Chapter 4: The Lateral Geniculate Nucleus (LGN) and Primary Visual Cortex (V1)

Introduction

In our quest to unravel the intricacies of visual perception, we have journeyed through the complex layers of the eye and ventured into the neural pathways that translate light into vision. In this chapter, we embark on a closer examination of a critical relay station in this intricate visual network: the Lateral Geniculate Nucleus (LGN) before moving to a discussion of neural processing in the Primary Visual Cortex (V1 or striate cortex).

The Lateral Geniculate Nucleus (LGN)

Before delving into the specifics of retinal connections, let's reintroduce the LGN, a key component of the thalamus. Nestled near the center of the brain, the thalamus acts as a gateway for sensory information, funneling it to various cortical regions for further processing. When it comes to vision, the LGN plays a pivotal role in this relay process (Sherman & Guillery, 2002).

Retinal Connectivity to LGN: Nasal and Temporal Portions

To comprehend the intricate connections between the retina and LGN, it's crucial to divide the retina into nasal and temporal regions:

• The nasal portion of the retina is situated closer to the nose and projects information

to the contralateral hemisphere. In simpler terms, the nasal portion of the right eye sends signals to the left hemisphere, while the nasal portion of the left eye connects to the right hemisphere (Guido, 2008).

• In contrast, the temporal portion of the retina is closer to the temples or sides of the head and follows an ipsilateral pathway. This means that the temporal portion of the right eye projects information to the right hemisphere, and the temporal portion of the left eye sends signals to the left hemisphere (Guido, 2008).

This unique connectivity pattern explains why information shown in the left-visual field will be transmitted to the right hemisphere and information from the right visual field will end up in the left hemisphere (see Figure 4.1).

Figure 4.1

Connection of eyes to the brain. Axons from the nasal half of each retina cross at the optic chiasm and project to the contralateral hemisphere, whereas axons from the temporal half remain on the ipsilateral side. This organization ensures that visual information from the left visual field is processed in the right cerebral hemisphere and information from the right visual field is processed in the left hemisphere.

"Neural pathway diagram" by Mads00 is licensed under CC BY-SA 4.0

LGN Layers and Pathway

With the retinal connection in mind, let's explore the LGN's internal structure and how it processes visual information. The LGN consists of six distinct layers (see Figure 4.2) that are connected to the retina in a specific format (Usrey & Alitto, 2015):

1. Layers 1, 4, and 6 predominantly receive input from the nasal portion of the retina (of the contralateral eye).
2. Layers 2, 3, and 5 primarily receive input from the temporal portion of the retina (of the ipsilateral eye).

This division ensures a systematic processing of visual information, with each LGN layer specialized in receiving data from specific regions of the retina.

Receptive Fields in LGN

Just as we've encountered in the retina, the cells within the LGN have receptive fields— specific areas in visual space where light stimulation elicits a response in a given neuron. Within the LGN, these receptive fields follow a center-surround arrangement, similar to those observed with ganglion cells (Hubel & Wiesel, 1961).

Magnocellular and Parvocellular Layers

As we delve deeper into the LGN's layers, we encounter two distinctive cell types: the magnocellular and parvocellular layers (Livingstone & Hubel, 1988). These cells contribute to diverse aspects of visual processing:

​Magnocellular Layers (Layers 1 and 2): These layers receive input from the M ganglion cells, which are larger and sensitive to motion. The magnocellular layers are primarily involved in processing motion-related information and the location of objects in space.

Figure 4.2

Six layers of the LGN. The LGN consists of six distinct layers that receive separate inputs from the retina. Layers 1 and 2 form the magnocellular (M) pathway, which receives input from magnocellular retinal ganglion cells and is specialized for processing motion and coarse visual features. Layers 3 through 6 form the parvocellular (P) pathway, which receives input from parvocellular retinal ganglion cells and is specialized for processing color and fine spatial detail. The segregation of these pathways preserves different types of visual information as they are relayed from the retina to the primary visual cortex.

"The Lateral Geniculate Nucleus" by Ophthalmology is in the Public Domain, CC0

Parvocellular Layers (Layers 3, 4, 5, and 6): These layers receive input from the P ganglion cells, which are smaller and play a more significant role in color perception and perception of fine details.

The Primary Visual Cortex (V1)

The information processed within the LGN serves as the foundation for higher-level visual processing in the primary visual cortex, known as V1. Located in the occipital lobe at the back of the brain, V1 is where the initial shaping of our visual perception takes place. Here, the segregated magnocellular and parvocellular inputs continue to influence how we perceive the visual world.

To comprehend the processing within V1, we must first consider the nature of the receptive fields for cells in this part of the brain. These are distinct from the receptive fields found in ganglion cells or LGN cells, as V1 cells exhibit a specific response pattern. Instead of a center-surround arrangement, V1 cells respond most effectively to edges, particularly lines with specific orientations (Hubel & Wiesel, 1959). In essence, V1 cells are finely tuned to detect edges in the visual scene. Whereas retinal ganglion cells and LGN neurons have circular center-surround receptive fields that respond to spots of light, V1 neurons combine inputs from many LGN neurons to create elongated receptive fields that are selective for the orientation of edges. So, rather than responding to uniform areas of light or darkness, V1 neurons are specialized for detecting edges because edges define the boundaries of objects and provide the building blocks from which more complex visual features can be recognized in higher visual areas.

Hubel and Wiesel described three classic types of cortical cells in V1 (Hubel & Wiesel, 1962; 1965).

1. Simple Cortical Cells: These cells respond to edges with specific orientations. They are highly selective and will fire in response to an edge at a particular location and orientation. The orientation preferences can vary across different simple cortical cells, covering all possible angles. Although they may respond to a moving edge passing through that location, they are primarily selective for orientation and position rather than motion itself.
2. Complex Cortical Cells: Unlike simple cortical cells, complex cortical cells are less sensitive to the exact location of an edge within their receptive field. They respond to edges with a specific orientation and respond most strongly when the edge moves in a particular direction.
3. End-Stop Cells: These cells are specialized in responding to lines of a specific length, and the length of the line matters in their activation. Additionally, some end-stop cells are sensitive to corners or angles in the visual input.

Neurons with similar orientation preferences are grouped together into structures known as orientation columns (Blasdel & Salama, 1986). As an electrode moves across the surface of V1, the preferred orientation of neighboring neurons changes gradually, allowing all edge orientations to be represented across the cortex. V1 also contains ocular dominance columns, in which neighboring neurons receive stronger input from either the left or right eye. Information from both eyes is first combined in V1, providing the basis for binocular vision and depth perception.

Another important feature of V1 is that it is organized as a retinotopic map (Engel, 1997). This means that neighboring neurons in V1 receive input from neighboring locations on the retina. This orderly mapping is preserved throughout much of the early visual system, beginning in the retina and continuing through the LGN to V1. As a result, the spatial layout of the visual world is preserved in the cortex: stimuli that fall next to each other on the retina activate neighboring groups of neurons in V1. This organization allows the brain to keep track of where objects are located in the visual field.

Although V1 preserves the layout of the retina, it does not devote an equal amount of cortex to every part of the visual field. Instead, a disproportionately large number of V1 neurons are dedicated to processing information from the fovea, the small central region of the retina responsible for our sharpest vision. Even though the fovea occupies only a tiny portion of the retina, it contains the highest density of photoreceptors and provides the fine detail needed for tasks such as reading and recognizing faces. As a result, a much larger area of V1 is devoted to processing information from the fovea than from the peripheral retina. This unequal representation is known as cortical magnification (Horton & Hoyt, 1991).

Cortical magnification is not unique to V1. It is a general principle of sensory processing in which the brain allocates more neural tissue to processing information from regions that are especially important for perception. For example, even within the visual system, a disproportionately large proportion of neurons in the LGN are devoted to processing information from the fovea than from the peripheral retina. Later in this book, we will see the same principle in the somatosensory system, where the hands and lips occupy a much larger area of the somatosensory cortex than would be expected based on their physical size. In both vision and touch, the amount of cortex devoted to a body region reflects its importance for perception rather than its actual size.

Interestingly, V1 plays a pivotal role in our conscious experience of vision. Damage to V1 can result in a person reporting blindness, even though other visual pathways may remain intact and allow the person to perceive aspects of the visual world, like movement. Most of the visual information from the eyes is routed to the LGN and subsequently to V1. However, some information goes to the superior colliculus located at the top of the brainstem. From there, an alternate pathway ascends through other thalamic nuclei to extrastriate cortical areas, particularly regions involved in motion processing within the dorsal visual stream. This means that an individual with V1 damage may still exhibit "blindsight," wherein they report blindness but retain the ability to perceive and respond to motion. For example, a person with blindsight may insist that they cannot see an object moving across a screen, yet when asked to guess its direction of movement, they perform well above chance. This demonstrates that some visual processing can occur without conscious visual awareness (Weiskrantz, 1986; Cowey, 2010).

This overview of visual processing in V1 provides a foundation for understanding the intricate mechanisms that underlie our perception of the visual world. The distinctions among simple, complex, and end-stop cells, as well as the retinotopic map, cortical magnification highlight the complexity and sophistication of the human visual system, and the concept of blindsight highlights the importance of early visual brain areas in conscious perception.

2DG and column-like organization

The 2-deoxyglucose (2-DG) technique is a method used to map functional activity in the brain, including the primary visual cortex (V1). It works by taking advantage of the fact that active neurons require more energy than inactive neurons. During an experiment, an animal is injected with a radioactive form of glucose called 2-deoxyglucose (2-DG). Because neurons use glucose as their primary energy source, the most active neurons take up both normal glucose and the radioactive 2-DG (Sokoloff et al., 1977). Because neurons cannot metabolize 2-DG normally, it accumulates in active cells, making it possible to identify which neurons were most active during the experiment.

After the experimental condition is complete (for example, while the animal views visual stimuli such as lines with different orientations), the animal is humanely euthanized, and the brain is removed for analysis. Thin slices are prepared for autoradiography, a technique in which the radioactive 2-DG exposes photographic film. Areas that accumulated more 2-DG appear darker on the film, allowing researchers to identify which regions of the brain were most active during the task.

When Hubel and Wiesel and later researchers applied the 2-DG technique to animals viewing visual stimuli, they found that activity in V1 was not distributed evenly across the cortex. Instead, it appeared in organized stripes or bands. Some of these cortical columns responded more strongly to input from one eye than the other, forming ocular dominance columns, while others responded preferentially to lines of a particular orientation, forming orientation columns. Interspersed among the orientation columns are regions known as blobs, which contain neurons that are particularly involved in processing color information (we will return to this in the chapter on color vision; Livingstone & Hubel, 1984). These findings provided strong evidence that the primary visual cortex is organized into functional cortical columns, with neighboring groups of neurons specialized for processing different aspects of visual information.

Neural Fatigue

If you stare at the large red rectangle on the left side of Figure 4.3 for about 1 minute and then shift your gaze to the small red dot on the right side of  Figure 4.3 you will experience an illusion (Blakemore & Campbell, 1969). Although the lines on the right are actually equally spaced, the bars at the top will appear more closely spaced, while the bars at the bottom will appear farther apart.

This illusion occurs because different groups of neurons are tuned to different spatial frequencies (the spacing between lines). While you are looking at the adapting stimulus on the left, neurons whose receptive fields correspond to the upper region of the image respond most strongly to the low spatial frequency pattern (widely spaced lines). At the same time, neurons whose receptive fields correspond to the lower region respond most strongly to the high spatial frequency pattern (closely spaced lines). This is shown in Figure 4.4.

Figure 4.3

Viewers fixate on the adaptation stimulus (left), which contains regions of low and high spatial frequency, for approximately one minute before shifting gaze to the red fixation point in the test stimulus (right). Although the horizontal bars in the test stimulus are physically equally spaced, the upper bars appear more closely spaced and the lower bars appear farther apart because adaptation temporarily reduces the responsiveness of neurons tuned to the spatial frequencies.

"Spatial frequency adaptation." by Kahan, T.A. is licensed under CC BY-NC-SA 4.0 

Figure 4.4

Continuous viewing of the adapting stimulus temporarily reduces the responsiveness of neurons tuned to the viewed spatial frequency (red downward arrows). Because these adapted neurons respond less strongly than usual, neighboring neurons tuned to slightly different spatial frequencies contribute relatively more to perception. Following adaptation to low spatial frequencies (top), the test pattern appears to have a higher spatial frequency, whereas adaptation to high spatial frequencies (bottom) causes the same pattern to appear to have a lower spatial frequency.

"Spatial frequency neural adaptation." by Kahan, T.A. is licensed under CC BY-NC-SA 4.0 

Because these neurons have been responding continuously for about a minute, they become adapted (sometimes referred to as neural fatigue). Adapted neurons still respond to visual input, but their response is temporarily weaker than normal. In Figure 4.4, this reduced response is represented by the red arrows.

When you then shift your gaze to the test pattern on the right side of Figure 4.3, the bars do not appear to have the same spatial frequency. Under normal circumstances, neurons tuned to medium spatial frequencies would produce the strongest response. However, the neurons tuned to low spatial frequencies in the upper region are still adapted. Although they continue to respond, they now respond less strongly than before. As a result, neurons tuned to slightly higher spatial frequencies contribute relatively more to perception, causing the top bars to appear more closely spaced than they really are. In the lower region, the opposite occurs. The neurons that are tuned to high spatial frequencies have become adapted and therefore respond less strongly than usual. This allows neurons tuned to slightly lower spatial frequencies to have a greater influence on perception, making the bottom bars appear more widely spaced than they actually are.

This illusion demonstrates that perception depends on the relative pattern of activity across a population of neurons rather than the activity of a single neuron. By temporarily reducing the responsiveness of the neurons that were most active during adaptation, the brain interprets the same physical stimulus differently, producing the illusion.

Conclusions

The lateral geniculate nucleus and primary visual cortex represent the first major stages of cortical visual processing. Although the retina begins the process of analyzing visual information, the LGN organizes and relays this information to V1 while maintaining separate pathways for different types of visual signals. Within V1, neurons become selective for features such as edge orientation, line length, and binocular input, transforming simple retinal signals into representations of the contours and structure of objects. The organization of V1 into orientation columns, ocular dominance columns, retinotopic maps, and regions devoted disproportionately to the fovea illustrates that the visual cortex is highly structured, with neighboring groups of neurons specialized for processing different aspects of the visual scene. Studies of visual adaptation and blindsight further demonstrate that our perception depends not only on the activity of individual neurons but also on coordinated patterns of activity across populations of neurons and multiple visual pathways. Together, the LGN and V1 provide the foundation upon which all higher levels of visual perception are built. In later chapters, we will examine how information leaving V1 is processed by extrastriate visual areas that analyze increasingly complex properties of objects (e.g., motion and color).

References

Blakemore, C., & Campbell, F. W. (1969). On the existence of neurons in the human visual system selectively sensitive to the orientation and size of retinal images. The Journal of Physiology, 203(1), 237–260. https://doi.org/10.1113/jphysiol.1969.sp008862

Blasdel, G. G., & Salama, G. (1986). Voltage-sensitive dyes reveal a modular organization in monkey striate cortex. Nature, 321, 579–585. https://doi.org/10.1038/321579a0

Cowey, A. (2010). The blindsight saga. Experimental Brain Research, 200(1), 3–24. https://doi.org/10.1007/s00221-009-1914-2

Engel, S. (1997). Retinotopic organization in human visual cortex and the spatial precision of functional MRI. Cerebral Cortex, 7(2), 181–192. https://doi.org/10.1093/cercor/7.2.181

Guido, W. (2008). Refinement of the retinogeniculate pathway. The Journal of Physiology, 586(18), 4357–4362. https://doi.org/10.1113/jphysiol.2008.157115

Horton, J. C., & Hoyt, W. F. (1991). The representation of the visual field in human striate cortex. Archives of Ophthalmology, 109(6), 816–824. https://doi.org/10.1001/archopht.1991.01080060080030

Hubel, D. H., & Wiesel, T. N. (1961). Integrative action in the cat's lateral geniculate body. The Journal of Physiology, 155(2), 385–398. https://doi.org/10.1113/jphysiol.1961.sp006635

Hubel, D. H., & Wiesel, T. N. (1959). Receptive fields of single neurones in the cat's striate cortex. The Journal of Physiology, 148(3), 574–591. https://doi.org/10.1113/jphysiol.1959.sp006308

Hubel, D. H., & Wiesel, T. N. (1962). Receptive fields, binocular interaction and functional architecture in the cat's visual cortex. The Journal of Physiology, 160(1), 106–154. https://doi.org/10.1113/jphysiol.1962.sp006837

Hubel, D. H., & Wiesel, T. N. (1965). Receptive fields and functional architecture in two nonstriate visual areas (18 and 19) of the cat. Journal of Neurophysiology, 28(2), 229–289. https://doi.org/10.1152/jn.1965.28.2.229

Livingstone, M. S., & Hubel, D. H. (1984). Anatomy and physiology of a color system in the primate visual cortex. Journal of Neuroscience, 4(1), 309–356. https://doi.org/10.1523/JNEUROSCI.04-01-00309.1984

Livingstone, M. S., & Hubel, D. H. (1988). Segregation of form, color, movement, and depth: Anatomy, physiology, and perception. Science, 240(4853), 740–749. https://doi.org/10.1126/science.3283936

Sherman, S. M., & Guillery, R. W. (2002). The role of the thalamus in the flow of information to the cortex. Philosophical Transactions of the Royal Society B: Biological Sciences, 357(1428), 1695–1708. https://doi.org/10.1098/rstb.2002.1161

Sokoloff, L., Reivich, M., Kennedy, C., Des Rosiers, M. H., Patlak, C. S., Pettigrew, K. D., Sakurada, O., & Shinohara, M. (1977). The [14C]deoxyglucose method for the measurement of local cerebral glucose utilization. Journal of Neurochemistry, 28(5), 897–916. https://doi.org/10.1111/j.1471-4159.1977.tb10649.x

Tootell, R. B. H., Hamilton, S. L., Switkes, E., & Switkes, J. (1988). Functional anatomy of macaque striate cortex. V. Spatial frequency. Journal of Neuroscience, 8(5), 1610–1624. https://doi.org/10.1523/JNEUROSCI.08-05-01610.1988

Usrey, W. M., & Alitto, H. J. (2015). Visual functions of the thalamus. Annual Review of Vision Science, 1, 351–371. https://doi.org/10.1146/annurev-vision-082114-035920

Weiskrantz, L. (1986). Blindsight: A Case Study and Implications. Oxford University Press.