Visual processing begins in the retina and that information eventually moves its way back to the visual cortex, after first passing through the lateral geniculate nucleus (LGN). The processing of the visual image begins in the retina and is processed similarly in the LGN, but the processing that occurs in the in the visual cortex is far different. This made it difficult to discover how visual cortical neurons processed visual images. This article deals with visual processing in the retina, the LGN and how the experiments by David Hubel and Torsten Wiesel led to the discovery of the two main visual cortical neurons (simple and complex cells) and how they process vision. The processing of the visual image after leaving the visual cortex will not be discussed.

Visual Pathway from Retina to Cortex
Visual processing begins in the retina. Light or darkness hits the retina after first traveling through the cornea, the aqueous humor, lens and vitreous humor (Figure 1).

external image hdc_0001_0002_0_img0114.jpgFigure 1
Light travels to the back of the retina where the photoreceptors are located. The photoreceptors are located towards the back of the retina next to the pigment epithelium (Figure 6) because the disks as well of the photopigment needs to be recycled (Purves, 2004). Had the photoreceptors been located far away from the pigment epithelium, this task would be much more complicated. If the photoreceptor is depolarized, it will release the neurotransmitter. There is no action potential in the photoreceptors, bipolar, or horizontal cells. This is because an action potential is not needed as the distance between cells is not great (Kandel, 2000). If the bipolar cells are depolarized, which depends on their interaction with neurotransmitters, they will release neurotransmitter onto the ganglion cells. All ganglion cells exit the back of eye at the optic disk. Since this region contains only ganglion cells, there are no photoreceptors or bipolar cells. This is where the blindspot is located as no vision can occur here as there are no photoreceptors to detect light. The optic disk becomes the optic nerve after exiting the eye. Half of the optic nerve fibers, those that are from the nasal hemiretina, cross as the optic chiasm (Figure 3).
figure_10_29_labeled.jpgFigure 2

bvis1.gifFigure 3
Half of the optic nerve fibers, those that are from the nasal hemiretina, cross as the optic chiasm (Figure 3). Images projected onto the right temporal retina, are from the left side of the visual field. This mean that these same images are seen by the left nasal retina. These images eventually need to terminate at the same location, and the crossing at the optic chiasm allows for this. This new group of fibers are called the optic tract, and contain the same visual images (Figure 4). The optic tract terminates in the lateral geniculate nucleus (LGN) of the thalamus. The LGN has projections, called optic radiations, back to the visual cortex. This is where the visual cortical neurons are located.
1204_Optic_Nerve_vs_Optic_Tract.jpgFigure 4

The retina contains two types of photoreceptors. Rods and cones. Both of these photoreceptors have the same response to light, which will be discussed shortly. The rods are more sensitive to light than cones are, because the cones contain less photopigment than the rods (Kandel, 2000). Comparing the two, it takes 100 photons of light in a cone to have a comparable affect as 1 photon on a rod (Purves, 2004). The fovea is different from the rest of the retina because the fovea has an extremely high density of cones. The cones in the fovea have less convergence, meaning there is one ganglion cell, for each bipolar cell, for each cone, which results in higher levels of spatial acuity. Bipolar cells and ganglion cells will be discussed shortly. This is why when focusing on an object, the fovea is directed at that object. The rods would not allow for high levels of spacial acuity. Rods tend to be more densely populated in the periphery of the retina and have higher rates of convergence leading to less spacial acuity (Purves, 2004).
external image retina-layout.jpegFigure 5

As light hits the photoreceptors, there is a change in the shape of the photopigment. This change activates transducin, which activates phospodiesterase. Phosphodiesterase deactivates cyclic GMP (cGMP) which causes the sodium channels to close. The photoreceptor is hyperpolarized and no longer releases any neurotransmitter on the bipolar cell (Purves, 2004). This discovery was made by a Japanese neurophysiologist, and was extremely unexpected. The thought that photoreceptors are depolarized in the dark went against conventional thought (Hubel, 1995).
Screen shot 2013-12-02 at 3.40.12 PM.pngFigure 6
Purves, Dale. "Central Visual Pathways." Neuroscience. 3rd ed. Sunderland, MA: Sinauer, 2004. 259-80. Print.

Ganglion Cells
Receptive Field
After the anatomy of the eye was known, experiments were conducted to see how ganglion cells responded to light. Scientists struggled to find any interesting evidence as they used diffuse light, which resulted in little or no response from a ganglion cell. It was Stephen Kuffler's work while at Johns Hopkins that determined that there are on-center and off-center receptive fields. Instead of using diffuse light, Kuffler was able to fill the retina with diffuse light, or hit certain parts of the retina with more intense bursts of light, while still providing light to the rest of the retina. Kuffler's animal of choice was a cat, which turned out to be an ideal mammal as it's retina is more simple than other mammals or fish used in the past. Kuffler stuck an electrode through the sclera and into the retina and on a ganglion cell. By shining and moving a small spot of light at a spot on the retina where the electrode was located, he could change the amount that the ganglion cell would fire. The area that the light was shone that elicited a response from that ganglion cell was determined to be the receptive field. Kuffler found that there was a center and a surround portion of a ganglion cell, because as light moved towards the outside of the receptive field, there was a different response in the ganglion cell. In an on-center ganglion cell, the ganglion cell fired more when light was shone in the center of the receptive field, and most when the entire center was filled with light with darkness in the surround. As the light moved towards the surround of the receptive field, the ganglion cell fired less (Figure 7). When the entire receptive field was filled with light, there was almost no response form the ganglion cell, which is the reason why scientists had struggled while studying vision in the past. They filled the entire receptive field with light, not knowing that a receptive field contained a center and surround orientation. Diffuse light was thought to be the optimal stimulus, which is now known to be far from the truth. An off-center ganglion cell was found to fire most when the center was filled with darkness and least when the surround was filled with darkness. This is the opposite of the on-center ganglion cell. Why is it that off-center cells exist? It is thought that are just as many dark objects as there are light objects in the world. If only the light objects could be detected, we would be missing out on lots of visual stimuli (Hubel, 1995).

Ganglion cell #2.jpg Figure 7

Bipolar Cell Receptive Fields
After discovering that ganglion cells contained a center and surround organization, the next task was to discover if bipolar cells also had center and surround receptive fields. They found that they did, but that horizontal cells played a crucial role in determining the size of the surround field. Horizontal fields are part of the indirect pathway, named so because it does not synapse on ganglion cells directly. Horizontal cells are synapsed on by photoreceptors and are capable of synapsing back onto photoreceptors from a different area (Figure 8). Horizontal cells are hyperpolarized by light, similar to bipolar cells and are depolarized by darkness. While bipolar cells can have connections with one or up to a few photoreceptors, their branching is not widespread enough to account for the entire center-surround receptive field. Horizontal cells are capable of spanning across the entire receptive field. When horizontal cells are excited, they inhibit the photoreceptors in the center receptive field. The ganglion cell does not prefer to have the entire receptive field covered with light or darkness, and prefer means that it will fire less than at its maximum frequency. The ganglion cell was found to fire most when there was a different amount of stimulus in the center and surround. Incorporating the horizontal cells into the equation results in the bipolar cell receiving the information from photoreceptors a far distance away without having a direct synapse with them. For example, if this is an on-center bipolar cell, it will prefer light in the center and darkness in the surround. If there is darkness in the surround, the photoreceptors will be depolarized, leading to a release of neurotransmitter onto the horizontal cell. The horizontal cell will be depolarized. This horizontal cell will then synapse on the photoreceptors from the on-portion of the receptive field. Horizontal cells inhibit whatever they synapse on, so the photoreceptors will be inhibited, or in other terms hyperpolarized. If there is already light shining on the center portion, the photoreceptor will be hyperpolarized by the light and also hyperpolarized by the horizontal cell. This photoreceptor will release extremely little neurotransmitter, leading to an increased firing rate in the bipolar cell and ganglion cell (Hubel,1995).

Figure 8

Receptive Field
Hubel found that the cells in the LGN had similar receptive fields as retinal ganglion cells. Why does the LGN exist in the first place then? The LGN does not receive input only from the optic tract, but from the visual cortex, the reticular formation, and the LGN has fibers that synapse on other areas of the LGN. The input from the optic tract becomes modified before traveling back to the visual cortex (Heeger, 2006).
Slide9.jpgFigure 9

Layers of LGN
The LGN has 6 layers (Figure 10). Each layer receives input from a single eye, so each LGN contains 3 layers receiving input from the ipsilateral eye and 3 layers receive input from the contralateral eye. There are 4 parvocellular layers, receiving input from small ganglion cells, and there are 2 magnocellular layers, receiving input from large ganglion cells. Parvocellular layers receive input regarding form/shape and color while magnocellular layers receive input regarding motion. Layer 1 receives input from the contralateral eye, 2 ipsilateral eye, 3 ipsilateral, 4 contralateral, 5 ipsilateral, 6 contralateral. The LGN gets this type of input since the contralateral eye has fibers that decussate at the optic chiasm (Kandel, 2000).
Screen shot 2013-12-16 at 3.14.20 PM.png

Figure 10
Kandel, Eric R. "Central Visual Pathways." Principles of Neuroscience. 4th ed. New York:McGraw-Hill, 2000. 529-38. Print.

The LGN has projections to the visual cortex through the optic radiations (Figure 11). The optic radiations terminate in area 1 or V1 (Kandel, 2000).
vnov072i19.jpgFigure 11


Figure 12

Visual Cortical Neurons
When visual cortical neurons were first studied in the 1950's, there was little knowledge. An electrode was placed in the visual cortex and diffuse light was shone on the retina. There was little response in the visual cortical neurons because diffuse light is not the preferred stimulus in the retina so there was also no response in V1. It was later discovered that what little activity they were able to see in V1 was not actually coming from V1 cells but from the optic radiations leaving the LGN. Hubel and Wiesel then discovered after many hours of experimenting that cells in V1 responded when edges were present in the visual field. They found that the orientation of the edge and what type of stimulus the edge was (black edge or light edge), was crucial to the cells response (Hubel, 1995). Each cell preferred a certain orientation, and the more that edge or bar was rotated, the less the cell would fire, and eventually the cell would not respond at all (Figure 13).
Screen shot 2013-12-17 at 9.04.22 AM.png Figure 13
Purves, Dale. "Central Visual Pathways." Neuroscience. 3rd ed. Sunderland, MA: Sinauer, 2004. 259-80. Print.

Screen shot 2013-12-17 at 9.26.52 AM.pngFigure 14
Purves, Dale. "Central Visual Pathways." Neuroscience. 3rd ed. Sunderland, MA: Sinauer, 2004. 259-80. Print.

Figure 14 depicts on image of how the cortex is organized. The cells in the same vertical column all prefer the same orientation of a bar. Moving laterally, the new column contains cells that prefer a slightly different orientation (Kandel, 2000). The cat preferred one orientation, but still fired at similar orientations (Figure 13). This will be explained in the next section.

Simple Cells
Simple cells are located in V1 and combine input from LGN cells that have neighboring receptive fields. When these receptive fields are combined, the shape of the center receptive field is a bar or line (Figure 15). This shape is identical to the shape that was shone on the cat's retina. Figure 14 shows the LGN cells synapsing on 1 simple cell, which leads to that simple cell firing when those cells in the LGN fire. The cells in the LGN fire if that image is shown onto the retina (Hubel, 1995).

Slide19.jpgFigure 15

Figure 16

This video is of the experiment that Hubel and Wiesel conducted. The electrode is hooked up to a speaker, so the sound that is heard is the simple cell firing. Figure 15 is a drawing created by Hubel and Wiesel depicting what the receptive field of a simple cell looks like. The simple cell in the video has a receptive field identical to (a) in Figure 16. As the light bar changes orientation, there is a change in response of the simple cell. This is because there is little light on the on-center and more light on the off-surround which is not preferred. When diffuse light is shone, the simple cell does not respond at all. The center is an on-center, so it fires when light is shown in that region. When the pen in the video is placed in the location of the on-center portion of the simple cell, there is no response. When the light moves outside of the center receptive fields of the retina, it moves outside the center receptive field of the LGN which leads to lack of firing by the LGN. If the LGN cells do not fire then the simple cell will not become excited.

Complex Cells
Complex Cells are the other type of visual cortical neuron. This was actually the first type of cell that Hubel and Wiesel discovered. Complex cells are similar to simple cells because they prefer a certain orientation, but the major difference is that complex cells prefer motion within their receptive field. A stationary bar does not produce a strong response, while a moving bar through the receptive field yields the strongest response. Complex cells do not contain on or off responses (Hubel, 1995). The cell only needs a correctly oriented stimulus moving through the receptive field, and will fire at the same frequency anywhere in the receptive field (Figure 17).

Slide20.jpgFigure 17

Complex cells receive their input from multiple simple cells. An obvious question to ask is why a complex cell does not fire if a stationary stimulus is shone? Why is a single simple cell's response not enough to evoke a response in a complex cell? It is thought that the for the complex cell to depolarize enough to elicit a response, multiple simple cells must fire on it. Simple cells may be rapidly adapting, which can be seen in the video (Hubel, 1995). There is a point early on in the video where the light bar is remains on the screen while someone fills in the center receptive field. The cell does not continue to fire as frequently as when the stimulus first appeared. Every time that the visual stimuli moves in the receptive field, it covers the on-center portion of a new simple cell. This spatial summation leads to the complex cell being depolarized to the point of firing an action potential.
Screen shot 2013-12-17 at 11.09.08 AM.pngFigure 18
Kandel, Eric R. "Central Visual Pathways." Principles of Neuroscience. 4th ed. New York:McGraw-Hill, 2000. 529-38. Print.

In this video Hubel and Wiesel are investigating complex cells. This complex cell has little direction specificity. Some complex cells do not fire when the stimulus is moved in the opposite direction. Similar to the simple cell, the complex cell has a preferred orientation of a stimulus. As the orientation of the stimulus changes, the response of the complex cell changes accordingly, ultimately leading to no firing at all even when moved in the preferred direction (Hubel, 1995). Hubel and Wiesel drew a schematic to show how directional selectivity works (FIgure 19). If the stimulus moves in one direction, the simple cells (shown in white) are inhibited by the green cell. The green cell would be inhibiting the next cell in the sequence, hyperpolarizing it to the point where it would be able to reach threshold. If the stimulus were to move in the opposite direction, the cell that would be inhibited would already have fired.

78.jpgFigure 19

Cortex Layers
The cortex has six layers and the main input from the LGN terminates on cells in layer 4. SInce the LGN's layers are separated by eye, layer 4 of the cortex is also separated into columns that receive input from one eye. These are called ocular dominant columns (Figure 20). After layer 4, the information sent from the LGN begins to be combined with the other eye (Figure 20). So where are these simple and complex cells located? Hirsh et. al. found that layer 4 had a large density of simple cells, while layers 6,5, and 2+3 had a large density of complex cells (Figure 21, 1998). Some of the simple cells project back down to 5 and 6, most likely to the complex cells, which then project back up to more superficial layers of the cortex (Hirsch, 1998 Hubel, 1962). Kelly and Hessen found that stellate cells were most likely simple cells while pyramidal cells were complex cells (Figure 22, 1974). The findings of this study were similar to what Hirsh et. al. found. The stellate cells are located in layer IV and the pyramidal cells are located in layer VI, V, as well as II and III (1998).
Screen shot 2013-12-18 at 9.34.09 AM.pngFigure 20
Purves, Dale. "Central Visual Pathways." Neuroscience. 3rd ed. Sunderland, MA: Sinauer, 2004. 259-80. Print.

Screen shot 2013-12-18 at 9.54.26 AM.pngFigure 21Screen shot 2013-12-18 at 10.24.42 AM.png Figure 22 Kandel, Eric R. "Central Visual Pathways." Principles of Neuroscience. 4th ed. New York:McGraw-Hill, 2000. 529-38. Print.

How does this contribute to vision?
A quote from Hubel
"Many people, including myself, still have trouble accepting the idea that the interior of a form… does not itself excite cells in our brain,… that our awareness of the interior as black or white.… depends only on cells' sensitivity to the borders. The intellectual argument is that perception of an evenly lit interior depends on the activation of cells having fields at the borders and on the absence of activation of cells whose fields are within the borders, since such activation would indicate that the interior is not evenly lit. So our perception of the interior as black, white, gray or green has nothing to do with cells whose fields are in the interior—hard as that may be to swallow.… What happens at the borders is the only information you need to know: the interior is boring (Kandel, 2000)."

Remember how diffuse light does not illicit a response in the cortex? Early experimenters thought that the interior was not boring, but in reality it is. Our eyes need edges and differences in light. There is complex processing that occurs in the 3 areas discussed that aid in forming the visual image that we perceive. Our eyes also need change. The visual image needs to change in order for us to see (Riggs, 1952). We think that our fovea is focused on an image but it constantly moving. Remember how the simple cell does not continue to fire if the stimulus is constant? If the eye does not move the visual image disappears entirely. The change our eyes need is a change in the locations of the edges.

Depolarize--positive change in cell voltage
Hyperpolarize--negative change in cell voltage
Ipsilateral--same side
Contralateral-opposite side
Hemiretina--one half of the retina
Acuity--being able to detect detail
Electrode--a device that allows one to connect a neuron to an electrical device
Diffuse--to spread over a wide area

Supplementary Reading
What the resting membrane potential is and how an action potential is propagated.

First half focuses on structures and functions of parts of the eye other than the retina.

How to find your blindspot.

Actions of the ocular muscles

Different types of eye movements that can be made.

True and False
1. Simple and complex cells both prefer a certain orientation
2. Ganglion cells do not have the same center-surround orientation as bipolar cells.
3. Rods are more densely concentrated in the fovea.
4. All complex cells display directional selectivity.
5. There are 6 layers in the LGN
6. Simple cells feed into complex cells
7. The optic nerve occurs before the optic tract.
1. Why does diffuse light not elicit a response in the retina, LGN or visual cortex?
2. Describe how a neighboring cells in the LGN contribute to the receptive field of a simple cell.
3. How does light hitting a photoreceptor lead to its hyperpolarization?
4. How do horizontal cells contribute to the firing of the bipolar cell?

1. T, 2. F, 3. F, 4. F, 5. T, 6. T, 7. T


1. Heeger, David. "Perception." Lecture. NYU, New York. New York University. Web. 12 Dec. 2013

2. Hirsch, Judith. et. al. "Ascending Projections of Simple and Complex Cells in Layer 6 of the Cat Striate Cortex." The Journal of Neuroscience 18.19 (1998): 8086- 094.Journal of Neuroscience. Journal of Neuroscience. Web. 8 Dec. 2013.

3. Hubel, David. "Eye, Brain, and Vision." Eye, Brain, and Vision. Harvard, n.d. Web. 10 Dec. 2013.

4. Hubel, David, and Torsten Wiesel. "Receptive fields, binocular interaction and functional architecture in the cat's visual cortex."
" Journal of Physiology 160 (1962): 106-54. National Center for Biotechnology Information. Web. 9 Dec. 2013.

5. Kandel, Eric R. "Central Visual Pathways." Principles of Neuroscience. 4th ed. New York:McGraw-Hill, 2000. 529-38. Print.

6. Kelly, J., and D. Van Essen. "Cell structure and function in the visual cortex of the cat." Journal of Physiology 238 (1974): 515- 47.National Center for Biotechnology Information. Web. 10 Dec. 2013.

8. Purves, Dale. "Central Visual Pathways." Neuroscience. 3rd ed. Sunderland, MA: Sinauer, 2004. 259-80. Print.

9. Riggs, L. A., and F. Ratliff. "The effects of counteracting the normal movements of the eye." Journal of the Optical Society of America 42 (1952): 872- 873.