Comparison of two image reconstruction algorithms for early-photon fluorescence molecular tomography
One of the most effective approaches for improving the spatial resolution of fluorescent tomograms is to apply the state-of-the-art reconstruction algorithms based on compressed sensing (CS) theory. The paper presents a numerical experiment on fluorescent image reconstruction where we compare two CS algorithms: the algebraic reconstruction technique (ART) with total variation regularization, ART-TV, and the ART with fast iterative shrinkage-thresholding, ART-FIST. For modeling measurement data and constructing the sensitivity matrix we assume a virtual fluorescence tomograph with a scanning fiber probe that illuminates and collects light in reflectance geometry with source-receiver separations corresponding to the macroscopic regime. We show that ART-FIST outperforms ART-TV in reproduction accuracy and is capable of achieving as good spatial resolution as 0.2 mm at depths to 9 mm inclusive, and is quite competitive in this sense with mesoscopic fluorescence tomography.
File with abstract
Alexander B. Konovalov
Federal State Unitary Enterprise “Russian Federal Nuclear Center – Zababakhin All-Russia Research Institute of Technical Physics”
File with report