22Recent studies have shown that acoustically distorted sentences can be perceived as either unintelligible or intel-23 ligible depending on whether one has previously been exposed to the undistorted, intelligible versions of the 24 sentences. This allows studying processes specifically related to speech intelligibility since any change between 25 the responses to the distorted stimuli before and after the presentation of their undistorted counterparts cannot 26 be attributed to acoustic variability but, rather, to the successful mapping of sensory information onto memory 27 representations. To estimate how the complexity of the message is reflected in speech comprehension, we ap-28 plied this rapid change in perception to behavioral and magnetoencephalography (MEG) experiments using 29 vowels, words and sentences. In the experiments, stimuli were initially presented to the subject in a distorted 30 form, after which undistorted versions of the stimuli were presented. Finally, the original distorted stimuli 31 were presented once more. The resulting increase in intelligibility observed for the second presentation of the 32 distorted stimuli depended on the complexity of the stimulus: vowels remained unintelligible (behaviorally 33 measured intelligibility 27%) whereas the intelligibility of the words increased from 19% to 45% and that of the 34 sentences from 31% to 65%. This increase in the intelligibility of the degraded stimuli was reflected as an enhance-35 ment of activity in the auditory cortex and surrounding areas at early latencies of 130-160 ms. In the same re-36 gions, increasing stimulus complexity attenuated mean currents at latencies of 130-160 ms whereas at 37 latencies of 200-270 ms the mean currents increased. These modulations in cortical activity may reflect feedback 38 from top-down mechanisms enhancing the extraction of information from speech. The behavioral results suggest 39 that memory-driven expectancies can have a significant effect on speech comprehension, especially in acousti-40 cally adverse conditions where the bottom-up information is decreased. 41 © 2015 Published by Elsevier Inc. responses to a set of vocoded sentences before and after the subject 73 was trained to perceive these sentences correctly in a learning phase NeuroImage xxx (2015) xxx-xxx ⁎ Corresponding author at: Department of Neuroscience and Biomedical Engineering (NBE), School of Science, Aalto University, PO Box 12200, FI-00076 AALTO, Finland.E-mail address: maria.hakonen@aalto.fi (M. Hakonen). YNIMG-12658; No. of pages: 13; 4C: 6, 8, 9 http://dx. where normal speech and vocoded speech were paired.
109In MEG and EEG measurements, auditory stimuli elicit a series of 110 transient responses, the most prominent of which is the auditory N1 re-111 sponse, measured electrically, and its magnetic counterpart, the N1m
112(for reviews, see Näätänen and Picton, 1987;May and Tiitinen, 2010).
113In the case of long-duration stimuli (N 300 ms), the transient responses , 2010, 2011, 2012 Liikkanen et al., 2007;Gray...