Studying the presence, organization and interaction of multiple proteins within intact biological tissue is a challenging but highly rewarding field of fundamental research and medical diagnosis. Indeed, in situ molecular knowledge unlocks the capability to understand the nature and function of living matter. Working with pristine living tissues guaranties the reliability of the conclusions since the 3D structuration, microenvironment and fluid circulation are preserved. For accurate biological conclusions, it would thus be key to identify multiple kind of molecules simultaneously while preserving the thick sample absolutely unspoiled and alive.
Optical imaging is the best compromise to have a molecular sensitivity while ensuring spatiotemporal resolution and limited sample invasiveness. Reaching a molecular identification requires to optically harvest specific signature from the targeted molecules. This means that (a) the probed signal should vary with different targeted molecules; (b) the spatial resolution should be high enough to isolate specific region containing the specific pools of molecules.
SPECIPHIC key objective is thus to simultaneously identify multiple bio-molecules deep into living unlabeled biological tissue. To achieve this objective, the proposed paradigm consists in: 1) performing with a super-resolution label-free measurement, and 2) deciphering molecular signatures within this enhanced signal with machine learning. This should unlock unprecedented possibilities for accurate studies in unmodified samples in a complete stealthy manner with applications target in fundamental biology as well as medicine.