Logo Logo
Help
Contact
Switch language to German
Cortico-hippocampal activations for high entropy visual stimulus: an fMRI perspective
Cortico-hippocampal activations for high entropy visual stimulus: an fMRI perspective
We perceive the environment around us in order to act upon it. To gain the desirable outcome effectively, we not only need the incoming information to be processed efficiently but we also need to know how reliable this information is. How this uncertainty is extracted from the visual input and how is it represented in the brain are still open questions. The hippocampus reacts to different measures of uncertainty. Because it is strongly connected to different cortical and subcortical regions, the hippocampus has the resources to communicate such information to other brain regions involved in visual processing and other cognitive processes. In this thesis, we investigate the aspects of uncertainty to which the hippocampus reacts. Is it the uncertainty in the ongoing recognition attempt of a temporally unfolding stimulus or is it the low-level spatiotemporal entropy? To answer this question, we used a dynamic visual stimulus with varying spatial and spatiotemporal entropy. We used well-structured virtual tunnel videos and the corresponding phase-scrambled videos with matching local luminance and contrast per frame. We also included pixel scrambled videos with high spatial and spatiotemporal entropy in our stimulus set. Brain responses (fMRI images) from the participants were recorded while they watched these videos and performed an engaging but cognitively independent task. Using the General Linear Model (GLM), we modeled the brain responses corresponding to different video types and found that the early visual cortex and the hippocampus had a stronger response to videos with higher spatiotemporal entropy. Using independent component analysis, we further investigated which underlying networks were recruited in processing high entropy visual information. We also discovered how these networks might influence each other. We found two cortico-hippocampal networks involved in processing our stimulus videos. While one of them represented a general primary visual processing network, the other was activated strongly by the high entropy videos and deactivated by the well-structured virtual tunnel videos. We also found a hierarchy in the processing stream with information flowing from less stimulus-specific to more stimulus-specific networks.
Hippocampus, entropy, fMRI, visual perception, partial least squares, independent component analysis, cortico-hippocampal networks, granger causality
Dalal, Nisha
2017
English
Universitätsbibliothek der Ludwig-Maximilians-Universität München
Dalal, Nisha (2017): Cortico-hippocampal activations for high entropy visual stimulus: an fMRI perspective. Dissertation, LMU München: Graduate School of Systemic Neurosciences (GSN)
[img] PDF
Dalal_Nisha.pdf

39MB

Abstract

We perceive the environment around us in order to act upon it. To gain the desirable outcome effectively, we not only need the incoming information to be processed efficiently but we also need to know how reliable this information is. How this uncertainty is extracted from the visual input and how is it represented in the brain are still open questions. The hippocampus reacts to different measures of uncertainty. Because it is strongly connected to different cortical and subcortical regions, the hippocampus has the resources to communicate such information to other brain regions involved in visual processing and other cognitive processes. In this thesis, we investigate the aspects of uncertainty to which the hippocampus reacts. Is it the uncertainty in the ongoing recognition attempt of a temporally unfolding stimulus or is it the low-level spatiotemporal entropy? To answer this question, we used a dynamic visual stimulus with varying spatial and spatiotemporal entropy. We used well-structured virtual tunnel videos and the corresponding phase-scrambled videos with matching local luminance and contrast per frame. We also included pixel scrambled videos with high spatial and spatiotemporal entropy in our stimulus set. Brain responses (fMRI images) from the participants were recorded while they watched these videos and performed an engaging but cognitively independent task. Using the General Linear Model (GLM), we modeled the brain responses corresponding to different video types and found that the early visual cortex and the hippocampus had a stronger response to videos with higher spatiotemporal entropy. Using independent component analysis, we further investigated which underlying networks were recruited in processing high entropy visual information. We also discovered how these networks might influence each other. We found two cortico-hippocampal networks involved in processing our stimulus videos. While one of them represented a general primary visual processing network, the other was activated strongly by the high entropy videos and deactivated by the well-structured virtual tunnel videos. We also found a hierarchy in the processing stream with information flowing from less stimulus-specific to more stimulus-specific networks.