Search

The ability to generate temporal predictions is fundamental for adap-tive behavior. Precise timing at the time-scale of seconds is critical,for instance to predict trajectories or to select relevant information.What mechanisms form the basis for such accurate timing? Recentevidence suggests that (1) temporal predictions adjust sensoryselection by controlling neural oscillations in time and (2) the motorsystem plays an active role in inferring “when” events will happen.We hypothesized that oscillations in the delta and beta bands areinstrumental in predicting the occurrence of auditory targets. Partici-pants listened to brief rhythmic tone sequences and detected targetdelays while undergoing magnetoencephalography recording. Priorto target occurrence, we found that coupled delta (1–3 Hz) and beta(18–22 Hz) oscillations temporally align with upcoming targets andbias decisions towards correct responses, suggesting that delta–beta coupled oscillations underpin prediction accuracy. Subsequentto target occurrence, subjects update their decisions using the mag-nitude of the alpha-band (10–14 Hz) response as internal evidence oftarget timing. These data support a model in which the orchestrationof oscillatory dynamics between sensory and motor systems isexploited to accuratelyselect sensory information in time.Keywords: auditory, motor, neuronal oscillations, sensorimotor, timingIntroductionAccurately predicting when future events will occur facilitatessensory processing, speeds up behavior, and optimizes theallocation of attentional resources in time (Correa et al. 2005;Nobre et al. 2012). Predictive timing, by analogy to the notionof predictive coding (Knill and Richards 1996; Friston 2005)inthe time domain, requires the construction of an internalmodel of observed temporal regularities to infer precisely theoccurrence of future events. Because these regularities canhappen at different time-scales (seconds, hours, days…), dis-tinct neural systems and computations may be used to generatetemporally adaptive predictions (Ivry and Schlerf 2008; Moril-lon et al. 2009). The seconds time-scale, in particular, is highlyrelevant for online human behavior, as a large number of phe-nomena pertaining to perception and action (i.e., speech,movement etc.) occur at this scale. Here we investigate howthe brain extracts temporal regularities that emerge from theperception of isochronous beats at thistime-scale.The inclination to automaticallysynchronizeour movementsto external rhythms suggests that temporal regularities are par-ticularly relevant to the motor system (Schubotz et al. 2000;Grahn and Brett 2007; Bengtsson et al. 2009; Teki et al. 2011).This system is arguably at the origin of the contingent negativevariation (CNV), an anticipatory electrophysiological compo-nent that precedes relevant sensory or motor events (Walteret al. 1964; Pfeutyet al. 2003; Praamstra et al. 2006; Cravo et al.2011) and facilitates performance when task timing is predict-able (Hillyard 1973; Niemi and Naatanan 1981; Stefanics et al.2010). It is also involved in duration perception, in particularfor durations shorter than 2 s (Ivry and Schlerf 2008; Morillonet al. 2009). Recent considerations have suggested that themotor system internally simulates movement synchronizedwith future events to anticipate their occurrence and facilitatetheir processing (Schubotz 2007; Tian and Poeppel 2010;Arnal 2012; Arnal and Giraud 2012). According to this idea,corollary discharges that are contingent on movement simula-tion propagate to sensory areas to align ongoing activity withpredicted events. Here, we hypothesized that temporal predic-tions are instantiated through sensorimotor oscillatory interac-tions. Such a mechanism would arguably permit the extractionof temporal regularities in order to infer when an event shouldoccur.Cortical oscillations are usually seen as a means towardsflexibly communicating between distant neuronal populations(Engel et al. 2001; Fries 2005). Recent findings also suggestthat oscillations, which reflect fluctuations of neuronal excit-ability, can be temporally adjusted to optimize sensory selec-tion (Schroeder and Lakatos 2009). Accordingly, expectationsalign the phase of delta oscillations in time, which in turnaccelerates response timing (Lakatos et al. 2008; Stefanics et al.2010). Furthermore, ongoing delta phase modulates theweighting of sensory events in the decision-making processduring sequential information processing (Wyart et al. 2012;Cravo et al. 2013). Beta oscillations arguably play a comple-mentary function during temporal expectations and the accu-mulation of sensory evidence (Donner et al. 2009; Saleh et al.2011; Fujioka et al. 2012; de Lange et al. 2013).In sum, sensory selection (stimulus detection or sensoryweighting) can be passively or actively (through prediction)regulated through modulation of the prestimulus oscillatorystate. However, whether they govern subjects’ accuracy remainsunclear. That is, whether oscillations play an instrumental rolein determining whether an event occurs at the expected timehasnotbeenelucidated.This study aimed at determining the neurophysiologicalmechanisms underpinning (1) the generation of a temporalprediction and (2) the subsequent evaluation of this predic-tion. We used a delayed-target detection task in which subjectswererequiredtodetectwhetherthelasttoneofanisochronoussequence was delayed with regard to the beat (Fig. 1A). Ouranalysis design (Fig. 1B) allows us to distinguish how 2 crucialstages of the processing chain are implemented: (1) the pre-dictive stage, which reflects the neural activity patterns used topredict “when” and (2) the decisional stage, in which subjects’decisions (i.e., subjective reports) are made (see Materials andmethods and Fig. 1C).