Rhythmic temporal expectation boosts neural activity by increasing neural gain.

Temporal orienting improves sensory processing, akin to other top-down biases. However, it is unknown whether these improvements reflect increased neural gain to any stimuli presented at expected time points, or specific tuning to task-relevant stimulus aspects. Furthermore, while other top-down biases are selective, the extent of trade-offs across time is less well characterised. Here, we tested whether gain and/or tuning of auditory frequency processing in humans is modulated by rhythmic temporal expectations, and whether these modulations are specific to time points relevant for task performance. Healthy participants (N=23) of either sex performed an auditory discrimination task while their brain activity was measured using magneto- and electroencephalography (M/EEG). Acoustic stimulation consisted of sequences of brief distractors interspersed with targets, presented in a rhythmic or jittered way. Target rhythmicity not only improved behavioural discrimination accuracy and M/EEG-based decoding of targets, but also of irrelevant distractors preceding these targets. To explain this finding in terms of increased sensitivity and/or sharpened tuning to auditory frequency, we estimated tuning curves based on M/EEG decoding results, with separate parameters describing gain and sharpness. The effect of rhythmic expectation on distractor decoding was linked to gain increase only, suggesting increased neural sensitivity to any stimuli presented at relevant time points.SIGNIFICANCE STATEMENTBeing able to predict when an event may happen can improve perception and action related to this event, likely due to alignment of neural activity to the temporal structure of stimulus streams. However, it is unclear whether rhythmic increases in neural sensitivity are specific to task-relevant targets, and whether they competitively impair stimulus processing at unexpected time points. By combining magneto/encephalographic (M/EEG) recordings, neural decoding of auditory stimulus features, and modelling, we found that rhythmic expectation improved neural decoding of both relevant targets and irrelevant distractors presented and expected time points, but did not competitively impair stimulus processing at unexpected time points. Using a quantitative model, these results were linked to non-specific neural gain increases due to rhythmic expectation.