fr | en
Laboratoire Angevin de Recherche en Ingénierie des Systèmes

Separated by coma

IMAGIN Research project

Active molecular imagin and unmixing

 

Group :  Information, Signal, Image and Life Sciences

Labelling: none

Duration: 48 months (2021 - 2025)

Funding: ANR

Staff involved from LARIS: David Rousseau, Valentin Gilet (PhD student)

Project partners: LASIRe (Laboratoire Avancé de Spectroscopie pour les Intéractions la Réactivité et l'Environnement), IRIT (Institut de Recherche en Informatique de Toulouse), KU LEUVEN / Département de chimie Lab for nanobiology and l'Université de Liège / CIRM, vibra-santé hub, Laboratory of pharmaceutical analytical chemsitry

 

Abstract

Many microscopy techniques provide images for which contrast is brought by the measurement at each pixel of a molecular signature, like in multiplexed and hyperspectral imaging, or fluorescence lifetime imaging. For most biological/chemical systems, microscopy data can be decomposed into relatively few individual components whose characteristics (pure signatures and spatial distribution of relative concentrations) allow to model the intrinsic information encoded by the measurement. When multilinear analysis is concerned, a key insight is that it is not necessary to use all data points (e.g. spectral pixels) to perform this decomposition. It might even be that the decision on which pixels to acquire can be taken dynamically and spectral unmixing achieved on-the-fly. The IMAGIN project aims at demonstrating the feasibility and the relevance of such a breakthrough image acquisition and processing framework that do not jeopardizing chemical selectivity or spatial resolution. It will also show that resorting to this framework results in shorter measurement time, reduced data storage and, most importantly, minimized risk of potential photodamage. IMAGIN will pursue a result-oriented working plan addressing three main issues: (i) sparse targeted acquisition at essential spectral pixels, (ii) operational unmixing to extract the signatures of individual chemical components and (iii) image processing approaches to produce the concentration distribution maps associated to those components. These three tasks will be tackled by a unique methodological approach that will be validated and implemented in the scanning and widefield microscopy systems available in our laboratories. In addition, beyond proof-of-concept, the project aims at investigating challenging biological and biomedical systems, highly heterogeneous, and sensitive to phototoxicity, such as live cells, falsified medicines, or tissues obtained from preclinical studies so as to get interesting insights into the pathology of the patients under study.

Scroll