Cortical representation of sensory predictions investigated with multivariate pattern analysis and high-field fMRI
Final Report Abstract
This project investigated the neural mechanisms underlying predictions about the sensory outcomes of our own actions. In particular, the project aimed to 1) identify where in the human brain these predictions originate, 2) test whether stimulus visibility determines whether actionbased predictions lead to suppression or enhancement, and 3) investigate which aspects of action-based predictive mechanisms are based on bottom-up processes and which on top-down processes using high-field fMRI. In all work packages, similarities and differences with predictions about externally generated stimuli were studied by including both active and passive conditions. Due to the Covid-19 pandemic and recurrent technical issues at the MRI scanner, the project got severely delayed. As a result, the third study investigating feedforward and feedback processes using high-field fMRI could not be conducted, and analyses of the first two studies are still ongoing. Preliminary fMRI results of the first study show that predictions about the orientation of predicted upcoming visual stimuli pre-activated visual cortex prior to stimulus presentation, and that this preactivation carried information about the orientation of the predicted upcoming stimulus. There were no significant differences between active and passive conditions, suggesting similar predictive mechanisms. During stimulus presentation, increased activity in visual cortex could be observed for active compared to passive conditions, but decoding accuracies did not differ, suggesting similar stimulus representations across conditions. Behavioural results showed no difference in perceptual performance across conditions. The second study manipulated stimulus visibility by presenting low contrast visual stimuli either with or without added noise. Behavioural results showed that participants were better at discriminating the orientation of the stimuli when they elicited the stimuli themselves by button press, compared to when the stimuli were presented automatically. This effect was more pronounced in the low visibility (noise) condition, though this interaction did not reach significance. Preliminary fMRI results showed a significant interaction between action and visibility in visual cortex, with increased activity for actively generated stimuli only when stimulus visibility was low. In sum, preliminary results suggest that predictions about self-generated and externally generated visual stimuli are both represented in visual cortex prior to stimulus presentation. Furthermore, stimulus processing seems to be modulated by stimulus visibility, with enhanced neural activity and perception for actively generated stimuli compared to externally generated stimuli when visibility is low.
Publications
-
Different contributions of efferent and reafferent feedback to sensorimotor temporal recalibration. Scientific Reports, 11(1).
Arikan, Belkis Ezgi; van Kemenade, Bianca M.; Fiehler, Katja; Kircher, Tilo; Drewing, Knut & Straube, Benjamin
-
Non‐stimulated regions in early visual cortex encode the contents of conscious visual perception. Human Brain Mapping, 43(4), 1394-1402.
van Kemenade, Bianca M.; Wilbertz, Gregor; Müller, Annalena & Sterzer, Philipp
-
The effect of self‐generated versus externally generated actions on timing, duration, and amplitude of blood oxygen level dependent response for visual feedback processing. Human Brain Mapping, 43(16), 4954-4969.
Kavroulakis, Eleftherios; van Kemenade, Bianca M.; Arikan, Belkis Ezgi; Kircher, Tilo & Straube, Benjamin
