Split-slice training and hyperparameter tuning of RAKI networks for simultaneous multi-slice reconstruction.
RAKI
deep learning
hyperparameters
image reconstruction
simultaneous multi-slice
training augmentation
Journal
Magnetic resonance in medicine
ISSN: 1522-2594
Titre abrégé: Magn Reson Med
Pays: United States
ID NLM: 8505245
Informations de publication
Date de publication:
06 2021
06 2021
Historique:
received:
09
06
2020
revised:
11
11
2020
accepted:
12
11
2020
pubmed:
18
12
2020
medline:
21
5
2021
entrez:
17
12
2020
Statut:
ppublish
Résumé
Simultaneous multi-slice acquisitions are essential for modern neuroimaging research, enabling high temporal resolution functional and high-resolution q-space sampling diffusion acquisitions. Recently, deep learning reconstruction techniques have been introduced for unaliasing these accelerated acquisitions, and robust artificial-neural-networks for k-space interpolation (RAKI) have shown promising capabilities. This study systematically examines the impacts of hyperparameter selections for RAKI networks, and introduces a novel technique for training data generation which is analogous to the split-slice formalism used in slice-GRAPPA. RAKI networks were developed with variable hyperparameters and with and without split-slice training data generation. Each network was trained and applied to five different datasets including acquisitions harmonized with Human Connectome Project lifespan protocol. Unaliasing performance was assessed through L1 errors computed between unaliased and calibration frequency-space data. Split-slice training significantly improved network performance in nearly all hyperparameter configurations. Best unaliasing results were achieved with three layer RAKI networks using at least 64 convolutional filters with receptive fields of 7 voxels, 128 single-voxel filters in the penultimate RAKI layer, batch normalization, and no training dropout with the split-slice augmented training dataset. Networks trained without the split-slice technique showed symptoms of network over-fitting. Split-slice training for simultaneous multi-slice RAKI networks positively impacts network performance. Hyperparameter tuning of such reconstruction networks can lead to further improvements in unaliasing performance.
Types de publication
Journal Article
Langues
eng
Sous-ensembles de citation
IM
Pagination
3272-3280Informations de copyright
© 2020 International Society for Magnetic Resonance in Medicine.
Références
Conturo TE, Lori NF, Cull TS, et al. Tracking neuronal fiber pathways in the living human brain. Proc Nat Acad Sci USA. 1999;96:10422-10427.
Ogawa S, Lee TM, Kay AR, Tank DW. Brain magnetic resonance imaging with contrast dependent on blood oxygenation. Proc Nat Acad Sci USA. 1990;87:9868-9872.
Mansfield P. Multi-planar image formation using NMR spin echoes. J Phys C. 1977;10:L55-L58.
Alexander AL, Lee JE, Lazar M, Field AS. Diffusion tensor imaging of the brain. Neurotherapeutics. 2007;4:316-329.
Lu H, Jensen JH, Ramani A, Helpern JA. Three-dimensional characterization of non-gaussian water diffusion in humans using diffusion kurtosis imaging. NMR Biomed. 2006;19:236-247.
Zhang H, Schneider T, Wheeler-Kingshott CA, Alexander DC. NODDI: practical in vivo neurite orientation dispersion and density imaging of the human brain. NeuroImage. 2012;61:1000-1016.
Larkman DJ, Hajnal JV, Herlihy AH, Coutts GA, Young IR, Ehnholm G. Use of multicoil arrays for separation of signal from multiple slices simultaneously excited. J Magn Reson Imaging. 2001;13:313-317.
Moeller S, Yacoub E, Olman CA, et al. Multiband multislice GE-EPI at 7 Tesla, with 16-fold acceleration using partial parallel imaging with application to high spatial and temporal whole-brain FMRI. Magn Reson Med. 2010;63:1144-1153.
Jesmanowicz A, Nencka AS, Li S-J, Hyde JS. Two-axis acceleration of functional connectivity magnetic resonance imaging by parallel excitation of phase-tagged slices and half k-space acceleration. Brain Connect. 2011;1:81-90.
Feinberg DA, Moeller S, Smith SM, et al. Multiplexed echo planar imaging for sub-second whole brain FMRI and fast diffusion imaging. PLoS ONE. 2010;5:e15710.
Blaimer M, Choli M, Jakob PM, Griswold MA, Breuer FA. Multiband phase-constrained parallel MRI. Magn Reson Med. 2013;69:974-980.
Setsompop K, Gagoski BA, Polimeni JR, Witzel T, Wedeen VJ, Wald LL. Blipped-controlled aliasing in parallel imaging for simultaneous multislice echo planar imaging with reduced g-factor penalty. Magn Reson Med. 2012;67:1210-1224.
Cauley SF, Polimeni JR, Bhat H, Wald LL, Setsompop K. Interslice leakage artifact reduction technique for simultaneous multislice acquisitions. Magn Reson Med. 2014;72:93-102.
Moeller S, Vu A, Auerbach E, Ugurbil K, Yacoub E. RO extended FOV SENSE/GRAPPA for multiband imaging with FOV shift. In Proceedings of the 22nd Scientific Meeting of ISMRM, Milan, Italy, 2014; 4396.
Krizhevsky A, Sutskever I, Hinton GE. ImageNet classification with deep convolutional neural networks. Technical Report. 2012. http://code.google.com/p/cuda-convnet/
Hammernik K, Klatzer T, Kobler E, Recht MP, Sodickson DK, Pock T, Knoll F. Learning a variational network for reconstruction of accelerated MRI data. Magn Reson Med. 2018;79:3055-3071.
Han Y, Yoo J, Kim HH, Shin HJ, Sung K, Ye JC. Deep learning with domain adaptation for accelerated projection-reconstruction MR. Magn Reson Med. 2018;80:1189-1205.
Hyun CM, Kim HP, Lee SM, Lee S, Seo JK. Deep learning for undersampled MRI reconstruction. Phys Med Biol. 2018;63:135007.
Aggarwal HK, Mani MP, Jacob M. MoDL: model based deep learning architecture for inverse problems. IEEE Trans Med Imaging. 2017;38:394-405.
Biswas S, Aggarwal HK, Poddar S, Jacob M. Model-based free-breathing cardiac MRI reconstruction using deep learned STORM priors: MoDL-STORM. ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing-Proceedings. Alberta, Canada. 2018;7:6533-6537. http://arxiv.org/abs/1807.03845
Schlemper J, Caballero J, Hajnal JV, Price A, Rueckert D. A deep cascade of convolutional neural networks for MR image reconstruction. In: Niethammer M, et al. eds. Information Processing in Medical Imaging. IPMI 2017. Lecture Notes in Computer Science. Springer, Cham2017:10265. http://doi-org-443.webvpn.fjmu.edu.cn/10.1007/978-3-319-59050-9_51
Zhu B, Liu JZ, Cauley SF, Rosen BR, Rosen MS. Image reconstruction by domain-transform manifold learning. Nature. 2018;555:487-492.
Knoll F, Hammernik K, Zhang C, et al. Deep-learning methods for parallel magnetic resonance imaging reconstruction: A survey of the current approaches, trends, and issues. IEEE Signal Processing Magazine. 2020;37:128-140
Akcakaya M, Moeller S, Weingartner S, Ugurbil K. Scan-specific robust artificial-neural-networks for k-space interpolation (RAKI) reconstruction: database-free deep learning for fast imaging. Magn Reson Med. 2019;81:439-453.
Moeller S, Kumar PP, Andersson J, et al. Diffusion imaging in the post HCP era. J Magn Reson Imaging. n/a. https://onlinelibrary.wiley.com/doi/abs/10.1002/jmri.27247
Zhang C, Moeller S, Weingartner S, Ugurbil K, Akcakaya M. Accelerated simultaneous multi-slice MRI using subject-specific convolutional neural networks. In 52nd Asilomar Conference on Signals, Systems, and Computers. Asilomar, CA: IEEE; 2018;10:1636-1640.
Zhang C, Hosseini SAH, Weingartner S, Ugurbil K, Moeller S, Akcakaya M. Optimized fast GPU implementation of robust artificial-neural-networks for k-space interpolation (raki) reconstruction. PLoS ONE. 2019;14:e0223315.
Nair V, Hinton GE. Rectified linear units improve restricted Boltzmann machines. In Proceedings of the 27th International Conference on Machine Learning. Haifa, Isreal; 2010:432.
Paszke A, Gross S, Chintala S, et al. Automatic differentiation in PyTorch. Technical Report. 2017.
RakiNetwork/README.md at master . NenckaLab/RakiNetwork. https://github.com/NenckaLab/RakiNetwork/blob/master/README.md
Kingma DP, Ba JL. Adam: a method for stochastic optimization. In 3rd International Conference on Learning Representations, ICLR 2015. San Diego, CA: ICLR; 2015:12.
Griswold MA, Jakob PM, Heidemann RM, et al. Generalized autocalibrating partially parallel acquisitions (GRAPPA). Magn Reson Med. 2002;47:1202-1210.
Geron A. Hands-on machine learning with Scikit-Learn and TensorFlow: concepts, tools, and techniques to build intelligent systems. Sebastopol, CA: O'Reilly Media.
Koch KM, Meier TB, Karr R, Nencka AS, Muftuler LT, McCrea M. Quantitative susceptibility mapping after sports-related concussion. Am J Neuroradiol. 2018;39:1215-1221.
Smith SM, Beckmann CF, Andersson J, et al. Resting-state fMRI in the Human Connectome Project. NeuroImage. 2013;80:144-168.
Gonzalez-Castillo J, Roopchansingh V, Bandettini PA, Bodurka J. Physiological noise effects on the flip angle selection in BOLD fMRI. Neuroimage. 2011;54:2764-2778.
Ferrazzi G, Bassenge JP, Wink C, et al. Autocalibrated multiband CAIPIRINHA with through-time encoding: proof of principle and application to cardiac tissue phase mapping. Magn Reson Med. 2019;81:1016-1030.
Wilcoxon F. Individual comparisons by ranking methods. Biomet Bull. 1945;1:80.
Nakarmi U, Cheng JY, Mardani M, Pauly JM, Vasanawala SS. Exploiting multi-scale image features for accelerating magnetic resonance imaging. ISMRM Workshop on Data Samping and Image Reconstruction. Sedona, AZ. 2020.
Muftuler LT, Arpinar VE, Koch K, et al. Optimization of hyperparameters for SMS reconstruction.Magnetic Resonance Imaging 2020; 73:91-103.
Ioffe S, Szegedy C. Batch normalization: Accelerating deep network training by reducing internal covariate shift. In 32nd International Conference on Machine Learning, ICML 2015. Lille, France. 2015;1:448-456.
Metz L, Poole B, Pfau D, Sohl-Dickstein J. Unrolled generative adversarial networks. arXiv, page preprint arXiv:1611.02163, 2016.
Cox RW. AFNI: software for analysis and visualization of functional magnetic resonance neuroimages. Comput Biomed Res. 1996;29:162-173.