Closer look at the fMRI data analysis pipeline and its application in anesthesia resting state experiment

Abstract

The main focus of the thesis is the resting state fMRI data analysis, with much emphasis on the anesthesia fMRI experiments. Under this central topic, three separate themes are developed: resting state fMRI data analysis overview, improved denoising techniques, and application to the Dexmedetomidine experiment data. In the first part, important and confusing resting state data analysis steps are explored indepth, focusing on how and why the pipeline is different from that of the task-based fMRI. In the second part, the Principal Component Analysis (PCA) based denoising technique is introduced and compared against the conventional fMRI denoising techniques. Finally, with the PCA denoising technique, the functional connectivity of the brainstem with the brain is assessed for the Dexmedetomidine-induced unconscious subjects. We found that the functional connectivity between the Locus Ceruleus (LC) in the brainstem and the Thalamus & Posterior Cingulate Cortex (PCC) is the neural correlates of the Dexmedetomidine-induced unconsciousness.