Figure 1.
Schematic representation of the paradigm.
(A) A trial started with the presentation of a visual cue “A” for auditory only trials (‘attentive’ trials) or “V” for audio-visual trials (‘distracted’ trials) (top and bottom respectively). For auditory only trials, subjects had to pay attention to a sequence of five sounds (duration: 50 ms, SOA: 150 ms). The fifth sound could be either standard or deviant according to a local (within a second) or a global (over tens of seconds) regularity (see method). For audio-visual trials, five colored rings (duration 166 ms, SOA: 250 ms) were presented concurrently to the sounds and the subject was instructed to estimate which of two colors was presented more often. (B) Schematic representation of a block of trials. An experimental block started with an ‘exposure phase’ in which only global standard stimuli were presented so that the participant could learn the regularity with perfectly predictable sounds. Half of the subjects were distracted during the exposure phase (100% distracted global standard trials, represented in black) while the other half was not (100% attentive global standard trials, represented in grey). We named these groups of subjects ‘distractively trained (AT) subjects’ and ‘attentively trained (DT) subjects’ respectively. The exposure phase was followed by a ‘test phase’ composed of 80% global standard stimuli and 20% global deviant. Attentive and distracted trials were inter-mixed during the test phase: 70% of global standard and 10% of global deviant were distracted trials, leaving an equal proportion of global standard and deviant in attentive trials (10% each).
Figure 2.
ERFs related to local and global violations.
Group averaged event-related fields for local (A, B) and global (C, D) violations in AT (A, C) and DT (B, D) subjects. In each panel, the first line of topographies represent deviant, standard and the difference between deviant and standard at 120 ms for the local effect and 380 ms for the global effect in attentive trials. The second line of topographies represents the same subtraction but for distracted trials. Colored squares indicate the corresponding time course depicted below the topographies. Black circles on topographies represent channels showing a significant difference between standard and deviant conditions (see method). Finally, a representative channel shows the time course of the deviancy effect both in attentive (blue and red lines) and distracted trials (green and black lines). As can be seen, local violations induced a mismatch effect in the ERFs which remained essentially unaffected by the visual task. However, a global effect was observed only on distracted trials. In attentive trials, both standard and deviant induced large late ERFs.
Figure 3.
Mean amplitude (± s.e.m.) for local (A) and global (B) effects.
For the local effect, we selected four right temporal sensors and measured the difference in amplitude between two peaks of the ERFs: we averaged the amplitude between 50 and 150 ms and between 150 and 250 ms and computed the difference between the two. The same procedure was used for the global effect but with a different set of sensors (as indicated by the empty head inset) and different time windows: 100–200 ms and 350–600 ms. Each color represents a combination of group and condition: attentive (red) and distracted (yellow) trials in AT subjects; attentive (green) and distracted (blue) trials in DT subjects. Dark and bright colors represent standard and deviant stimuli respectively. Brackets and stars represent results from t test comparing amplitudes for deviants and standard stimuli: *: p<0.05; **: p<0.01; ***: p<0.001.
Figure 4.
Temporal evolution of the global effect on attentive trials at the beginning of the test phase.
Panels A and B represent an average of six sensors showing ERFs following the global standard stimulus during the exposure phase (blue) and during the test phase (red) for AT subjects (A) and DT subjects (B). The amplitude of late ERFs increased from the exposure phase to the test phase. Panel C and D represent the mean amplitude across subjects for the same sensors and time windows as in figure 3B for trials 1 to 10 (for display purposes, data points were smoothed using a moving average with a window of 1). During the test phase, the amplitude of late ERFs progressively increased for global standard stimuli (red) and decreased for global deviant (black) in AT subjects (C). By contrast, the amplitude of ERFs in DT subjects remained stable along the test phase (D), showing that the brain responses to global standard and deviant in that phase never differed.
Figure 5.
Influence of auditory violations on the visual detection task.
In each panel is represented the topography (inset) and the time course of a spatial filter selective of the activity induced by the visual task. The spatial filter was computed as follow: we first subtracted Attentive ‘auditory only’ trials from Distracted ‘audio-visual’ trials, resulting in ‘purely’ visual ERFs and averaged the topographies between 500 and 650 ms after the onset of the first visual stimulus. We applied this filter to global deviant (green lines) and global standard (black lines) trials in AT (A) and DT (B) subjects in order to test whether violations of the auditory rule influenced the visual activity. Black dots show significant paired t-tests performed at each time sample, comparing ERFs amplitudes for global standard to global deviant.