Modeling the R2* relaxivity of blood at 1.5 Tesla
Metadata[+] Show full item record
BOLD (Blood Oxygenation Level Dependent) imaging is used in fMRI to show differences in activation of the brain based on the relative changes of the T2* (= 1/R2*) signal of the blood. However, quantification of blood oxygenation level based on the T2* signal has been hindered by the lack of a predictive model which accurately correlates the T2* signal to the oxygenation level of blood. The T2* signal decay in BOLD imaging is generated due to blood containing paramagnetic deoxyhemoglobin (in comparison to diamagnetic oxyhemoglobin). This generates local field inhomogeneities, which cause protons to experience different phase shifts, leading to dephasing and the MR signal decay. The blood T2* signal has been shown to decay with a complex behavior1, termed Non- Lorenztian, and thus is not adequately described by the traditional model of simple mono-exponential decay. Theoretical calculations show that diffusion narrowing substantially affects signal loss in our data. Over the past decade, several theoretical models have been proposed to describe this Non-Lorenztian behavior in the blood T2* signal in BOLD fMRI imaging. The goal of this project was to investigate different models which have been proposed over the years and determine a semi-phenomenological model for the T2* behavior using actual MR blood data.