August 2019
Volume 60, Issue 11
Open Access
ARVO Imaging in the Eye Conference Abstract  |   August 2019
Label free retinal cell imaging with dynamic full-field OCT
Author Affiliations & Notes
  • Jules Scholler
    Institut Langevin, ESPCI Paris, PSL Research University, Paris, France
  • Kassandra Groux
    Institut Langevin, ESPCI Paris, PSL Research University, Paris, France
  • Mathias Fink
    Institut Langevin, ESPCI Paris, PSL Research University, Paris, France
  • Claude Boccara
    Institut Langevin, ESPCI Paris, PSL Research University, Paris, France
  • Kate Grieve
    Vision Institute, Paris, France
    Quinze Vingt National Ophthalmology Hospital, Paris, France
  • Footnotes
    Commercial Relationships   Jules Scholler, None; Kassandra Groux, None; Mathias Fink, None; Claude Boccara, None; Kate Grieve, None
  • Footnotes
    Support  European Research Council SYNERGY Grant scheme (HELMHOLTZ, ERC Grant Agreement # 610110)
Investigative Ophthalmology & Visual Science August 2019, Vol.60, 009. doi:
  • Views
  • Share
  • Tools
    • Alerts
      ×
      This feature is available to authenticated users only.
      Sign In or Create an Account ×
    • Get Citation

      Jules Scholler, Kassandra Groux, Mathias Fink, Claude Boccara, Kate Grieve; Label free retinal cell imaging with dynamic full-field OCT. Invest. Ophthalmol. Vis. Sci. 2019;60(11):009.

      Download citation file:


      © ARVO (1962-2015); The Authors (2016-present)

      ×
  • Supplements
Abstract

Purpose : Microscopic imaging of 3D live tissue samples is important in disease modeling applications and development of new therapies, and non-invasive methods such as dynamic full-field OCT (D-FFOCT) hold promise for removing the need for complex and invasive fluorescent labeling. However in order for non-invasive methods to offer an effective alternative to fluorescent labeling, identification of specific cell types and behaviors is an essential step. Our purpose was to perform real time imaging with D-FFOCT of the local fluctuations in 3D samples, to use the fluctuations as features to classify cells, label-free, and to validate against ground truth data.

Methods : We used D-FFOCT to measure subcellular organelle dynamics at up to 150Hz in living cells at 3D sub-micrometer resolution. We used a combination of i) ground truth fluorescent labeling and inhibitor drugs, imaged with simultaneous coincident fluorescence and real-time D-FFOCT acquisitions, to establish true cell identities and behaviors; ii) a machine learning approach to classify retinal cells using only the dynamic dimension of the signal, disregarding any morphological information.

Results : We tested our methods on 3D primate retinal explants and human induced pluripotent stem cell (iPS)-derived retinal organoids at various stages of development. Comparison with fluorescence ground truth allowed us to identify live and dead cells, and use of inhibitors allowed discrimination of which subcellular organelles (e.g. mitochondria, vesicles) were responsible for the D-FFOCT signal. The machine learning framework was successful in classifying different retinal cell types with 95% accuracy.

Conclusions : We demonstrated real-time quantitative D-FFOCT which allowed 3D reconstruction and time-lapse acquisitions without artifacts, and were able to validate our machine learning-based cell classification against ground-truth data. D-FFOCT with label-free cell classification offers the possibility to follow and control the evolution of the same sample, such as an in vitro retinal explant or a growing organoid for disease modeling or cell therapy applications, at different stages of its development.

This abstract was presented at the 2019 ARVO Imaging in the Eye Conference, held in Vancouver, Canada, April 26-27, 2019.

 

a) D-FFOCT image of a retinal organoid.
b) Retinal organoid with dead cells labelled : D-FFOCT image (left), overlay (right) of D-FFOCT and fluorescence (green) images.
c) First two images: segmentation on D-FFOCT images of ex vivo retina (ONL and INL), last two images: classification of cells.

a) D-FFOCT image of a retinal organoid.
b) Retinal organoid with dead cells labelled : D-FFOCT image (left), overlay (right) of D-FFOCT and fluorescence (green) images.
c) First two images: segmentation on D-FFOCT images of ex vivo retina (ONL and INL), last two images: classification of cells.

×
×

This PDF is available to Subscribers Only

Sign in or purchase a subscription to access this content. ×

You must be signed into an individual account to use this feature.

×