Le vendredi 9 octobre à 14H à l’ENS – Salle Jean Jaurès –  24, rue Lhomond – 75005 Paris

Light control in scattering media and computational fluorescence imaging

In biological microscopy, light scattering represents the main limitation to image at depth.
Recently, a set of wavefront shaping techniques has been developed in order to manipulate coherent light in strongly disordered materials. The Transmission Matrix approach has shown its capability to inverse the effect of scattering and efficiently focus light. In practice, the matrix is usually measured using an invasive detector or low-resolution acoustic guide stars. In this thesis, we developed three different techniques based on linear fluorescence to reconstruct the transmission matrices, to and from a fluorescent object placed inside a scattering medium. The first approach consists of optimizing the spatial variance of low contrast speckle patterns. A second method extends this original idea of exploiting the low contrast fluorescent speckles. We developed a computational pipeline to better extract the physical information. It relies on demixing the fluorescent speckles using low rank factorizations and phase retrieval algorithms. It allows for reconstructing the transmission matrices to and from the object, and thus focusing and/or imaging (providing some memory effect) the object. A third technique provides a different and more comprehensive tool to retrieve the information. It is based on physics informed neural networks, that recently emerged as promising strategies in the field of imaging under challenging conditions.