Novel Experimental and Analysis Paradigms for Neuroimaging
Type of DegreePhD Dissertation
Electrical and Computer Engineering
MetadataShow full item record
Functional magnetic resonance imaging (fMRI) technique uses the blood-oxygen-level-dependent (BOLD) contrast to detect associated changes in blood flow. The BOLD signal is temporal correlated with cerebral electrophysiological activities. Therefore, the recording and decoding of the BOLD signal provides insight into human brain function and benefits enormous amount of brain cognitive studies in scientific and medical interest. Although fMRI has become the primary application in the field of brain cognitive science, elusive and enigmatic questions remain. The scope of this dissertation lies in the construction of novel signal processing methods and experimental paradigms with fMRI applications, which devotes to the strategies of decoding brain signals and states. In the first and second studies, we addressed the issue that the fMRI is an indirect measure of neural activity. Since the BOLD signal can be modeled as a convolution of the hemodynamic response function (HRF) and latent neural activity and HRF varies across both between individuals and different brain regions within an individual. Therefore, the correlation acquired from the BOLD-level data could lead to false inferences of functional connectivity. The aberrant neurochemical mechanism which control the shape of the HRF have been reported in autism spectrum disorders (ASD), schizophrenia (SZ) and bipolar disorder (BP). Therefore, we hypothesized that these aberrations would lead to differences in the shape of the HRF between these pathological populations and healthy controls, and the alterations would contribute to the differences in estimating functional connectivity in the BOLD space as compared to the latent neural space. We reported that raw fMRI data failed to detect group differences in connectivity analysis when compared that with deconvolved data. Our results are relevant for the understanding of hemodynamic and neurochemical aberrations in pathological groups. In the third study, we applied a novel real-time fMRI (rt-fMRI) neurofeedback technology to investigate the insight problem solving. Inspired by a transcranial direct current stimulation (tDCS) study, we hypothesized that rt-fMRI neurofeedback could enable subjects to up-regulate activity in their right Anterior Temporal Lobe (rATL) using neurofeedback, and it could mimic the effects of tDCS in facilitating subjects to solve the nine-dot puzzle. Our results show that approximately 40% of subjects were able to solve the problem using rt-fMRI neurofeedback, which is similar to the percentage of subjects who were able to solve the puzzle using tDCS. Our results indicated that the group that successfully solved the problem were able to up-regulate the activity over rATL through rt-fMRI neurofeedback while the group that didn’t solve the problem were not able to. Further, we contrasted the brain activation and network obtained from two groups to investigate the neural bias generated by up-regulation of activity in rATL. Our results provideda putative neural mechanisms underlying nine-dot problem solving, and a possible explanation for the top-down inputs into the rATL in enhancing or suppressing creative insight. Furthermore, our study demonstrates that neurofeedback could potentially be used to mimic effects similar to brain stimulation techniques such as tDCS.