Deep super-resolved vibrational imaging via tomography-informed scattering compensation

Objectif

Optical microscopy has transformed biology, yet imaging intact 3D tissues remains challenging: Light scattering limits imaging beyond superficial layers, and fluorescent dyes – central to many techniques – are often incompatible with live samples.

We propose to develop DeepVibes: a super-resolving, deep-tissue microscope based on Stimulated Raman Scattering (SRS). Label-free and chemically specific, it will be capable of imaging millimetres deep into strongly scattering samples, such as histopathological tissues and living brain organoids.

Rather than using chemical clearing, we will undo tissue scattering in real time through kilohertz-rate wavefront shaping during raster scanning. The shaping will be guided by precise 3D digital maps of the specimen. Building on our expertise in nonlinear microscopy and computational imaging, we will construct the maps using cutting-edge computational techniques and a pioneering strategy to transform SRS signals directly into information about the specimen's scattering properties. This will allow us to focus light deep inside turbid samples, surpassing the depth and field-of-view limits of classical adaptive optics. We will also adapt a proven fluorescence super-resolution method to achieve super-resolved SRS microscopy.

As proof-of-concept we will test DeepVibes in two high-impact applications:
(1) rapid, label-free histopathological diagnosis of thick tissues;
(2) non-invasive monitoring of developing brain organoids.

DeepVibes will unlock millimetre-deep, nanoscopic, label-free imaging of living tissues – opening new frontiers in biology and medicine.