Abstract
Aims. This work seeks to seeks
electrophysiological event-related potentials (EPRs) in
the human EEG that are indicative of the processing of
a metrical beat, based on purely temporal characteristics
of basic rhythmical tone sequences. Methods.
Motivated by pre-existing behavioural work (Povel and
Essens, 1985; Grube and Griffiths, 2009) on temporal
processing accuracy and subjective perception of the
catchiness of rhythmic patterns with a strongly vs. a
weakly metrical beat, a total of 784 sequences were
composed (each containing 7 or 8 tones distributed over
16 beat units of 200 ms). Evaluated based on
behavioural ratings, an optimally separable set of 28
strongly and 28 weakly metrical sequences was chosen
and employed in a passive listening EEG paradigm
looking at the processing of sequences in their original
format and with a violated ending. In order to analyse
ERP markers of metrical processing, continuous EEG
data were epoched and averaged across subjects (n=20).
Results. Grand average responses of initial analysis at
FZ display differences in the EEG for strongly
compared to weakly metrical sequences that indicate the
beat-based analysis of the unfolding of sequences, their
metrically plausible endings and violation of those.
Conclusions. The findings provide evidence for the
beat-based predictive encoding of such minimalistic
rhythms and the temporal prediction of their metrically
plausible endings. On-going analyses tease aim to tease
out the individual components and test the classification
of strongly vs. weakly metrical beat processing in a
machine learning-based multiple electrode approach
electrophysiological event-related potentials (EPRs) in
the human EEG that are indicative of the processing of
a metrical beat, based on purely temporal characteristics
of basic rhythmical tone sequences. Methods.
Motivated by pre-existing behavioural work (Povel and
Essens, 1985; Grube and Griffiths, 2009) on temporal
processing accuracy and subjective perception of the
catchiness of rhythmic patterns with a strongly vs. a
weakly metrical beat, a total of 784 sequences were
composed (each containing 7 or 8 tones distributed over
16 beat units of 200 ms). Evaluated based on
behavioural ratings, an optimally separable set of 28
strongly and 28 weakly metrical sequences was chosen
and employed in a passive listening EEG paradigm
looking at the processing of sequences in their original
format and with a violated ending. In order to analyse
ERP markers of metrical processing, continuous EEG
data were epoched and averaged across subjects (n=20).
Results. Grand average responses of initial analysis at
FZ display differences in the EEG for strongly
compared to weakly metrical sequences that indicate the
beat-based analysis of the unfolding of sequences, their
metrically plausible endings and violation of those.
Conclusions. The findings provide evidence for the
beat-based predictive encoding of such minimalistic
rhythms and the temporal prediction of their metrically
plausible endings. On-going analyses tease aim to tease
out the individual components and test the classification
of strongly vs. weakly metrical beat processing in a
machine learning-based multiple electrode approach
Original language | English |
---|---|
Publication date | 1 May 2018 |
Publication status | Published - 1 May 2018 |
Event | Neuroscience Day 2018: Rewarding Neuroscience - Aarhus University, Aarhus, Denmark Duration: 1 May 2018 → 1 May 2018 http://neurocampus.au.dk/neuroscience-day/neuroscience-day-2018/ |
Conference
Conference | Neuroscience Day 2018 |
---|---|
Location | Aarhus University |
Country/Territory | Denmark |
City | Aarhus |
Period | 01/05/2018 → 01/05/2018 |
Internet address |