Geometric Approaches to Increase the Expressivity of Deep Neural Networks for MR Reconstruction
This work addresses the challenge of optimizing deep neural network architectures for MRI reconstruction, which is incremental as it builds on existing geometric interpretations to enhance expressivity.
The paper tackled the problem of balancing network complexity and performance in deep learning for accelerated MRI reconstruction by proposing a geometric approach using bootstrapping and subnetwork aggregation with an attention module to increase neural network expressivity, resulting in significantly improved reconstruction performance with negligible complexity increases.
Recently, deep learning approaches have been extensively investigated to reconstruct images from accelerated magnetic resonance image (MRI) acquisition. Although these approaches provide significant performance gain compared to compressed sensing MRI (CS-MRI), it is not clear how to choose a suitable network architecture to balance the trade-off between network complexity and performance. Recently, it was shown that an encoder-decoder convolutional neural network (CNN) can be interpreted as a piecewise linear basis-like representation, whose specific representation is determined by the ReLU activation patterns for a given input image. Thus, the expressivity or the representation power is determined by the number of piecewise linear regions. As an extension of this geometric understanding, this paper proposes a systematic geometric approach using bootstrapping and subnetwork aggregation using an attention module to increase the expressivity of the underlying neural network. Our method can be implemented in both k-space domain and image domain that can be trained in an end-to-end manner. Experimental results show that the proposed schemes significantly improve reconstruction performance with negligible complexity increases.