Cross multivariate correlation coefficients as screening tool for analysis of concurrent EEG‐fMRI recordings

Abstract

Over the past decade, the simultaneous recording of electroencephalogram (EEG) and functional magnetic resonance imaging (fMRI) data has garnered growing interest because it may provide an avenue towards combining the strengths of both imaging modalities. Given their pronounced differences in temporal and spatial statistics, the combination of EEG and fMRI data is however methodologically challenging. Here, we propose a novel screening approach that relies on a Cross Multivariate Correlation Coefficient (xMCC) framework. This approach accomplishes three tasks:(1) It provides a measure for testing multivariate correlation and multivariate uncorrelation of the two modalities; (2) it provides criterion for the selection of EEG features; (3) it performs a screening of relevant EEG information by grouping the EEG channels into clusters to improve efficiency and to reduce computational load when searching for the best predictors of the BOLD signal. The present report applies this approach to a data set with concurrent recordings of steady-state-visual evoked potentials (ssVEPs) and fMRI, recorded while observers viewed phase-reversing Gabor patches. We test the hypothesis that fluctuations in visuo-cortical mass potentials systematically covary with BOLD fluctuations not only in visual cortical, but also in anterior temporal and prefrontal areas. Results supported the hypothesis and showed that the xMCC-based analysis provides straightforward identification of neurophysiological plausible brain regions with EEG-fMRI covariance. Furthermore xMCC converged with other extant methods for EEG-fMRI analysis.

Functional Source Separation for EEG-fMRI Fusion: Application to Steady-State Visual Evoked Potentials

“…In conclusion, the present study demonstrated that FSS, in conjunction with a multivariate correlation and feature selection approach proposed previously (Ji et al, 2018), yields robust ssVEP-fMRI co-variation maps with high face validity (i.e., high correspondence between BOLD and ssVEP in visual cortex, where strong responses are expected to during visual stimulation) and with high replicability across re-tests (i.e., across three experimental phases) in the same participants. The work is important to neurorobotics for several reasons: (1) This approach could be used to examine a wide range of hypotheses regarding the neurophysiology of visual cortex and its function for biological and artificial systems, (2) it provides a means to conduct the multi-modal information fusion and correlation analysis to recover the underlying neurophysiological mechanisms, and (3) the core idea of the approach are inspiring for other brain data processing and the knowledge can be transferred to other applications as well.…”

“…This is consistent with the sources of ssVEP established through intracranial recordings in humans and experimental animals (see e.g., Wittevrongel et al, 2018). Furthermore, the cross-participant and re-test convergence along with the large spatial extent of the ssVEP-fMRI covariation in visual cortex observed in the present study exceeds earlier analyses of this same data-set with different methods, i.e., parametric modulation analysis as implemented in SPM (Petro et al, 2017), and multivariate correlation analyses without the FSS preprocessing step (Ji et al, 2018). Because the covariation of both signal modalities across the time series in visual cortex can be considered a gold standard for analytical strategies, the fact that ssVEP-fMRI correlations were observed reliably in extended visual cortex suggests that FSS represents a desirable preprocessing step, positively affecting the validity of EEG-fMRI fusion.…”

“…As in the procedure of FSS optimization described in methods, the proper scaling (including sign) of the source signals is non-recoverable, we are mostly interested in the predictability reflects the proportion of the variance in the BOLD signal that is linearly predicted from ssVEP time series. Following the previous work (Wang and Zheng, 2014; Ji et al, 2018), we use the bivariate case of cross multivariate correlation coefficient (xMCC) for correlation analysis between EEG and fMRI, which is equal to the absolute value of the cross correlation coefficient. The metabolic (BOLD) events are slower and temporally lagged to the neuro-electric activity (EEG).…”

“…In addition, and of particular relevance in the absence of experimental stimulation, specific portions of the co-variance between modalities may reflect neural-hemodynamic dependencies in spontaneous, ongoing brain activity. In previous work, we have described a method for quantifying the spatial and temporal correlations between the time series of fMRI BOLD and suitable indices derived from the EEG signal (Ji et al, 2018). The present work examines in more detail the benefits of ssVEP preprocessing using Functional Source Separation (FSS) as an optimization step.…”

“…The present paper illustrates the usage of this FSS based approach for quantifying ssVEP-fMRI co-variation. Using a fusion algorithm that we have previously described (Ji et al, 2018), we applied the FSS preprocessing step to data from a 3-phase aversive conditioning paradigm that included a habituation, acquisition, and extinction phase. EEG and fMRI were recorded simultaneously while participants viewed periodically (10Hz) and rapidly phase-reversing Gabor patches (sine-wave gratings), evoking steady-state visual potentials (ssVEPs).…”