Optimization of Functional MRI methods for olfactory interventional studies at 3T

• Main Author: Ahluwalia, Vishwadeep
• Format: Others
• Published: VCU Scholars Compass 2009
• Subjects: fMRI; olfaction; Medical Specialties; Medicine and Health Sciences; Radiology
• Online Access: http://scholarscompass.vcu.edu/etd/1953; http://scholarscompass.vcu.edu/cgi/viewcontent.cgi?article=2952&context=etd

Functional MRI technique is vital in investigating the effect of an intervention on cortical activation in normal and patient population. In many such investigations, block stimulation paradigms are still the preferred method of inducing brain activation during functional imaging sessions because of the high BOLD response, ease in implementation and subject compliance especially in patient population. However, effect of an intervention can be validly interpreted only after reproducibility of a detectable BOLD response evoked by the stimulation paradigm is first verified in the absence of the intervention. Detecting a large BOLD response that is also reproducible is a difficult task particularly in olfactory Functional MRI studies due to the factors such as (a) susceptibility-induced signal loss in olfactory related brain areas and (b) desensitization to odors due to prolonged odor stimulation, which is typical when block paradigms are used. Therefore, when block paradigms are used in olfactory interventional Functional MRI studies, the effect of the intervention may not be easily interpretable due to the factors mentioned above. The first task of this thesis was to select a block stimulation paradigm that would produce a large and reproducible BOLD response. It was hypothesized that a BOLD response of this nature could be produced if within-block and across-session desensitization could be minimized and further, that desensitization could be minimized by reducing the amount of odor by pulsing the odor stimulus within a block instead of providing a continuous odor throughout the block duration. Once the best paradigm was selected, the second task of the thesis was to select the best model for use in general linear model (GLM) analysis of the functional data, so that robust activation is detected in olfactory related brain areas. Finally, the third task was to apply the paradigm and model that were selected as the best among the ones tested in this thesis, to an olfactory interventional Functional MRI study investigating the effect of food (bananas) eaten to satiety on the brain activation to the odor related to that food. The methods used in this thesis to ensure valid interpretation of an interventional effect, can serve as a template for the experimental design of future interventional Functional MRI studies.