Task relevance modulates the behavioural and neural effects of sensory predictions

Abstract

The brain is thought to generate internal predictions to optimize behaviour. However, it is unclear whether predictions signalling is an automatic brain function or depends on task demands. Here, we manipulated the spatial/temporal predictability of visual targets, and the relevance of spatial/temporal information provided by auditory cues. We used magnetoencephalography (MEG) to measure participants’ brain activity during task performance. Task relevance modulated the influence of predictions on behaviour: spatial/temporal predictability improved spatial/temporal discrimination accuracy, but not vice versa. To explain these effects, we used behavioural responses to estimate subjective predictions under an ideal-observer model. Model-based time-series of predictions and prediction errors (PEs) were associated with dissociable neural responses: predictions correlated with cue-induced beta-band activity in auditory regions and alpha-band activity in visual regions, while stimulus-bound PEs correlated with gamma-band activity in posterior regions. Crucially, task relevance modulated these spectral correlates, suggesting that current goals influence PE and prediction signalling.

Neural correlates of auditory pattern learning in the auditory cortex

“…Such difference suggests that within-sequence repetitions are not a requirement for the learning process, as long as the sequence contains salient information to be learned. Both effects in our study were observed mostly in the beta band, which has been implicated in sensory memory (Haenschel et al, 2000;Scholz et al, 2017) and sensory predictions (Auksztulewicz and Friston, 2016;Auksztulewicz et al, 2017;Pearce et al, 2010).…”

Content-based dissociation of hippocampal involvement in prediction

“…First, the two studies employed different behavioural tasks: participants performed orientation and contrast discrimination in Kok et al (2012a), whereas in the current study they discriminated grating phase. Although we do not have a clear hypothesis for how this would lead to differences in V1, previous work has shown that task demands influence expectation effects in visual cortex (50)(51)(52)(53). Additionally, there were differences in the grating presentations between the two studies.…”

No evidence that predictions and attention modulate the first feedforward sweep of cortical information processing

“…Higher cortical regions implement top-down predictions in hierarchically lower cortical regions though synchronization of activity in lower-frequency ranges (i.e., the alpha/beta band), whereas prediction violations are propagated from lower to higher cortical regions through synchronization of gamma band activity (Arnal & Giraud, 2012;Michalareas et al, 2016). Indeed, a recent MEG study found that invalid predictions increased gamma activity induced by task-relevant stimuli in V1, however, not until 130ms after stimulus presentation (Auksztulewicz, Friston, & Nobre, 2017). Another MEG study employing a probabilistic cuing task also reported stimulus-induced increases in gamma-band activity by attention, which, in contrast to Auksztulewisz et al, decreased as a function of stimulus predictability (Bauer, Stenner, Friston, & Dolan, 2014).…”