Several subcortical nuclei along the auditory pathway are involved in the processing of sounds. One of the most commonly used methods of measuring the activity of these nuclei is the auditory brainstem response (ABR). Due to its low signal-to-noise ratio, ABR's have to be derived by averaging over activity generated by thousands of artificial sounds such as clicks or tone bursts. This approach cannot be easily applied to natural listening situations (e.g. speech, music), which limits auditory cognitive neuroscientific studies to investigate mostly cortical processes. We propose that by individually training backward encoding models to reconstruct evoked ABRs from high-density electrophysiological data, spatial filters can be tuned to auditory brainstem activity. Since these individualized filters can be applied (i.e. generalized) to any other data set using the same spatial coverage, this could allow for the estimation of auditory brainstem activity from any continuous sensor level data. In this study, we established a proof-of-concept by using backward encoding models generated using a click stimulation rate of 30 Hz to predict ABR activity recorded using EEG from an independent measurement using a stimulation rate of 9 Hz. We show that individually predicted and measured ABR's are highly correlated (r ∼ 0.7). Importantly these predictions are stable even when applying the trained backward encoding model to a low number of trials, mimicking a situation with an unfavorable signal-to-noise ratio. Overall, this work lays the necessary foundation to use this approach in more interesting listening situations.
Bibliographical noteCopyright © 2020. Published by Elsevier Inc.
Fields of Science and Technology Classification 2012
- 501 Psychology