Date of Award

January 2019

Document Type

Thesis

Degree Name

Medical Doctor (MD)

Department

Medicine

First Advisor

Zachary Corbin

Second Advisor

David Pitt

Abstract

Visual conscious perception occurs in less than one second. Using cerebral cortex electrical readings from human subjects during a behavioral visual task, we studied the spatiotemporal sequence of these events. Nine subjects, implanted with 100-300 subdural and depth electrodes in both hemispheres, were enrolled to complete the threshold perception task. The analysis investigated the role of broadband gamma power changes in electric potential in the range of 40-115 Hertz (Hz). In addition to implementing the behavioral task with the research subjects, my contribution focused on using clustering algorithms to analyze the broadband gamma power time-series to visualize the “switch and wave” model during conscious perception using surface maps. I applied K-means clustering to individual data and group data for Perceived trials, Not Perceived trials, and Combined baselines trials. Based on this analysis, we were able to ascertain vertices on the brain surface with comparable time-courses and classify them into anatomical clusters. Four regions emerged from k-means clustering within the cerebral cortex with comparable within-cluster gamma power z-score time-courses for Perceived stimuli. Cluster 1, localized in the visual and higher association cortices, showed initial increases, then decreases, and late secondary decrease. Cluster 2 showed sustained deactivation in the default mode network, indicating switching off during conscious stimulus processing. Cluster 3 and 4, localized in the visual association cortex, medial temporal lobes, and frontoparietal association areas, showed intermediate and late increases. For Non-Perceived stimuli, only Cluster 1, localized at the visual cortex, showed significance. Only Perceived stimuli displayed a wave of activity that traveled forward and throughout the cerebral cortex with large-scale network switching. I conducted novel k-means clustering analysis with combined baseline trials. These conclusions are crucial for comprehending the basis of normal conscious visual perception, which ultimately may have implications for tackling brain disorders.

Comments

This thesis is restricted to Yale network users only. It will be made publicly available on 06/01/2022

Share

COinS