According to this link,


retinal ganglion cells (RGCs) receive input from overlapping receptive fields (RFs). This is also an idea used in convolutional neural networks for deep learning of images.

On the other hand, this one, for the primate retina, states that there is no such overlap: http://www.sciencedirect.com/science/article/pii/0042698995001670

So do primate RGCs have overlapping RFs? If not, how does the primate visual system deal with this sparse input? How does it fill in the blanks?


1 Answer 1


Short answer
Associated ON- and OFF-center retinal ganglion cells can show 100% overlap in their receptive field.

When looking at the elementary neurophysiology of the retina, we can see that a single cone generally synapses onto two bipolars. Because photoreceptor cells hyperpolarize when illuminated, glutamate release from their synapse is inhibited. In turn, OFF-center bipolar cells are hyperpolarized and ON-center bipolars are depolarized. The OFF-center bipolar synapses onto an OFF-center retinal ganglion cell (RGC), while the ON-center synapses onto an ON-center RGC (Fig. 1).

Retinal circuitry linking cones to bipolar cells (left panel) and bipolar cells to retinal ganglion cells (right panel). source: Washington University.

Then to your question - the basic circuitry shows that the ON- and OFF-center RGCs receive their input from a single cone and hence have 100% overlapping field. In the foveal region, bipolars link 1:1 on cones, as in Fig. 1, so in this example case, they show 100% overlap.

Note that there are an estimated 30 types of retinal ganglion cells, about half of them being not even described yet, as per a 2015 review article (Sanes & Masland, 2015). Hence, this answer is anywhere from exhaustive, but it does show there can be substantial overlap between RGCs.

- Sanes & Masland, Ann Rev Neurosci (2015); 38: 221-46


You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .