Modeling the mosaics formed by the somata of ON- and OFF-center X retinal ganglion cells

X. J. Zhan, John B Troy

Research output: Contribution to journalArticlepeer-review

Abstract

Purpose. The visual image is encoded at the retina in the pattern of activity across the arrays of different ganglion cell classes. The locations of the ganglion cells hence form the points at which the image is defined. A model which accurately characterizes the mosaic of X-cell locations is proposed here and tested against a real X-cell mosaic. Methods. Real ON-center and OFF-center er X-cell soma mosaics were provided by H. Wässle. Simulated mosaics were generated by 1) superimposing two equilateral triangular lattices, 2) shifting all lattice points a constant fixed amount, and 3) jittering each point randomly with statistics following a bivariate normal distribution. Movement of cells in the simulated mosaic was constrained only by soma size. We characterized a simulated mosaic by the fixed spatial shift and the jitter's standard deviation. Correspondence between real and simulated mosaics was assessed by Voronoi-Delaunay analysis. Results and Conclusions. Voronoi-Delaunay analysis revealed that real ON-center and OFF-center X-cell mosaics are statistically indistinguishable. The model lattice simulated the X-cell lattice best when the jitter's standard deviation was 23% of the spacing between elements in the starting triangular lattice and when the spatial shift between ON- and OFF-center lattices was half the spacing between the triangular lattice points.

Original languageEnglish (US)
JournalInvestigative Ophthalmology and Visual Science
Volume37
Issue number3
StatePublished - Feb 15 1996

ASJC Scopus subject areas

  • Ophthalmology
  • Sensory Systems
  • Cellular and Molecular Neuroscience

Fingerprint Dive into the research topics of 'Modeling the mosaics formed by the somata of ON- and OFF-center X retinal ganglion cells'. Together they form a unique fingerprint.

Cite this