Login / Signup

Intuitive representation of photopolarimetric data using the polarization ellipse.

Yakir Luc GagnonNicholas Justin Marshall
Published in: The Journal of experimental biology (2016)
Photopolarimetry is the spatial characterization of light polarization. Unlike intensity or wavelength, we are largely insensitive to polarization and therefore find it hard to explore the multidimensional data that photopolarimetry produces (two spatial dimensions plus four polarization dimensions). Many different ways for presenting and exploring this modality of light have been suggested. Most of these ignore circular polarization, include multiple image panes that make correlating structure with polarization difficult, and obscure the main trends with overly detailed information and often misleading colour maps. Here, we suggest a novel way for presenting the main results from photopolarimetric analyses. By superimposing a grid of polarization ellipses onto the RGB image, the full polarization state of each cell is intuitively conveyed to the reader. This method presents linear and circular polarization as well as ellipticity in a graphical manner, does not require multiple panes, facilitates the correlation between structure and polarization, and requires the addition of only three novel colours. We demonstrate its usefulness in a biological context where we believe it would be most relevant.
Keyphrases
  • machine learning
  • deep learning
  • healthcare
  • single cell
  • artificial intelligence
  • cell therapy
  • bone marrow
  • social media