicated layers such as complex convolution, batch normalization, non-linearities etc. $\mathbb{C}$DFNet leverages the inherently complex-valued nature of input k-space and learns richer representations. We demonstrate improved perceptual quality and recovery of anatomical structures through $\mathbb{C}$DFNet in contrast to its real-valued counterparts.