Hemodynamic measurements and modeling for functional magnetic resonance imaging
MetadataShow full item record
In imaging, short wavelength (high-frequency) particles scattered from targets typically yield greater spatial resolutions than longer wavelengths. X-Rays, for example are typically within 2 orders of magnitude of a nanometer wavelength to achieve desired resolutions for medical imaging. Although better for imaging, this poses a health risk for subjects as ionizing radiation and this limits its use. Functional Magnetic Resonance Imaging (fMRI) avoids this issue by using radiation of much larger wavelengths, 4.8 m (62.5 MHz), that are relatively harmless. Instead of scattering, these photons are used to excite protons between spin-states in an external magnetic field. Magnetization relaxation rates and dephasing as a function of space and time are then measured to reconstruct images. This dissertation develops experimental methods to understand and interpret the biophysical underpinnings of fMRI in terms of blood flow and oxygen concentration changes. In neuroscience, fMRI may be used to deduce brain activity. Brain activity is a general term related to neuronal firing rate, which metabolizes oxygen. Deoxygenated blood increases proton spin dephasing. This is the physical mechanism that ultimately yields contrast in the fMRI signal. This is known as Blood-Oxygen Level Dependent (BOLD) contrast. A critical piece of information in this process, hemodynamics, is the dynamics of cerebral (brain) oxygen concentrations in relation to blood flow. The hemodynamics of BOLD contrast fMRI and its relation to brain activity is vital. In this dissertation, I have classified hemodynamic data as a function of space and time in cerebral cortex as well as testing a rudimentary hemodynamic model. I have taken fMRI measurements in three human subjects to identify spatial and temporal hemodynamic trends in brain. Furthermore, I've analyzed laser-speckle imaging in three subjects to identify spatiotemporal trends in blood speed. The final portion of this dissertation relates developments of a hemodynamic model of BOLD.