UCL  IRIS
Institutional Research Information Service
UCL Logo
Please report any queries concerning the funding data grouped in the sections named "Externally Awarded" or "Internally Disbursed" (shown on the profile page) to your Research Finance Administrator. Your can find your Research Finance Administrator at https://www.ucl.ac.uk/finance/research/rs-contacts.php by entering your department
Please report any queries concerning the student data shown on the profile page to:

Email: portico-services@ucl.ac.uk

Help Desk: http://www.ucl.ac.uk/ras/portico/helpdesk
Publication Detail
Learning of inter-element-interval statistics in sound sequences - an MEG study
  • Publication Type:
    Conference
  • Authors:
    Andreou LV, Chait M
  • Publication date:
    2012
  • Name of conference:
    International Conference on Auditory Cortex
  • Conference start date:
    01/01/2012
Abstract
An important aspect of auditory scene analysis is the ability to accurately learn the time structure of acoustic sequences. This capacity is important for predicting when a future event will occur (and thus being able to respond to it optimally) and also for detecting changes in the input, indicative of new events in one's surroundings. In this MEG study we use sound sequences (tone-pips, separated by variable duration silence intervals) in order to investigate the extent to which naive, distracted, listeners are able to acquire the time structure of on-going acoustic input. To probe such mechanisms, we measured auditory cortical responses to the offset of long sound sequences. If listeners are able to accurately learn the temporal properties of a sequence and form expectancies with regards to the arrival time of an upcoming tone, sequence offsets should be detectable as soon as an expected tone fails to arrive (offset of the last audible tone in a sequence + duration of expected silence interval following that tone). Therefore latencies of auditory cortical offset responses (measured relative to the time of the last tone in a sequence) are indicative of the extent to which sequence time structure has been acquired. The MEG paradigm involved naive listeners passively listening to sounds while performing an irrelevant (decoy) visual task. The stimuli were long sequences of tone bursts separated by silence intervals (inter element interval; IEI). In Experiment 1, the sequences were isochronous (constant IEI). The IEI was set to one of three values: (1) IEI=50ms, (2) IEI=100ms (3) IEI=200ms. The sequences were of variable overall duration and presented in random order. These specific IEI durations were chosen because they are characteristic of the temporal properties of natural sequences. The stimulus set also contained long pure-tone stimuli in order to measure the baseline latency of auditory cortical offset responses. In order to evaluate IEI learning, we subtracted the relevant IEI from the measured offset-peak latency, to compute a "corrected" offset latency. After this subtraction we observed no statistical difference between IEI conditions (although the corrected latency in all three IEI conditions was significantly longer than the "simple" offset response latency measured relative to a pure tone). The results demonstrate, therefore, that brain responses of distracted listeners adapted to sequence IEI. In Experiment 2, we used non-isochronous, but temporally regular, sequences. These were created by using the same IEIs as in Experiment 1 but such that sequence silence durations alternated regularly between the 3 values e.g. 50-100-200-50-100-200 etc. (IEIs were presented with equal probability). The sequences were grouped into three conditions based on the duration of the last predictable interval (the expected IEI after the last audible tone): (1) last IEI=50ms, (2) last IEI=100ms and (3) last IEI=200ms. As in Experiment 1, the stimulus set also contained pure tone stimuli. Results show that, as compared to isochronous sequences, there is a significant latency increase in the response to the 50ms and 200ms IEIs (around 50ms and 35 ms, respectively). Interestingly, the 100ms interval appears to be learnt equally efficiently in both experiments. Overall, our results demonstrate that the auditory cortex is sensitive to the temporal structure of isochronous sequences, even when this information is not behaviourally relevant. However, the temporal structure of regular sequences is not learnt as precisely.
Publication data is maintained in RPS. Visit https://rps.ucl.ac.uk
 More search options
UCL Researchers
Author
The Ear Institute
University College London - Gower Street - London - WC1E 6BT Tel:+44 (0)20 7679 2000

© UCL 1999–2011

Search by