How Big Data and High-performance Computing Drive Brain Science

Abstract

Brain science accelerates the study of intelligence and behavior, contributes fundamental insights into human cognition, and offers prospective treatments for brain disease. Faced with the challenges posed by imaging technologies and deep learning computational models, big data and high-performance computing (HPC) play essential roles in studying brain function, brain diseases, and large-scale brain models or connectomes. We review the driving forces behind big data and HPC methods applied to brain science, including deep learning, powerful data analysis capabilities, and computational performance solutions, each of which can be used to improve diagnostic accuracy and research output. This work reinforces predictions that big data and HPC will continue to improve brain science by making ultrahigh-performance analysis possible, by improving data standardization and sharing, and by providing new neuromorphic insights.

Keywords: Big data; Brain connectomes; Brain science; Deep learning; High-performance computing.

Figure 1

Research overview of big data and HPC methods in brain science A. The heatmap shows the changes in the number of articles published annually from 2000 to 2018 in four research directions of brain science: brain science with HPC; brain science with deep learning; brain science with big data; and brain science with neural networks. Articles in brain science with deep learning, brain science with big data, and brain science with neural networks reached their highest numbers in 2013, whereas articles in brain science with HPC reached its highest number in 2018. All articles were retrieved by searching using keywords “brain science, HPC” (BS-HPC), “brain science, deep learning” (BS-DL), “brain science, big data” (BS-BD), or “brain science, neural network” (BS-NN) in Google Scholar in September 2019. B. Combinations between brain science and big data or HPC methods. Big data provide a wealth of knowledge and data, from which neural networks and deep learning methods can extract features that represent brain functions, mechanisms, or diseases. Big data can also be used to build computational models. HPC provides storage space and formidable computing power for the study of brain science. HPC, high-performance computing.

Figure 2

General classification of research activities in brain science This figure shows the main research directions in contemporary brain science.

Figure 3

Research status of brain science in combination with other fields This figure shows the number of articles listed in Google™ Scholar each year from 2000 to 2018 by the following terms: BS, BS-NN, BSBD, BS-DL, and BS-HPC. This figure comprises three parts. The histogram shows trends in the number of articles on BS combined with each of the other four fields. Each thumbnail in the lower triangular area consists of a correlation ellipse, a scattergram of the corresponding rows and columns, and its LOWESS smoothing curve. The correlation ellipse indicates the correlation between corresponding rows and columns. A flatter oval indicates a stronger correlation. The LOWESS smoothing curve shows the trend between the two sets of data over time. Each thumbnail in the upper triangle contains a value that represents the correlation coefficient of the corresponding row and column. For example, the value 0.44 in the first row and the second column refers to the correlation coefficient between BS and BS-NN. BS, brain science; BS-NN, brain science with neural network; BS-BD, brain science with big data; BS-DL brain science with deep learning; BS-HPC, brain science with high-performance computing.