Thesis Committee

Keith D. Paulsen, Ph.D. (Chair)

David W. Roberts, M.D.

Stephen C. Kanick, Ph.D.

Scott C. Davis, Ph.D.

Sylvain Gioux, Ph.D.

Abstract

Fluorescence-guided surgery (FGS) is a newer approach for providing intraoperative normal tissue to tumor contrast. While FGS has shown great promise as a navigational tool, it is hindered by a qualitative assessment of fluorescence. This assertion is based on the fact that current clinical methods of FGS rely on raw fluorescence emissions which can be distorted by background tissue autofluorescence, tissue absorption and scattering, and the presence of additional chromophores. These distortions may be most critical near the end of surgery, when remaining disease must be identified within a cavity that has been physiologically marred by surgical intervention. To correct for such distortions paired measurements of reflectance and fluorescence were used to estimate tissue optical properties and inform a fluorescence correction model. Data was acquired with both a localized point-probe system, with multiple sub-millimeter spacings, and a wide-field imaging system.

This work details the development, validation, and refinement of spectroscopic analysis algorithms for the two most commonly used fluorophores for FGS in brain. Protoporphyrin IX (PpIX) is an endogenous fluorophore that is part of the heme synthesis pathway, which selectively accumulates in malignant tissue after administration of 5-aminolevulinic acid (ALA), a non-fluorescent pro-drug, but has yet to pass FDA approval in the US. Fluorescein sodium (FS) is a vascular-targeted marker, which accumulates in areas of blood-brain barrier breakdown, making it far less selective but benefits from its low cost and ability to be used without specialized equipment. The presented work has three main focuses: (1) development and application of a subdiffuse reflectance model to improve optical property extraction from in vivo probe data, (2) validation and implementation of wide-field quantitative techniques for in vivo mapping of PpIX biodistribution, and (3) development and validation of spectral analysis algorithms for the small Stoke's shift fluorophore fluorescein sodium (FS), analogous to those that exist for PpIX. The work to be presented aims to address these concerns through a combination of Monte Carlo modeling, experimental measurements of tissue simulating optical phantoms, and clinical measurements/histology acquired as a part of ongoing clinical studies at Dartmouth-Hitchcock Medical Center.