Limits...
Implicit learning of predictable sound sequences modulates human brain responses at different levels of the auditory hierarchy.

Lecaignard F, Bertrand O, Gimenez G, Mattout J, Caclin A - Front Hum Neurosci (2015)

Bottom Line: We observed a decrease of the MMN with predictability and interestingly, a similar effect at earlier latencies, within 70 ms after deviance onset.Following these pre-attentive responses, a reduced P3a was measured in the case of predictable deviants.We conclude that early and late deviance responses reflect prediction errors, triggering belief updating within the auditory hierarchy.

View Article: PubMed Central - PubMed

Affiliation: Lyon Neuroscience Research Center, CRNL, INSERM, U1028 - CNRS, UMR5292, Brain Dynamics and Cognition Team Lyon, France ; University Lyon 1 Lyon, France ; MEG Department, CERMEP Imaging Center Lyon, France.

ABSTRACT
Deviant stimuli, violating regularities in a sensory environment, elicit the mismatch negativity (MMN), largely described in the Event-Related Potential literature. While it is widely accepted that the MMN reflects more than basic change detection, a comprehensive description of mental processes modulating this response is still lacking. Within the framework of predictive coding, deviance processing is part of an inference process where prediction errors (the mismatch between incoming sensations and predictions established through experience) are minimized. In this view, the MMN is a measure of prediction error, which yields specific expectations regarding its modulations by various experimental factors. In particular, it predicts that the MMN should decrease as the occurrence of a deviance becomes more predictable. We conducted a passive oddball EEG study and manipulated the predictability of sound sequences by means of different temporal structures. Importantly, our design allows comparing mismatch responses elicited by predictable and unpredictable violations of a simple repetition rule and therefore departs from previous studies that investigate violations of different time-scale regularities. We observed a decrease of the MMN with predictability and interestingly, a similar effect at earlier latencies, within 70 ms after deviance onset. Following these pre-attentive responses, a reduced P3a was measured in the case of predictable deviants. We conclude that early and late deviance responses reflect prediction errors, triggering belief updating within the auditory hierarchy. Beside, in this passive study, such perceptual inference appears to be modulated by higher-level implicit learning of sequence statistical structures. Our findings argue for a hierarchical model of auditory processing where predictive coding enables implicit extraction of environmental regularities.

No MeSH data available.


Experimental design. (A) Schematic view of a complete cycle in predictable (left) and unpredictable conditions (right). Rectangles symbolize single tones with standards and deviants colored in gray and black, respectively. Sound duration was 70 ms with stimulus onset asynchrony (SOA) set to 610 ms. In every condition, each cycle entails seven deviants, each of them being preceded by a number of standards ranging from 2 to 8. A chunk of n standards corresponds to n+1 tones (n consecutive standards and the following deviant), as illustrated by the shaded area in condition PF. Chunks are sorted by their size in predictable condition, whereas these are shuffled in unpredictable ones. (B) Variation of the size of chunks (black circle) within cycles, over sound sequence in predictable (left) and unpredictable conditions (right). Each sequence is composed of 16 cycles and examples of shuffled cycles are presented for unpredictable conditions. Shaded areas delineating one cycle in both sequence types highlight their difference with respect to sound predictability.
© Copyright Policy
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC4584941&req=5

Figure 1: Experimental design. (A) Schematic view of a complete cycle in predictable (left) and unpredictable conditions (right). Rectangles symbolize single tones with standards and deviants colored in gray and black, respectively. Sound duration was 70 ms with stimulus onset asynchrony (SOA) set to 610 ms. In every condition, each cycle entails seven deviants, each of them being preceded by a number of standards ranging from 2 to 8. A chunk of n standards corresponds to n+1 tones (n consecutive standards and the following deviant), as illustrated by the shaded area in condition PF. Chunks are sorted by their size in predictable condition, whereas these are shuffled in unpredictable ones. (B) Variation of the size of chunks (black circle) within cycles, over sound sequence in predictable (left) and unpredictable conditions (right). Each sequence is composed of 16 cycles and examples of shuffled cycles are presented for unpredictable conditions. Shaded areas delineating one cycle in both sequence types highlight their difference with respect to sound predictability.

Mentions: To design the predictable sequences (Figure 1), we did not use a fixed number of standards between two deviants as in Scherg et al. (1989), because this cannot be mirrored in the unpredictable sequence without inducing different refractoriness effects. This issue could be avoided by the construction of a statistical structure unfolding over a larger time-scale. Precisely, the rule that we designed increments the number of standards progressively within a cycle: it starts with one deviant after two standards, followed by one deviant after three standards and so on until one deviant after eight standards. From now on, a chunk with n standards will refer to a series of n standard sounds ending with a deviant stimulus (n ranging from 2 to 8). The 42-tone cycle, composed of seven incrementing chunks, was repeated 16 times in the sequence, thus leading to a total of 560 standards and 112 deviants. For the unpredictable sequences, each cycle was shuffled so as to permute the order of the seven chunks with the constraint that no chunk with n standard was preceded or followed by a chunk with either n-1 or n+1 standards. Additionally, the transition between two cycles was such that no successive chunks with n standards could occur. Altogether this randomization allowed to (1) avoid any global rule to emerge in the unpredictable sequence and (2) have exactly the same number of chunks with n standards in predictable and unpredictable conditions. Note that the number of deviants presented at a 2–3 chunk timescale may differ between UF and PF (for instance, the set of 16 sounds that precede a “chunk of 8 standards” deviant comprises exactly one deviant in PF and two deviants on average in UF) but the fact that adaptation saturates rapidly [2–3 standard repetitions, (Demarquay et al., 2011)] led us to assume that this particularity did not introduce any significant adaptation effect difference between PF and UF, in the current analysis that we conducted with standards just preceding deviants.


Implicit learning of predictable sound sequences modulates human brain responses at different levels of the auditory hierarchy.

Lecaignard F, Bertrand O, Gimenez G, Mattout J, Caclin A - Front Hum Neurosci (2015)

Experimental design. (A) Schematic view of a complete cycle in predictable (left) and unpredictable conditions (right). Rectangles symbolize single tones with standards and deviants colored in gray and black, respectively. Sound duration was 70 ms with stimulus onset asynchrony (SOA) set to 610 ms. In every condition, each cycle entails seven deviants, each of them being preceded by a number of standards ranging from 2 to 8. A chunk of n standards corresponds to n+1 tones (n consecutive standards and the following deviant), as illustrated by the shaded area in condition PF. Chunks are sorted by their size in predictable condition, whereas these are shuffled in unpredictable ones. (B) Variation of the size of chunks (black circle) within cycles, over sound sequence in predictable (left) and unpredictable conditions (right). Each sequence is composed of 16 cycles and examples of shuffled cycles are presented for unpredictable conditions. Shaded areas delineating one cycle in both sequence types highlight their difference with respect to sound predictability.
© Copyright Policy
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC4584941&req=5

Figure 1: Experimental design. (A) Schematic view of a complete cycle in predictable (left) and unpredictable conditions (right). Rectangles symbolize single tones with standards and deviants colored in gray and black, respectively. Sound duration was 70 ms with stimulus onset asynchrony (SOA) set to 610 ms. In every condition, each cycle entails seven deviants, each of them being preceded by a number of standards ranging from 2 to 8. A chunk of n standards corresponds to n+1 tones (n consecutive standards and the following deviant), as illustrated by the shaded area in condition PF. Chunks are sorted by their size in predictable condition, whereas these are shuffled in unpredictable ones. (B) Variation of the size of chunks (black circle) within cycles, over sound sequence in predictable (left) and unpredictable conditions (right). Each sequence is composed of 16 cycles and examples of shuffled cycles are presented for unpredictable conditions. Shaded areas delineating one cycle in both sequence types highlight their difference with respect to sound predictability.
Mentions: To design the predictable sequences (Figure 1), we did not use a fixed number of standards between two deviants as in Scherg et al. (1989), because this cannot be mirrored in the unpredictable sequence without inducing different refractoriness effects. This issue could be avoided by the construction of a statistical structure unfolding over a larger time-scale. Precisely, the rule that we designed increments the number of standards progressively within a cycle: it starts with one deviant after two standards, followed by one deviant after three standards and so on until one deviant after eight standards. From now on, a chunk with n standards will refer to a series of n standard sounds ending with a deviant stimulus (n ranging from 2 to 8). The 42-tone cycle, composed of seven incrementing chunks, was repeated 16 times in the sequence, thus leading to a total of 560 standards and 112 deviants. For the unpredictable sequences, each cycle was shuffled so as to permute the order of the seven chunks with the constraint that no chunk with n standard was preceded or followed by a chunk with either n-1 or n+1 standards. Additionally, the transition between two cycles was such that no successive chunks with n standards could occur. Altogether this randomization allowed to (1) avoid any global rule to emerge in the unpredictable sequence and (2) have exactly the same number of chunks with n standards in predictable and unpredictable conditions. Note that the number of deviants presented at a 2–3 chunk timescale may differ between UF and PF (for instance, the set of 16 sounds that precede a “chunk of 8 standards” deviant comprises exactly one deviant in PF and two deviants on average in UF) but the fact that adaptation saturates rapidly [2–3 standard repetitions, (Demarquay et al., 2011)] led us to assume that this particularity did not introduce any significant adaptation effect difference between PF and UF, in the current analysis that we conducted with standards just preceding deviants.

Bottom Line: We observed a decrease of the MMN with predictability and interestingly, a similar effect at earlier latencies, within 70 ms after deviance onset.Following these pre-attentive responses, a reduced P3a was measured in the case of predictable deviants.We conclude that early and late deviance responses reflect prediction errors, triggering belief updating within the auditory hierarchy.

View Article: PubMed Central - PubMed

Affiliation: Lyon Neuroscience Research Center, CRNL, INSERM, U1028 - CNRS, UMR5292, Brain Dynamics and Cognition Team Lyon, France ; University Lyon 1 Lyon, France ; MEG Department, CERMEP Imaging Center Lyon, France.

ABSTRACT
Deviant stimuli, violating regularities in a sensory environment, elicit the mismatch negativity (MMN), largely described in the Event-Related Potential literature. While it is widely accepted that the MMN reflects more than basic change detection, a comprehensive description of mental processes modulating this response is still lacking. Within the framework of predictive coding, deviance processing is part of an inference process where prediction errors (the mismatch between incoming sensations and predictions established through experience) are minimized. In this view, the MMN is a measure of prediction error, which yields specific expectations regarding its modulations by various experimental factors. In particular, it predicts that the MMN should decrease as the occurrence of a deviance becomes more predictable. We conducted a passive oddball EEG study and manipulated the predictability of sound sequences by means of different temporal structures. Importantly, our design allows comparing mismatch responses elicited by predictable and unpredictable violations of a simple repetition rule and therefore departs from previous studies that investigate violations of different time-scale regularities. We observed a decrease of the MMN with predictability and interestingly, a similar effect at earlier latencies, within 70 ms after deviance onset. Following these pre-attentive responses, a reduced P3a was measured in the case of predictable deviants. We conclude that early and late deviance responses reflect prediction errors, triggering belief updating within the auditory hierarchy. Beside, in this passive study, such perceptual inference appears to be modulated by higher-level implicit learning of sequence statistical structures. Our findings argue for a hierarchical model of auditory processing where predictive coding enables implicit extraction of environmental regularities.

No MeSH data available.