Date of Graduation

5-2015

Document Type

Dissertation (PhD)

Program Affiliation

Medical Physics

Degree Name

Doctor of Philosophy (PhD)

Advisor/Committee Chair

Dianna Cody, Ph.D.

Committee Member

Dawid Schellingerhout, M.D.

Committee Member

Lucia LeRoux, Ph.D.

Committee Member

John Rong, Ph.D.

Committee Member

Donna Reeve, M.S.

Committee Member

Veera Baladandayuthapani, Ph.D.

Abstract

Calcific and hemorrhagic foci of susceptibility are frequently encountered on routine brain MR studies. Both etiologies cause variations in local magnetic field strength, leading to dark regions on the MR images that cannot be classified. Single-energy CT (SECT) can be used to identify lesions with attenuation over 100 HU as calcific, however lesions with lower attenuation cannot be reliably identified. While calcific lesions are unlikely to cause harm, hemorrhagic lesions carry a risk of subsequent intracranial bleeding; as such, identification of hemorrhage is vital in preventing the inappropriate use of anticoagulant medications in patients with hemorrhagic lesions.

Given there currently exists no clinically available means of differentiating between these two lesions over their full biological attenuation ranges, there is a clear need for a reliable imaging method to differentiate low intensity calcification and hemorrhage. Recently, several vendors have released new computed tomography (CT) scanner models with dual-energy capabilities, which may be successfully applied to this issue. By acquiring data at two different energies, dual-energy CT (DECT) collects information about the energy-dependent attenuation changes in a material and may help distinguish between two materials with similar linear attenuation measurements which would be impossible to differentiate using SECT.

This work applies the unique capabilities of DECT to the characterization of intracranial hemorrhage and calcification using biologically-relevant and spectrally-equivalent models. Lesion and acquisition parameters were varied to elucidate their impact on DECT’s ability to differentiate and subsequently classify intracranial lesions. DECT’s characterization ability was shown to improve with increasing CTDIvol, image thickness and lesion size. Using an optimized protocol, intracranial lesions were correctly classified with over 90% confidence down to a minimum attenuation of 56 HU, representing a significant improvement beyond the 100 HU limit imposed by the current standard. Since this data collection spanned several years, a dual-energy quality control program was designed to validate the comparison of collected data. The added characterization ability of DECT will assist physicians in the correct prescription of anticoagulant medications, hopefully sparing hemorrhagic patients from prophylaxis that might cause them harm.

Keywords

Medical Physics, Imaging, Computed Tomography, Dual-Energy CT, Hemorrhage, Calcification, Differentiation