Brain big data processing with massively parallel computing technology: challenges and opportunities

Brain data processing has been embracing the big data era driven by the rapid advances of neuroscience as well as the experimental techniques for recording neuronal activities. Processing of massive brain data has become a constant in neuroscience research and practice, which is vital in revealing the hidden information to better understand the brain functions and malfunctions. Brain data are routinely non‐linear and non‐stationary in nature, and existing algorithms and approaches to neural data processing are generally complicated in order to characterize the non‐linearity and non‐stationarity. Brain big data processing has pressing needs for appropriate computing technologies to address three grand challenges: (1) efficiency, (2) scalability and (3) reliability. Recent advances of computing technologies are making non‐linear methods viable in sophisticated applications of massive brain data processing. General‐purpose Computing on the Graphics Processing Unit (GPGPU) technology fosters an ideal environment for this purpose, which benefits from the tremendous computing power of modern graphics processing units in massively parallel architecture that is frequently an order of magnitude larger than the modern multi‐core CPUs. This article first recaps significant speed‐ups of existing algorithms aided by GPGPU in neuroimaging and processing electroencephalogram (EEG), functional magnetic resonance imaging (fMRI), magnetoencephalography (MEG) and etc. The article then demonstrates a series of successful approaches to processing EEG data in various dimensions and scales in a massively parallel manner: (1) decomposition: a massively parallel Ensemble Local Mean Decomposition (ELMD) algorithm aided by GPGPU can decompose EEG series, which forms the basis of further time‐frequency transformation, in real‐time without sacrificing the precision of processing; (2) synchronization measurement: a parallelized Nonlinear Interdependence (NLI) method for global synchronization measurement of multivariate EEG with speed‐up of more than 1000 times, and it was successful in localization of epileptic focus; and (3) dimensionality reduction: a large‐scale Parallel Factor Analysis which excels in run‐time performance and scales far better by hundreds of times than conventional approach does, and it supports fast factorization of EEG with more than 1000 channels. Through these practices, the massively parallel computing technology manifests great potentials in addressing the grand challenges of brain big data processing. Copyright © 2016 John Wiley & Sons, Ltd.

[1]  Rajiv Ranjan,et al.  IK-SVD: Dictionary Learning for Spatial Big Data via Incremental Atom Update , 2014, Computing in Science & Engineering.

[2]  Jun Zhao,et al.  Accelerating the reconstruction of magnetic resonance imaging by three-dimensional dual-dictionary learning using CUDA , 2014, 2014 36th Annual International Conference of the IEEE Engineering in Medicine and Biology Society.

[3]  R. Ilmoniemi,et al.  Magnetoencephalography-theory, instrumentation, and applications to noninvasive studies of the working human brain , 1993 .

[4]  Rajesh P. N. Rao,et al.  Short-time windowed covariance: A metric for identifying non-stationary, event-related covariant cortical sites , 2014, Journal of Neuroscience Methods.

[5]  Lizhe Wang,et al.  Fast and Scalable Multi-Way Analysis of Massive Neural Data , 2015, IEEE Transactions on Computers.

[6]  Christopher Nimsky,et al.  Hybrid Visualization for White Matter Tracts using Triangle Strips and Point Sprites , 2006, IEEE Transactions on Visualization and Computer Graphics.

[7]  Albert Y. Zomaya,et al.  Recent advances in autonomic provisioning of big data applications on clouds , 2015, IEEE Trans. Cloud Comput..

[8]  Adelino R. Ferreira da Silva,et al.  A Bayesian multilevel model for fMRI data analysis , 2011, Comput. Methods Programs Biomed..

[9]  Byron M. Yu,et al.  Dimensionality reduction for large-scale neural recordings , 2014, Nature Neuroscience.

[10]  Lizhe Wang,et al.  Global Synchronization Measurement of Multivariate Neural Signals with Massively Parallel Nonlinear Interdependence Analysis , 2014, IEEE Transactions on Neural Systems and Rehabilitation Engineering.

[11]  Hüseyin Gürüler,et al.  Rapid Automated Classification of Anesthetic Depth Levels using GPU Based Parallelization of Neural Networks , 2015, Journal of medical systems.

[12]  Chao Yang,et al.  Ultra-Scalable CPU-MIC Acceleration of Mesoscale Atmospheric Modeling on Tianhe-2 , 2015, IEEE Transactions on Computers.

[13]  Achim Streit,et al.  Enabling collaborative MapReduce on the Cloud with a single-sign-on mechanism , 2014, Computing.

[14]  Mustafa Coskun,et al.  Determining the Appropriate Amount of Anesthetic Gas Using DWT and EMD Combined with Neural Network , 2014, Journal of Medical Systems.

[15]  Rüdiger Westermann,et al.  The application of GPU particle tracing to diffusion tensor field visualization , 2005, VIS 05. IEEE Visualization, 2005..

[16]  Tim McGraw,et al.  Stochastic DT-MRI Connectivity Mapping on the GPU , 2007, IEEE Transactions on Visualization and Computer Graphics.

[17]  Markus Gipp,et al.  Correlation analysis on GPU systems using NVIDIA’s CUDA , 2011, Journal of Real-Time Image Processing.

[18]  Norden E. Huang,et al.  Ensemble Empirical Mode Decomposition: a Noise-Assisted Data Analysis Method , 2009, Adv. Data Sci. Adapt. Anal..

[19]  Michael Garland,et al.  Understanding throughput-oriented architectures , 2010, Commun. ACM.

[20]  H. Kudo,et al.  GPU-Based PET Image Reconstruction Using an Accurate Geometrical System Model , 2012, IEEE Transactions on Nuclear Science.

[21]  Frank Lindseth,et al.  Medical image segmentation on GPUs - A comprehensive review , 2015, Medical Image Anal..

[22]  Michael Lees,et al.  Design and Evaluation of a Data-Driven Scenario Generation Framework for Game-Based Training , 2017, IEEE Transactions on Computational Intelligence and AI in Games.

[23]  Joshua R. Smith,et al.  The local mean decomposition and its application to EEG perception data , 2005, Journal of The Royal Society Interface.

[24]  Rasmus Bro,et al.  Improving the speed of multi-way algorithms:: Part I. Tucker3 , 1998 .

[25]  Stephen D. Laycock,et al.  GPU Accelerated Generation of Digitally Reconstructed Radiographs for 2-D/3-D Image Registration , 2012, IEEE Transactions on Biomedical Engineering.

[26]  Jinjun Chen,et al.  A security framework in G-Hadoop for big data computing across distributed Cloud data centres , 2014, J. Comput. Syst. Sci..

[27]  Wei Xue,et al.  A case study of large-scale parallel I/O analysis and optimization for numerical weather prediction system , 2014, Future Gener. Comput. Syst..

[28]  D. Rugar,et al.  Nuclear magnetic resonance imaging with 90-nm resolution. , 2007, Nature Nanotechnology.

[29]  Saeid Sanei,et al.  EEG signal processing , 2000, Clinical Neurophysiology.

[30]  Tao Yuan,et al.  Parallel Processing of Massive Remote Sensing Images in a GPU Architecture , 2014, Comput. Informatics.

[31]  Paul Springer,et al.  A Study of Productivity and Performance of Modern Vector Processors , 2012 .

[32]  Hung-Yu Lin,et al.  4D MR phase and magnitude segmentations with GPU parallel computing. , 2015, Magnetic resonance imaging.

[33]  Albert Y. Zomaya,et al.  Task-Tree Based Large-Scale Mosaicking for Massive Remote Sensed Imageries with Dynamic DAG Scheduling , 2014, IEEE Transactions on Parallel and Distributed Systems.

[34]  Konrad P Kording,et al.  How advances in neural recording affect data analysis , 2011, Nature Neuroscience.

[35]  Hans Knutsson,et al.  A GPU accelerated interactive interface for exploratory functional connectivity analysis of FMRI data , 2011, 2011 18th IEEE International Conference on Image Processing.

[36]  Xiaoli Li,et al.  Towards energy-efficient parallel analysis of neural signals , 2011, Cluster Computing.

[37]  Klaus Lehnertz,et al.  Measuring synchronization in coupled model systems: A comparison of different approaches , 2007 .

[38]  Eros Comunello,et al.  Diffusion tensor fiber tracking on graphics processing units , 2008, Comput. Medical Imaging Graph..

[39]  Dustin Scheinost,et al.  A Graphics Processing Unit Accelerated Motion Correction Algorithm and Modular System for Real-time fMRI , 2013, Neuroinformatics.

[40]  Justin C. Williams,et al.  Massively Parallel Signal Processing using the Graphics Processing Unit for Real-Time Brain–Computer Interface Feature Extraction , 2009, Front. Neuroeng..

[41]  Lizhe Wang,et al.  Massively Parallel Neural Signal Processing on a Many-Core Platform , 2011, Computing in Science & Engineering.

[42]  Xiaoli Li,et al.  Interaction dynamics of neuronal oscillations analysed using wavelet transforms , 2007, Journal of Neuroscience Methods.

[43]  Adelino Ferreira da Silva,et al.  cudaBayesreg: Bayesian Computation in CUDA , 2010, R J..

[44]  Andrzej Cichocki,et al.  PARAFAC algorithms for large-scale problems , 2011, Neurocomputing.

[45]  Karl Pearson F.R.S..  LIII. On lines and planes of closest fit to systems of points in space , 1901 .

[46]  Pierre Comon,et al.  Independent component analysis, A new concept? , 1994, Signal Process..

[47]  Wei Liu,et al.  Spatial Regularization of Functional Connectivity Using High-Dimensional Markov Random Fields , 2010, MICCAI.

[48]  Naren Ramakrishnan,et al.  Towards Chip-on-Chip Neuroscience: Fast Mining of Frequent Episodes Using Graphics Processors , 2009, ArXiv.

[49]  N. Alon,et al.  Resolution enhancement in MRI. , 2006, Magnetic resonance imaging.

[50]  Hong Bao,et al.  GPGPU-Aided Ensemble Empirical-Mode Decomposition for EEG Analysis During Anesthesia , 2010, IEEE Transactions on Information Technology in Biomedicine.

[51]  Guy B. Williams,et al.  A New Fast Accurate Nonlinear Medical Image Registration Program Including Surface Preserving Regularization , 2014, IEEE Transactions on Medical Imaging.

[52]  David R. Kaeli,et al.  Multi GPU implementation of iterative tomographic reconstruction algorithms , 2009, 2009 IEEE International Symposium on Biomedical Imaging: From Nano to Macro.

[53]  Michael Lees,et al.  Towards a data‐driven approach to scenario generation for serious games , 2014, Comput. Animat. Virtual Worlds.

[54]  Daniel Rueckert,et al.  Fast Volume Reconstruction from Motion Corrupted Stacks of 2D Slices , 2015, IEEE Transactions on Medical Imaging.

[55]  Rajkumar Buyya,et al.  A Case for Cooperative and Incentive-Based Coupling of Distributed Clusters , 2005, 2005 IEEE International Conference on Cluster Computing.

[56]  Yves Goussard,et al.  GPU-accelerated regularized iterative reconstruction for few-view cone beam CT. , 2015, Medical physics.

[57]  Kim L. Boyer,et al.  Pathological image segmentation for neuroblastoma using the GPU , 2008, 2008 5th IEEE International Symposium on Biomedical Imaging: From Nano to Macro.

[58]  Hiroyuki Morikawa,et al.  A Packet Scheduling Technique on Optical Ring Networks for Increasing Utilization , 2011 .

[59]  J. Anthony Movshon,et al.  Comparison of Recordings from Microelectrode Arrays and Single Electrodes in the Visual Cortex , 2007, The Journal of Neuroscience.

[60]  Xiaofeng Gong,et al.  Tensor decomposition of EEG signals: A brief review , 2015, Journal of Neuroscience Methods.

[61]  Mohamed Akil,et al.  Special issue on parallel computing for real-time image processing , 2011, Journal of Real-Time Image Processing.

[62]  Rüdiger Westermann,et al.  MR image reconstruction using the GPU , 2006, SPIE Medical Imaging.

[63]  T. Schreiber,et al.  Surrogate time series , 1999, chao-dyn/9909037.

[64]  Rajiv Ranjan,et al.  Cloud monitoring for optimizing the QoS of hosted applications , 2012, 4th IEEE International Conference on Cloud Computing Technology and Science Proceedings.

[65]  Alexander V. Veidenbaum,et al.  Large-scale neural circuit mapping data analysis accelerated with the graphical processing unit (GPU) , 2015, Journal of Neuroscience Methods.

[66]  Tao Zhang,et al.  Towards real-time detection of seizures in awake rats with GPU-accelerated diffuse optical tomography , 2015, Journal of Neuroscience Methods.

[67]  Tsuneya Kurihara,et al.  Efficient registration method of medical images using GPU , 2011, Medical Imaging.

[68]  Klaus Lehnertz,et al.  A distributed computing system for multivariate time series analyses of multichannel neurophysiological data , 2006, Journal of Neuroscience Methods.

[69]  Markus Diesmann,et al.  Advancing the Boundaries of High-Connectivity Network Simulation with Distributed Computing , 2005, Neural Computation.

[70]  H. Martin Bücker,et al.  Parallel Minimum p-Norm Solution of the Neuromagnetic Inverse Problem for Realistic Signals Using Exact Hessian-Vector Products , 2008, SIAM J. Sci. Comput..

[71]  Kang Zhang,et al.  An ensemble local means decomposition method and its application to local rub-impact fault diagnosis of the rotor systems , 2012 .