Phase alignment as a neural mechanism for crossmodal temporal prediction

990 

Abstract Submission 

Rebecca Burke¹, Jonathan Daume², Till Schneider¹, Andreas Engel¹

¹University Medical Center Hamburg-Eppendorf (UKE), Hamburg, Germany, ²Cedars-Sinai Medical Center, Los Angeles, CA, USA

Rebecca Burke  
University Medical Center Hamburg-Eppendorf (UKE)
Hamburg, Germany

Jonathan Daume  
Cedars-Sinai Medical Center
Los Angeles, CA, USA

Till Schneider  
University Medical Center Hamburg-Eppendorf (UKE)
Hamburg, Germany

Andreas Engel  
University Medical Center Hamburg-Eppendorf (UKE)
Hamburg, Germany

Our representation of time is embedded within multisensory perception, such as sight, sound or touch (Buzsáki, 2017). However, despite being a crucial aspect of daily life, the neural dynamics of crossmodal temporal predictions remain elusive. The objective of this study was to investigate neural correlates of tactile-to-visual influences on temporal prediction using Magnetoencephalography (MEG). We hypothesized increased inter-trial phase consistency (ITPC) in the low-frequency delta range [0.5-3Hz], corresponding to the length of the temporal prediction intervals. Additionally, stronger ITPC values should correlate with a steeper slope of the psychometric function, indicating phase alignments as a likely cause for more consistent temporal predictions.

The study was conducted within one MEG session employing a modified version of the time prediction task by Roth (2013) and Daume (2021). Participants (N=23) observed a visual stimulus moving towards an occluder. Shortly before reaching the occluder, the visual stimulus faded in luminance to make the visual offset less informative. Instead, participants received a brief tactile stimulus to the ipsilateral hand at the timepoint of disappearance, generating a temporal expectation regarding its reappearance on the opposite side of the occluder. After variable time intervals, a visual stimulus reappeared, and participants had to indicate whether this was "too early" or "too late" compared to the movement before disappearance. A non-predictive control condition involved participants judging the variable luminance of the reappearing visual stimulus compared to its initial luminance in the beginning of the trial. The order of conditions was randomized, and feedback was provided at the end of each block.
Psychometric curves were fitted to the behavioural data of each participant and condition, and MEG recordings were analysed using time-frequency representations obtained through wavelet convolution. To compare spectral power and ITPC estimates between conditions within frequency bands showing significant differences to the pre-stimulus baseline, we used cluster-based permutation statistics. Pearson's correlations were employed to examine the relationship between ITPC or power estimates and the steepness of each participant's psychometric function.

ITPC analysis revealed strong increases in the delta range [0.5-3Hz] around stimulus disappearance and reappearance. Delta ITPC was significantly stronger during temporal prediction compared to the control condition (cluster-p=.02) in time bins around 200ms to 900ms after disappearance, indicated by cluster-based permutation analyses. This cluster included sensors from right temporal and frontal regions, contralaterally to the visuo-tactile stimulus presentation. Spectral power analysis also showed significant increases in the delta band around movement onset, disappearance, and reappearance compared to baseline. However, delta ITPC, but not delta power, correlated with the the steepness of the psychometric curve (r=0.47, p=.01). Notably, this correlation was not observed in the non-predictive control condition.

Our findings suggest that the increase in delta ITPC is likely due to a phase reset driven by the temporal prediction process rather than evoked neural activity. Furthermore, our results indicate that phase alignments occur during crossmodal visuo-tactile-to-visual temporal predictions, even with a combination of non-rhythmic and discrete stimulation. This highlights the broad applicability of phase resets as a mechanism for predicting timing across various types of stimuli. Overall, this study provides valuable insights into the neural mechanisms involved in anticipating upcoming crossmodal events by elucidating the role of phase alignments.

Executive Function, Cognitive Control and Decision Making

Space, Time and Number Coding ¹

EEG/MEG Modeling and Analysis ²

Perception and Attention Other

Cognition

MEG

Other - Temporal Prediction; Crossmodal; Phase reset