A study on tensor and matrix models for super-resolution fluorescence microscopy

Description

Super-resolution techniques for fluorescence microscopy areinvaluable tools for studying phenomena that take place atsub-cellular scales, thanks to their capability of overcominglight diffraction. Yet, achieving sufficient temporal resolutionfor imaging live-cell processes remains a challenging prob-lem. Exploiting the temporal fluctuations (blinking) of fluo-rophores is a promising approach that allows employing stan-dard equipment and harmless excitation levels. In this work,we study a novel constrained tensor modeling approach thattakes this temporal diversity into account to estimate the spa-tial distribution of fluorophores and their overall intensities.We compare this approach with an also novel matrix-basedformulation which promotes structured sparsity via a continu-ous approximation of the cardinality function, as well as withother state-of-the-art methods.

Abstract

International audience

Additional details