Imagerie volumétrique rapide pour la nanoscopie des cellules vivantes // Fast Volumetric imaging for Live cell nanoscopy: bringing super-resolution to life.

State of the art: The diffraction limit that has long restrained the observation of biological systems is being currently outpaced by new approaches combining optics and photophysical control of fluorescent emitters. Super-localization microscopy (dSTORM/PALM/PAINT), by constraining excited fluorophores to emit at different times, enables localization of molecules within 10nm lateral precision. NanoBio team proposes novel methods to enrich single molecule imaging, in particular for 3D imaging with new optical concepts (Nat. Photonics 2015,20211,2, Nat. Commun. 20193) enabling in-depth localization of molecules in complex samples (organoid, embryos). Part of these developments have been achieved in close collaboration with Abbelight, a French SME founded in 2016, which focuses on the development of super-resolution add-on transforming inverted microscopes into nanoscopes.

While single-molecule localization microscopy (SMLM) provides high spatial resolution, the actual challenge is to address live cell imaging where biological events occur on timescales exceeding the temporal resolution of SMLM, leading to blurred or inaccurate image reconstructions. We proposed an interdisciplinary PhD project, where a new optical implementation for excitation/detection of the fluorescence emission will be developed to permit volumetric imaging, along with the implementation of fast deep learning-based processing workflow. The final goal of these developments is to address biological questions, in particular on cancer cell migration into surrounding tissues during metastasis.

Objectives and Methodology: With our synergistic approach, we are aiming at addressing existing challenges in acquisition speed and 3D resolution, to enable observations of live cell processes with nanoscopic precision. Benefiting from the competitive research environment of supervisors, and the training and network provided by LIGHTinPARIS, the PhD candidate will be central for the development of pioneering instrument combining optical and image processing innovations.

Adapting the nanoscope to live imaging: we propose to further develop single shot volumetric imaging, in particular we will investigate the use of inclined excitation which permits to increase the observed volume without increasing the background which is mandatory for localization process, but also the implementation of alternative detection similar to Oblique Plane Microscopy (OPM)

Increasing acquisition speed: by combining the volumetric imaging capability of the OPM with deep learning algorithms, both the number of acquired frames and the needed signal to noise ratio will be reduced. Using our bank of images, machine learning methods will be deployed to generate “synthetic frames” and boost the reconstruction of imaged structures, without compromising the desired 3D resolution.

Applications : after calibration steps on reference samples, the new setup will be applied to key biological questions with close collaborators in biology, to decipher mechanism at the nanoscale such as the role of various proteins in cancer migration.
------------------------------------------------------------------------------------------------------------------------------------------------------------------------
------------------------------------------------------------------------------------------------------------------------------------------------------------------------

State of the art: The diffraction limit that has long restrained the observation of biological systems is being currently outpaced by new approaches combining optics and photophysical control of fluorescent emitters. Super-localization microscopy (dSTORM/PALM/PAINT), by constraining excited fluorophores to emit at different times, enables localization of molecules within 10nm lateral precision. NanoBio team proposes novel methods to enrich single molecule imaging, in particular for 3D imaging with new optical concepts (Nat. Photonics 2015,20211,2, Nat. Commun. 20193) enabling in-depth localization of molecules in complex samples (organoid, embryos). Part of these developments have been achieved in close collaboration with Abbelight, a French SME founded in 2016, which focuses on the development of super-resolution add-on transforming inverted microscopes into nanoscopes.

While single-molecule localization microscopy (SMLM) provides high spatial resolution, the actual challenge is to address live cell imaging where biological events occur on timescales exceeding the temporal resolution of SMLM, leading to blurred or inaccurate image reconstructions. We proposed an interdisciplinary PhD project, where a new optical implementation for excitation/detection of the fluorescence emission will be developed to permit volumetric imaging, along with the implementation of fast deep learning-based processing workflow. The final goal of these developments is to address biological questions, in particular on cancer cell migration into surrounding tissues during metastasis.

Objectives and Methodology: With our synergistic approach, we are aiming at addressing existing challenges in acquisition speed and 3D resolution, to enable observations of live cell processes with nanoscopic precision. Benefiting from the competitive research environment of supervisors, and the training and network provided by LIGHTinPARIS, the PhD candidate will be central for the development of pioneering instrument combining optical and image processing innovations.

Adapting the nanoscope to live imaging: we propose to further develop single shot volumetric imaging, in particular we will investigate the use of inclined excitation which permits to increase the observed volume without increasing the background which is mandatory for localization process, but also the implementation of alternative detection similar to Oblique Plane Microscopy (OPM)

Increasing acquisition speed: by combining the volumetric imaging capability of the OPM with deep learning algorithms, both the number of acquired frames and the needed signal to noise ratio will be reduced. Using our bank of images, machine learning methods will be deployed to generate “synthetic frames” and boost the reconstruction of imaged structures, without compromising the desired 3D resolution.

Applications : after calibration steps on reference samples, the new setup will be applied to key biological questions with close collaborators in biology, to decipher mechanism at the nanoscale such as the role of various proteins in cancer migration.
------------------------------------------------------------------------------------------------------------------------------------------------------------------------
------------------------------------------------------------------------------------------------------------------------------------------------------------------------

Début de la thèse : 01/10/2025

Programme COFUND LIGHTinPARIS

Université Paris-Saclay GS Physique

572 Ondes et Matière

-connaissances en optique  -connaissances en deep learning  -intéressé(e) par le developpement instrumental  -connaissance en python  -intéressé(e) par le travail interdisciplinaire 
-knowledge in optics  -interested in instrumental development  -knowledge in python  -knowledge in deep learning -interested in interdisciplinary work