Output details
11 - Computer Science and Informatics
University of Birmingham
Compressive sensing based reconstruction in bioluminescence tomography improves image resolution and robustness to noise
<28>This paper describes a compressive-sensing based algorithm for the reconstruction of in vivo bioluminescence. There are two important advances: instrument noise is fully accounted for; and an adaptive sparsity weighting is used to account for the changing nature of the best solution. The method quantified bioluminescence source location up to 30% more accurately than a standard method (error: 1.1mm vs 1.7mm) using experimental data from a phantom sample. An earlier version of this work was the subject of a prize-winning student poster presentation by Basevi at the 2012 Optical Society of America Biomedical Topical Meeting.