You are viewing the SNL 2017 Archive Website. For the latest information, see the Current Website.

Poster B49, Wednesday, November 8, 3:00 – 4:15 pm, Harborview and Loch Raven Ballrooms

Neurobiological mechanisms of efficient encoding: A pilot EEG study.

Nicholas Walker1, Christian Stilp2, Keith Kleunder3, Julia Evans1, Meredith Scheppele1;1Univeristy of Texas at Dallas, 2University of Louisville, 3Purdue University

Kluender, Stilp and Kiefte (2013) propose that efficient encoding of covarying perceptual attributes is an important mechanism of auditory learning that may have implications for language learning. Prior work has shown that participants encode covarying statistical structures in the auditory system (Stilp, Rogers, & Kluender, 2010). Changes in brainstem, thalamocortical, and corticothalamic processing may explain efficient perceptual encoding. EEG provides a tool for examining perceptual encoding of information as sounds are being heard. The purpose of this project is to determine if tracking of covarying stimulus properties is due to efficient encoding (bottom-up changes) or top-down processing. Methods: Seven college-aged participants with normal hearing, and no history of language or neurological disorders completed the experiment. The protocol consisted of four phases: passive exposure, two testing blocks and passive exposure. Stimuli from Experiment 1 in Stilp et al. (2010) were used. Passive exposures consisted of 250 random presentation of the sounds used in the experiment (1000 ms ISI). Testing blocks were identical to blocks of Experiment 1 from Stilp et al. (2010). EEG analysis is restricted to the passive exposure blocks to examine changes in encoding of information and assess the feasibility of using time-frequency analysis to examine encoding of information. EEG data were recorded using a 64-channel BrainVision actiCHamp system. Bad channels were deleted prior to analysis. Ocular and single channel artifacts were removed using infomax ICA with EEGLAB (Delorme & Makeig, 2004). EEG data were segmented into 5 second epochs around stimuli. Epochs were then average referenced and baseline corrected to the prestimulus interval. Continuous wavelet transforms were computed at 100 log-spaced frequencies from 3Hz to 100Hz. Wavelet power was computed and baseline corrected to a prestimulus interval (-250 to – 50 ms) using a decibel correction. Weighted intertrial phase clustering was computed across conditions (Cohen, 2014). Results: From pre-testing to post-testing, there is a significant difference in theta phase clustering and power at left and frontal electrode sites (T7, F7, FZ) during the N1 component, from 100-200 ms after sound onset, with decreased theta phase clustering and power. Conversely, there is a significant increase in alpha power from pre-testing to post-testing from 100-200 ms. Discussion: Efficient encoding of covarying information in audition has been proposed as a learning mechanism important for speech perception (Kluender et al., 2013). The measured changes in brain activity are occurring during the perception of the sound stimuli, which suggests changes in the way the brain is processing the auditory information. Decreases in theta power and phase clustering may suggest an increase in processing efficiency or habituation to the task. However, increases in alpha power have been associated with increased attention and memory (Bastiaansen, Posthuma, Groot, & de Geus, 2002; Bishop, Hardiman, & Barry, 2010; Klimesch, 1999, 2012) and are in line with improved processing of the stimuli. These differences in theta and alpha frequency bands are occurring during the N1 response which is generated by thalamic and primary auditory cortex, and may suggest efficient encoding is occurring prior to higher level perceptual processes.

Topic Area: Perception: Auditory

Back to Poster Schedule