A Deep Cascade of Convolutional Neural Networks for MR Image Reconstruction

The acquisition of Magnetic Resonance Imaging (MRI) is inherently slow. Inspired by recent advances in deep learning, we propose a framework for reconstructing MR images from undersampled data using a deep cascade of convolutional neural networks to accelerate the data acquisition process. We show that for Cartesian undersampling of 2D cardiac MR images, the proposed method outperforms the state-of-the-art compressed sensing approaches, such as dictionary learning-based MRI (DLMRI) reconstruction, in terms of reconstruction error, perceptual quality and reconstruction speed for both 3-fold and 6-fold undersampling. Compared to DLMRI, the error produced by the method proposed is approximately twice as small, allowing to preserve anatomical structures more faithfully. Using our method, each image can be reconstructed in 23 ms, which is fast enough to enable real-time applications.

[1]  Jian Sun,et al.  Deep Residual Learning for Image Recognition , 2015, 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).

[2]  Pavan K. Turaga,et al.  ReconNet: Non-Iterative Reconstruction of Images from Compressively Sensed Measurements , 2016, 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).

[3]  Yoram Bresler,et al.  MR Image Reconstruction From Highly Undersampled k-Space Data by Dictionary Learning , 2011, IEEE Transactions on Medical Imaging.

[4]  Akira Hirose,et al.  Learning Algorithms in Complex-Valued Neural Networks using Wirtinger Calculus , 2013 .

[5]  Jimmy Ba,et al.  Adam: A Method for Stochastic Optimization , 2014, ICLR.

[6]  Daniel Rueckert,et al.  Dictionary Learning and Time Sparsity for Dynamic MR Data Reconstruction , 2014, IEEE Transactions on Medical Imaging.

[7]  Jian Sun,et al.  Delving Deep into Rectifiers: Surpassing Human-Level Performance on ImageNet Classification , 2015, 2015 IEEE International Conference on Computer Vision (ICCV).

[8]  David L Donoho,et al.  Compressed sensing , 2006, IEEE Transactions on Information Theory.

[9]  Thomas Brox,et al.  U-Net: Convolutional Networks for Biomedical Image Segmentation , 2015, MICCAI.

[10]  Dumitru Erhan,et al.  Going deeper with convolutions , 2014, 2015 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).

[11]  Jong Chul Ye,et al.  A novel k-space annihilating filter method for unification between compressed sensing and parallel MRI , 2015, 2015 IEEE 12th International Symposium on Biomedical Imaging (ISBI).

[12]  Nasser M. Nasrabadi,et al.  Pattern Recognition and Machine Learning , 2006, Technometrics.

[13]  T. Kurihara,et al.  Compressed sensing MRI using sparsity induced from adjacent slice similarity , 2015, 2015 International Conference on Sampling Theory and Applications (SampTA).

[14]  P. Boesiger,et al.  SENSE: Sensitivity encoding for fast MRI , 1999, Magnetic resonance in medicine.

[15]  Patrice Y. Simard,et al.  Best practices for convolutional neural networks applied to visual document analysis , 2003, Seventh International Conference on Document Analysis and Recognition, 2003. Proceedings..

[16]  Vladlen Koltun,et al.  Multi-Scale Context Aggregation by Dilated Convolutions , 2015, ICLR.

[17]  Won-Ki Jeong,et al.  Compressed sensing reconstruction of dynamic contrast enhanced MRI using GPU-accelerated convolutional sparse coding , 2016, 2016 IEEE 13th International Symposium on Biomedical Imaging (ISBI).

[18]  Jian Sun,et al.  Deep ADMM-Net for Compressive Sensing MRI , 2016, NIPS.

[19]  Mathews Jacob,et al.  Accelerated Dynamic MRI Exploiting Sparsity and Low-Rank Structure: k-t SLR , 2011, IEEE Transactions on Medical Imaging.

[20]  K. T. Block,et al.  Undersampled radial MRI with multiple coils. Iterative image reconstruction using a total variation constraint , 2007, Magnetic resonance in medicine.

[21]  Dong Liang,et al.  MR Image Reconstruction with Convolutional Characteristic Constraint (CoCCo) , 2015, IEEE Signal Processing Letters.

[22]  Xiaoou Tang,et al.  Image Super-Resolution Using Deep Convolutional Networks , 2014, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[23]  Daniel Rueckert,et al.  Real-Time Single Image and Video Super-Resolution Using an Efficient Sub-Pixel Convolutional Neural Network , 2016, 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).

[24]  Leslie Ying,et al.  Accelerating magnetic resonance imaging via deep learning , 2016, 2016 IEEE 13th International Symposium on Biomedical Imaging (ISBI).

[25]  Jeffrey A Fessler,et al.  On NUFFT-based gridding for non-Cartesian MRI. , 2007, Journal of magnetic resonance.

[26]  Junzhou Huang,et al.  Fast multi-contrast MRI reconstruction. , 2014, Magnetic resonance imaging.

[27]  Michael Elad,et al.  ESPIRiT—an eigenvalue approach to autocalibrating parallel MRI: Where SENSE meets GRAPPA , 2014, Magnetic resonance in medicine.

[28]  Geoffrey E. Hinton,et al.  Learning representations by back-propagating errors , 1986, Nature.

[29]  K. Scheffler,et al.  Analysis and compensation of eddy currents in balanced SSFP , 2005, Magnetic resonance in medicine.

[30]  M. Lustig,et al.  Compressed Sensing MRI , 2008, IEEE Signal Processing Magazine.

[31]  정홍,et al.  Improved k-t BLAST and k-t SENSE using FOCUSS , 2007 .

[32]  Volker Rasche,et al.  Continuous radial data acquisition for dynamic MRI , 1995, Magnetic resonance in medicine.

[33]  Jens Frahm,et al.  Magnetic resonance imaging in real time: Advances using radial FLASH , 2010, Journal of magnetic resonance imaging : JMRI.

[34]  L. Ying,et al.  Accelerating SENSE using compressed sensing , 2009, Magnetic resonance in medicine.

[35]  Michael Lustig,et al.  Faster Imaging with Randomly Perturbed, Undersampled Spirals and |L|_1 Reconstruction , 2004 .

[36]  Kaiming He,et al.  Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networks , 2015, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[37]  Jong Chul Ye,et al.  Improved k–t BLAST and k–t SENSE using FOCUSS , 2007, Physics in medicine and biology.

[38]  Daniel K Sodickson,et al.  Low‐rank plus sparse matrix decomposition for accelerated dynamic MRI with separation of background and dynamic components , 2015, Magnetic resonance in medicine.