Neural circuit leading to the W3 ganglion cell in the mouse retina (Zhang 2012).
The mammalian retina contains about 80 types of nerve cell, from the photoreceptors in the input layer to the ganglion cells in the output layer. The ganglion cells alone comprise about 30 different types. Each type tiles the visual field and is responsible for communicating one particular image feature to the rest of the brain. So we can think of the retina as 30 different visual processing circuits all interleaved in one network [1]. Our goal is to understand their function by answering these three questions: What are the features each ganglion cell computes [2-3]? How is that done with the circuit elements of the retina [4-5]? Why does Nature prefer these computations over others [6-7]? Given the unsurpassed experimental access that the retina offers to input signals, output signals, interneurons, and synapses, this has been a particularly fertile area of dynamic neuroscience.
