strict warning: Only variables should be passed by reference in /home/howyourb/public_html/modules/book/book.module on line 559.
In lesson 1 we have seen how neurons can use electrical potentials and impulses to "encode" features of the outside world, and in lesson 2 we saw how they can combine information using converging synaptic connections. In the lesson in week 3 we shall look at how these mechanisms are at play to deliver visual information from the eyes to our brains, and in week 4 we shall look in some detail at how the brain then processes this information.
Key Concepts to Remember from Lesson 3:
Imagine light to be a stream of tiny "wave packets" (photons) which rush out from a light source at phenomenal speed, and which can penetrate transparent materials, but not opaque ones. When hitting opaque ones they either bounce back or get absorbed. In order to allow us to see, our eyes must collect the photons that bounced off the objects around us, and try to work out the properties of those objects from those reflected objects.
The eye largely works like a pinhole camera, and exploits the fact that photons travel in straight lines (rays) to "project" an orderly image of the outside world onto the retina at the back of the eye.
Each photon has a "wave length", and this could be anything from very large (e.g. several thousand meters for a "photon" of a longwave radio broadcast) to very tiny indeed (eg. ca a third of a nanometer (billionth of a meter) for the photon of a gamma ray). Only photons of a limited range of relatively short wavelengths ca (400-700 nanometers) are visible to the human eye because photons of these wavelengths are just "the right size" to be absorbed by photopigment molecules (opsins) in our eyes.
Each retina contains about 130 million photoreceptors (the eye is a "130 mega pixel camera") but about 120 million of those are highly light sensitive rod cells which are used during night vision only, leaving less than 10 million "cone cells" for daytime vision.
Rods and cones contain different types of photopigment molecules (opsins), which makes them sensitive to slightly different light wavelengts. Normal humans have cones containing one of three different types of cone opsin ('iodopsins'), whcih makes these cones sensitive to red, green or blue light respectively.
Like other neurons, rods and cones have high intracellular K+, high extracellular Na+, and a cell membrane that is polarized due to K+ leakage. However, unlike other neurons, these photoreceptors have Na+ leakage channels in their membranes which are open in the dark, so that Na+ can enter the cell (the 'dark current'), causing photoreceptors to be depolarized, and their synpases to be active, releasing glutamate.
When opsins in the photoreceptor cells are struck by photons, that sets off a biochemical cascade which closes the Na+ "dark current" channels, causing the cell to become more polarized, and its synapses to release less glutamate. The amount of glutamate released by a photoreceptor is therefore a measure of the amount of "darkness" that falls onto the pixel that the receptor cell occupies.
Although there are almost 10 million cones and about 120 million rods in the retina, the optic nerve, which carries information from the eye to the brain, has only about 1.3 million nerve fibers or so. The neural circuitry of the retina therefore has to achieve a form of "information compression", throw away some detail, convey only the important bits.
It achieves this through a network of bipolar, horizontal and amacrine cells which converges onto retinal ganglion cells (RGCs), and which uses lateral inhibition to create receptive fields with opposing centre-surround organization (ON-centre or OFF-centre cells). Such centre-surround receptive fields are effectively devices for measuring local contrast. If there is no local contrast (say you are staring at a uniform white piece of paper) then there is nothing to see.
Similarly "colour-opponent" receptive fields are created by centre surround interactions. These make cells sensitive to colour ("chromatic") contrast rather than brightness ("luminance") contrast.
Retinal ganglion cells (which carry visual information to the brain) come in 3 flavours:
M-cells project to the magnocellular layers of the lateral geniculate (LGN), are colour blind but highly sensitive to luminance contrast, and feed heavily into the brain's motion centres.
P-cells projectto the parvocellular layers of the LGN, are often red-green opponent, and project to the parts of the brain most interested in shape and fine detail.
Non-M non-P cells project to the koniocellular layers of the LGN, are often yellow-blue sensitive, and project into the parts of the brain most interested in colour.