Near-Infrared (NIR) spectroscopic tomography provides a means of imaging with high intrinsic contrast based upon tissue composition, as well as potentially in conjunction with exogenous agents for molecular-specific imaging. A particular strength of this imaging modality is that it can provide broadband spectral information along with high temporal resolution. In this project, we will further develop the tomographic approach to multidimensional imaging of the breast with the goal of identifying optimal ways to sample localized regions of tissue with rapid multi-spectral measurements. The strategy will use tomographic information as an initial guide to optimize the source-detector locations deployed for multidimensional spectroscopy with the goal of defining the methods which provide the optimal set of spatial, spectral and temporal information about tumor and normal breast tissue. A hybrid approach based on both frequency-domain and continuous wave spectroscopy is proposed to capitalize on the strengths of each method, while circumventing their respective limitations. The intended design will become a hybrid of the tomography methodology under development at Dartmouth and the spectroscopy scheme pioneered at UC Irvine. This project wil leverage existing NIR tomography facilities at the Dartmouth-Hitchcock Medical Center in order to test specific multidimensional spectroscopy hypotheses while also developing the additional algorithms (in conjunction with the Software Core) and hardware required to complete the proposed studies. Women recruited to participate in the ongoing Alternative Breast Imaging Modalities study at Dartmouth will be asked to participate in the additional spectroscopy procedures planned here, which will provide a sufficiently large pool of subjects for evaluation. In addition, a small animal NIR imaging system already exists and will be further developed as part of the Project 2 specific aims to study fundamentally new methods in multidimensional tomography using both (i) second-derivative spectroscopic methods for broadband tomography as well as, (ii) tomography in conjunction with cellular and molecular imaging agents. Specifically, we will examine the ability to image vascular and cellular targeted contrast agents in a rodent tumor model. In collaboration with the Hardware Core, we will also evaluate alternative approaches to multispectral tomography and refine the most promising methods which can be scaled up for human use, or which can be used robustly in basic research studies with experimental tumor models.
Showing the most recent 10 out of 74 publications
