Where and When? : The oculomotor signature of spatio-temporal prediction

Details

Supervisors

Missal, Marcus

Degree label

Master [120] en sciences biomédicales, à finalité approfondie

Abstract

The human brain could be described as a "prediction machine". The ability to use past experiences, information about our current state and the surrounding environment to predict the future is crucial. It allows us to increase the chances of achieving desired outcomes while avoiding or preparing for future adversity. This ability is directly related to our level of uncertainty about the future: where and when an event will occur. Uncertainty could be described as "expected surprise”. For example, temporal surprise occurs when an anticipated event starts with variable timing. Similarly, spatial surprise can be described as an uncertainty about the event location. The central objective of this project was to investigate the interaction between two sources of surprise (spatial and temporal) through the analyses of eye movements. We hypothesised that temporal and spatial surprise will modulate each other when the future location of a stimulus is known. This hypothesis was tested in humans using a saccadic reaction time task in which cues indicated the future stimulus position and its onset time. Results were partially in accordance with the supported hypothesis. Indeed, this conditional processing depending on the certainty of a spatial information has been detected within a specific type of saccade: anticipatory saccades. Furthermore, results claimed distinctiveness between cognitive processes underlying visually-guided and anticipatory saccades when facing spatial and temporal surprise. This dichotomy of processes for the spatial and temporal surprises suggest that anticipatory saccades and visually-guided saccades use information differently. These results broaden the knowledge about the functioning of these two systems suggesting an interaction of the spatial and temporal information that depends on the surprise level when trying to anticipate a future event and an independence of these two when actually making an action toward a certain goal. The implication of this study has been illustrated by two conceptual models of optimal saccade execution that provide an interesting opening for further research. Specifically, future experiments are needed to test the modelled behaviours in the context of impulsivity-related and anticipation dysfunctions.