Dynamic Estimation of the Auditory Temporal Response Function From MEG in Competing-Speaker Environments

Abstract

Objective A central problem in computational neuroscience is to characterize brain function using neural activity recorded from the brain in response to sensory inputs with statistical confidence. Most of existing estimation techniques, such as those based on reverse correlation, exhibit two main limitations: first, they are unable to produce dynamic estimates of the neural activity at a resolution comparable with that of the recorded data, and second, they often require heavy averaging across time as well as multiple trials in order to construct statistical confidence intervals for a precise interpretation of data. In this paper, we address the above-mentioned issues for estimating auditory temporal response function (TRF) as a parametric computational model for selective auditory attention in competing-speaker environments. Methods The TRF is a sparse kernel which regresses auditory MEG data with respect to the envelopes of the speech streams. We develop an efficient estimation technique by exploiting the sparsity of the TRF and adopting an ℓ1-regularized least squares estimator which is capable of producing dynamic TRF estimates as well as confidence intervals at sampling resolution from single-trial MEG data. Results We evaluate the performance of our proposed estimator using evoked MEG responses from the human brain in an auditory attention experiment with two competing speakers. The TRFs are estimated dynamically over time using the proposed technique with multi-second resolution, which is a significant improvement over previous results with a temporal resolution of the order of a minute. Conclusion Application of our method to MEG data reveals a precise characterization of the modulation of M50 and M100 evoked responses with respect to the attentional state of the subject at multi-second resolution. Significance Our proposed estimation technique provides a high resolution real-time attention decoding framework in multi-speaker environments with potential application in smart hearing aid technology.

Neuro-Current Response Functions: A Unified Approach to MEG Source Analysis under the Continuous Stimuli Paradigm

“…To this end, we construct regularizers based on a convex norm of the NCRF matrix Φ, to both capture the structural properties of the NCRFs and facilitate algorithm development. The structural properties of interest in this case are spatial sparsity over the cortical source space, sparsity of the peaks/troughs, smoothness in the lag domain, and rotational invariance (Ding and Simon, 2012b;Akram et al, 2017).…”

“…In order to promote smoothness in the lag domain and sparsity of the peaks/troughs, we adopt a concept from Chen et al (2001), in which a temporally smooth time series is approximated by a small number of Gabor atoms over an over-complete dictionary G ∈ R L×L , for someL ≥ L (Feichtinger and Strohmer, 2012;Akram et al, 2017). To this end, we first perform a change of variables τ m := Gθ m , Φ = ΘG , and S := G S, where θ m ∈ RL ×3 are the coefficients of the m th NCRF over the dictionary G and Θ ∈ R 3M ×L is a matrix containing θ m s across its rows.…”

“…A commonly used regularization scheme in this context is the Tikhonov regularization and its variants for promoting smoothness (Lalor et al, 2006). Other estimation schemes such as boosting and 1 -regularization promote sparse solutions (David et al, 2007;Akram et al, 2017). Here, we introduce a structured regularization by penalizing a specific mixed-norm of the NCRF matrix to recover spatio-temporally sparse solutions over the Gabor coefficients:…”

“…Thanks to the grounding of the TRF model in linear system theory, several techniques from the system identification literature have been utilized for TRF estimation, such as the normalized reverse correlation (Theunissen et al, 2001), ridge regression (Machens et al, 2004), boosting (David et al, 2007), and SPARLS (Akram et al, 2017), some of which are available as software packages (Theunissen, 2007(Theunissen, , 2010Crosse et al, 2016;Brodbeck, 2017). While these methods have facilitated the characterization of the functional roles of various TRF components in sensory and cognitive processing of auditory stimuli, they predominantly aim at estimating TRFs over the MEG/EEG sensor space.…”

Stimulus-aware spatial filtering for single-trial neural response and temporal response function estimation in high-density EEG with applications in auditory research

“…of both attended and unattended speakers, to the neural activity of the listener (Di Liberto et al, 2015;Ding and Simon, 2012a,b;Power et al, 2012). TRFs not only have high temporal precision but are also sensitive to attentional modulation (Akram et al, 2017;Power et al, 2012). Forward modeling thus comes with the advantages of being able to investigate the TRFs and gain a better understanding of how our brain handles auditory stimuli Ding and Simon, 2012a,b), and also the possibility to identify the brain regions involved with stimulus processing (Das et al, 2016;Ding and Simon, 2012b;Etard et al, 2018;Power et al, 2012).…”