Laser Microbeam and Medical Program    »     People  |   Education & Dissemination  |   Research & Resources  |   Publications
LAMMP Seminar Video
Intrinsically-Regularized Data Processing Algorithms for Fluorescence Molecular Tomography
Vivian Pera, PhD Candidate

Fluorescence molecular tomography (FMT) is an optical technique that uses near-infrared light to perform quantitative, three-dimensional imaging of fluorophores in whole animals noninvasively. It is becoming an important tool in preclinical imaging of small animals and has been employed to image tumors and assess response to anti-cancer therapeutics. Due to the highly scattering nature of light propagation in tissue, however, FMT is an ill-posed inverse problem that requires regularization to solve. This is usually accomplished by adding a regularization term to the cost function used to compute the FMT image, which makes the resulting image highly sensitive to the value of the regularization parameter as well as the assumptions used to regularize the problem. In this work, we attempt to sidestep this dependency on a regularization parameter by employing algorithms that are intrinsically regularized. In other words, regularization is accomplished only by physically relevant assumptions (e.g., nonnegativity of optical signals) without relying on user-specified parameters. For this we adapt classical and recent signal processing approaches from the radar and computer vision fields. We use these approaches to develop intrinsically-regularized data processing algorithms for two small-animal FMT instruments under construction in our lab. The first is a "diffuse fluorescence flow cytometer" that can detect and localize very rare fluorescently-labeled circulating cells in the limb of a mouse. The second is a time-resolved hyperspectral tomographic imaging system designed to enable robust demixing and localization of at least four concurrent fluorophores.

LAMMP Seminars

View selected seminars online

Supported by


P41EB015890