An exploratory data analysis method for identifying brain regions and frequencies of interest from large-scale neural recordings

Abstract

High-resolution whole brain recordings have the potential to uncover unknown functionality but also present the challenge of how to find such associations between brain and behavior when presented with a large number of regions and spectral frequencies. In this paper, we propose an exploratory data analysis method that sorts through a massive quantity of multivariate neural recordings to quickly extract a subset of brain regions and frequencies that encode behavior. This approach combines existing tools and exploits low-rank approximation of matrices without a priori selection of regions and frequency bands for analysis. In detail, the spectral content of neural activity across all frequencies of each recording contact is computed and represented as a matrix. Then, the rank-1 approximation of the matrix is computed using singular value decomposition and the associated singular vectors are extracted. The temporal singular vector, which captures the salient features of the spectrogram, is then correlated to the trial-varying behavioral signal. The distribution of correlations for each brain region is efficiently computed and used to find a subset of regions and frequency bands of interest for further examination. As an illustration, we apply this approach to a data set of local field potentials collected using stereoelectroencephalography from a human subject performing a reaching task. Using the proposed procedure, we produced a comprehensive set of brain regions and frequencies related to our specific behavior. We demonstrate how this tool can produce preliminary results that capture neural patterns related to behavior and aid in formulating data-driven hypotheses, hence reducing the time it takes for any scientist to transition from the exploratory to the confirmatory phase.

Internal states as a source of subject-dependent movement variability and their representation by large-scale networks

“…Inducing movement variability using our motor task. Our motor task was a center-out delay arm reach where subjects won virtual money by controlling a cursor on a screen to reach a target with an instructed speed despite a chance of encountering a random physical perturbation [70,[81][82][83] . Subjects performed this task in the EMU using a behavioral control system, which consisted of three elements: a computer screen, an InMotion2 robotic manipulandum (Interactive Motion Technologies, USA), and a Windows-based laptop computer [81] .…”

Non-motor Brain Regions in Non-dominant Hemisphere Are Influential in Decoding Movement Speed

“…Subjects performed goal-directed reaching movements with speed instructions that have been previously described (Johnson et al, 2014; Breault et al, 2017, 2018, 2019a,b; Kerr et al, 2017). Movements were made using a robotic manipulandum from the InMotion ARM Interactive Therapy System (Interactive Motion Technologies, Watertown, MA, USA) and were displayed as a cursor on an attached computer screen (Figure 2A).…”

Emerging techniques in statistical analysis of neural data

“…For example, state-space models are used to understand how latent variables (states) influence neural and behavioral measurements or to simply explain how and why control systems in the central nervous system operate the way they do. This special issue includes pioneering studies that describe methods to sift through large amounts of data to identify brain regions and frequency bands of interest (Breault et al 2018); to construct models from multi-scale neural data ranging from spike trains from individual neurons (Chen et al 2018 to EEG recordings from populations of neurons (Talukdar et al 2018); and to decode behavior from neural data (Han et al 2018), with applications to neuroprosthetics and brain-machine interfaces. Network connectivity studies in the contexts of brain state changes (Luckett et al 2018, Xiao et al 2018 and language are also presented (Grappe et al 2018).…”