A deeper look into visual cognition: low dimensionality in visual exploration dynamics and memory retrieval

The general aim of this project is to characterize the relationship between spontaneous eye movement patterns and spontaneous brain activity. This is rooted on two main hypotheses: 1) apparently complex and variable behaviours are in part driven by low dimensional intrinsic dynamics; 2) these intrinsic behavioural dynamics match intrinsic neural data (i.e., resting-state). In the first chapter, we review the hypothesis of low dimensionality in behaviour. We show several examples of studies that show that data reduction in low dimensional components commonly occurs in cognitive data, also when measured with high resolution signals. In the second chapter, we show that visual exploration is low dimensional and driven by intrinsic factors. In the following chapter, we show that these behavioural phenotypes correlate with spontaneous brain activity (as measured with hd-EEG) recorded one year later. Finally, in the fifth chapter, we describe a behavioural correlate of free visual exploration: visual memory recall. By transforming audio recordings in 2D maps with a novel method, we show that there is an overlap between gaze and memory topography. Gaze fixations outperform purely stimulus-driven models, and this unbalance varies in clusters of participants.

The general aim of this project is to characterize the relationship between spontaneous eye movement patterns and spontaneous brain activity. This is rooted on two main hypotheses: 1) apparently complex and variable behaviours are in part driven by low dimensional intrinsic dynamics; 2) these intrinsic behavioural dynamics match intrinsic neural data (i.e., resting-state). In the first chapter, we review the hypothesis of low dimensionality in behaviour. We show several examples of studies that show that data reduction in low dimensional components commonly occurs in cognitive data, also when measured with high resolution signals. In the second chapter, we show that visual exploration is low dimensional and driven by intrinsic factors. In the following chapter, we show that these visual exploration phenotypes correlate with spontaneous brain activity (as measured with hd-EEG) recorded one year later. Finally, in the fifth chapter, we describe a behavioural correlate of free visual exploration: visual memory recall. By transforming audio recordings in 2D maps with a novel method, we show that there is an overlap between gaze and memory topography. Gaze fixations outperform purely stimulus-driven models, and this unbalance varies in clusters of participants.