South-Center Minzu University,a. College of Electronic and Information Engineering; b. Hubei Key Lab of Intelligent Wireless Communication; c. College of Computer Science,Wuhan 430074,China
Transformer has shown great potential to improve image recovery performance. A reconstruction method of the accelerated magnetic resonance image (MRI) based on a wavelet-decomposed U-shaped Transformer network is studied. The core unit of the reconstruction network is designed based on combining the Swin Transformer with the Unet. By fusing multi-scale features of images, the learning ability of the network is improved and better reconstruction performance is obtained. By using wavelet transform to decompose the input image, the input feature dimension of Swin Transformer is reduced, so as to effectively reduce the computational complexity of the reconstruction network. The wavelet domain loss is adopted to constrain network training, with better recovering the structure and texture information of the image. The experimental results on Calgary-Campinas brain MR Dataset verify the effectiveness of the proposed method in improving the quality of reconstructed images and balancing the complexity of system.
BLAIMERM, BREUERF, MUELLERM, et al. SMASH, SENSE, PILS, GRAPPA: How to choose the optimal method[J]. Topics in Magnetic Resonance Imaging, 2004, 15(4): 223-236.
[2]
LUSTIGM, PAULYJ M. SPIRiT: Iterative self‐consistent parallel imaging reconstruction from arbitrary k‐space[J]. Magnetic Resonance in Medicine, 2010, 64(2): 457-471.
[3]
PRUESSMANNK P, WEIGERM, BORNERTP, et al. Advances in sensitivity encoding with arbitrary k-Space trajectories[J]. Magnetic Resonance in Medicine, 2001, 46: 638-651.
[4]
GRISWOLDM A, JAKOBP M, HEIDEMANNR M, et al. Generalized autocalibrating partially parallel acquisitions (GRAPPA)[J]. Magnetic Resonance in Medicine, 2002, 47(6): 1202-1210.
[5]
LUSTIGM, DONOHOD L, SANTOSJ M, et al. Compressed sensing MRI[J]. IEEE Signal Processing Magazine, 2008, 25(2): 72-82.
[6]
LIANGD, LIUB, WANGJ, et al. Accelerating SENSE using compressed sensing[J]. Magnetic Resonance in Medicine, 2009, 62(6): 1574-1584.
[7]
LUSTIGM, DONOHOD, PAULYJ M. Sparse MRI: The application of compressed sensing for rapid MR imaging[J]. Magnetic Resonance in Medicine, 2007, 58(6): 1182-1195.
[8]
HONGM, YUY, WANGH, et al. Compressed sensing MRI with singular value decomposition-based sparsity basis[J]. Physics in Medicine & Biology, 2011, 56(19): 6311.
[9]
WANGS, XIAOT, LIUQ, et al. Deep learning for fast MR imaging: A review for learning reconstruction from incomplete k-space data[J]. Biomedical Signal Processing and Control, 2021, 68: 102579.
[10]
RONNEBERGERO, FISCHERP, BROXT. U-net: Convolutional networks for biomedical image segmentation[C]//International Conference on Medical image computing and computer-assisted intervention. Switzerland: Springer, 2015: 234-241.
[11]
SOUZAR, BENTOM, NOGOVITSYNN, et al. Dual-domain cascade of U-nets for multi-channel magnetic resonance image reconstruction[J]. Magnetic resonance imaging, 2020, 71: 140-153.
[12]
YANGG, YUS, DONGH, et al. DAGAN: Deep de-aliasing generative adversarial networks for fast compressed sensing MRI reconstruction[J]. IEEE transactions on medical imaging, 2017, 37(6): 1310-1321.
[13]
DOSOVITSKIYA, BeyerL, KOLESNIKOVA, et al. An image is worth 16×16 words: Transformers for image recognition at scale[C]//International Conference on Learning Representations. Vienna: ICLR, 2021: 1-14.
[14]
SHAMSHADF, KHANS, ZAMIRS W, et al. Transformers in medical imaging: A survey[J]. arXiv:2022, 2201.09873.
[15]
HUANGJ, FANGY, WUY, et al. Swin transformer for fast MRI[J]. Neurocomputing, 2022, 493: 281-304.
[16]
LIUZ, LINY, CAOY, et al. Swin transformer: Hierarchical vision transformer using shifted windows[C]// International Conference on Computer Vision. Montreal: IEEE, 2021: 10012-10022.
[17]
FABIANZ, TINAZB, SOLTANOLKOTABIM. HUMUS-net: Hybrid unrolled multi-scale network architecture for accelerated MRI reconstruction[C]//Neural Information Processing System. SanDiego: NeurIPS, 2022.
[18]
QUX, GUOD, NINGB, et al. Undersampled MRI reconstruction with patch-based directional wavelets[J]. Magnetic Resonance Imaging, 2012, 30(7): 964-977.
[19]
BLOCKK T, UECKERM, FRAHMJ. Undersampled radial MRI with multiple coils. Iterative image reconstruction using a total variation constraint[J]. Magnetic Resonance in Medicine, 2007, 57(6): 1086-1098.
[20]
HANY, SUMWOOL, YeC J. k-Space deep learning for accelerated MRI[J]. IEEE Transactions on Medical Imaging, 2020, 39(2): 377-386.
[21]
RANM, XIAW, HUANGY, et al. MD-Recon-Net: A parallel dual-domain convolutional neural network for compressed sensing MRI[J]. IEEE Transactions on Radiation and Plasma Medical Sciences, 2021, 5(1): 120-135.
[22]
LEED, YOOJ, TAK S, et al. Deep residual learning for accelerated MRI using magnitude and phase networks[J]. IEEE Transactions on Biomedical Engineering, 2018, 65(9): 1985-1995.
[23]
HEK, ZHANGX, RENS, et al. Deep residual learning for image recognition[C]//IEEE Conference on Computer Vision and Pattern Recognition. Las Vegas: IEEE, 2016: 770-778.
[24]
LIANGJ, CAOJ, SUNG, et al. Swinir: Image restoration using swin transformer[C]//International Conference on Computer Vision. Montreal: IEEE, 2021: 1833-1844.
[25]
TOUBRONH, CordM, SABLAYROLLESA, et al. Going deeper with image transformers[C]//International Conference on Computer Vision. Montreal: IEEE, 2021: 32-42.
[26]
LAIW S, HUANGJ B, AHUJAN, et al. Fast and accurate image super-resolution with deep laplacian pyramid networks[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2018, 41(11): 2599-2613.
[27]
SOUZAR, LUCENAO, GARRAFAJ, et al. An open, multi-vendor, multi-field-strength brain MR dataset and analysis of publicly available skull stripping methods agreement[J]. NeuroImage, 2018, 170: 482-494.