Reading 4: The cerebral cortex

Question 1

Cerebral cortex in small animals vs large animals/humans

Answer 1

• The cerebral cortex is a 2-4 mm sheet of tissue approximately 120 square cm in surface area. It is almost all you see when you look at an intact human brain.
• In small mammals, like rats and mice, the cortical sheet is smooth; however, in larger mammals, like dogs, cats, humans and whales, the cortex is wrinkled up.

Question 2

Why does our cortex have folds

Answer 2

• This folding may have evolved to enable a large cortical surface area to fit inside the skull.
• The pattern of folds in the human cerebral cortex shows quite a bit of variability from one individual to another; nevertheless, it is possible to identify anatomical landmarks that divide the human cortex into distinct regions.

Question 3

The central sulcus and lateral sulcus

Answer 3

• The lateral sulcus is a large deep groove that separates the temporal lobe from the rest of the brain.
• The temporal lobe is like an arm that extends rostrally. You can grab it and pull it back or break it off.
• The location and course of the central sulcus is somewhat variable from one brain to another, and it can be harder to find. It extends from the dorsal-medial apex of the cortex to the lateral sulcus.
• The frontal lobe corresponds to the region of cortex dorsal to the lateral sulcus and rostral to the central sulcus.
• The parietal lobe is caudal to the
  central sulcus and dorsal to the temporal lobe.
• The boundary of the occipital lobe on the lateral surface of the brain runs from a small notch at the top of the brain to a small notch at the bottom of the brain. As we’ll see, the dorsal notch is the beginning of a large sulcus on the medial surface of the cortex.

Question 4

Gyri and Sulci of the Frontal Lobe

Answer 4

• rostral to the central sulcus, the precentral sulcus runs parallel to the central sulcus and extends from the top of the brain to the lateral sulcus.
• The cortex between these two sulci forms the precentral gyrus = primary motor cortex (main output from the cortex for the control of voluntary movement).
• the superior frontal sulcus and the inferior frontal sulcus: these two sulci divide the rostral regions of the frontal lobes into the superior frontal gyrus, the middle frontal gyrus, and the inferior frontal gyrus.
• The rostral regions of the frontal lobes are involved in complex aspect of movement control and in executive functions like decision-making.
• In the left hemisphere, part of inferior frontal gyrus just anterior to the precentral sulcus forms Broca’s area, which is involved in the control of speech.

Question 5

Identify the Sulci and Gyri of the Frontal lobe

Answer 5

Question 6

Identify the Sulci and Gyri of the Temporal lobe

Answer 6

• As the cortex grows over the course of embryonic development it takes on a C shape. The temporal lobe is the bottom of the C. It sticks out like an arm, separate from the rest of the brain.
• The temporal lobe is divided by three sulci that run parallel to the lateral sulcus: the superior temporal sulcus, the middle temporal sulcus, and the inferior temporal sulcus. The inferior temporal sulcus is on the ventralsurface of the cortex and is not visible in Fig. 4.
• The three sulci divide the temporal lobe into the superior temporal gyrus, the middle temporal gyrus and the inferior temporal gyrus.
• A caudal region of the superior temporal sulcus forms Wernicke’s area, which is involved in language comprehension.

Question 7

What happens if you grab a hold of the temporal lobe and pull it back.

Answer 7

You will discover more cortex hiding behind it. This region, called the insula, is especially involved in sensations from the body and their relation to emotional states.
* The feeling of a full stomach, physical pain and emotional distress are all associated with activation of the insula.

Question 8

Sulci and Gyri of the Parietal Lobe

Answer 8

• In the parietal lobe, the postcentral sulcus runs posterior and parallel to the central sulcus (Fig. 6).
• The gyrus between these two sulci is called the postcentral gyrus. This is primary somatic sensory cortex.
• The other main sulcus that divided the parietal lobe is the intraparietal sulcus. It starts at the postcentral sulcus and runs caudally toward the midline. The intraparietal sulcus divides the parietal lobe posterior to the postcentral sulcus into two regions. The region dorsal to the intraparietal sulcus is called the superior parietal gyrus. The region ventral to the intraparietal sulcus is subdivided into two regions: the supramarginal gyrus and the angular gyrus.

Question 9

Occipital lobe

Answer 9

The boundary, on the lateral cortical surface, between the parietal,
temporal and occipital lobes is a bit vague. It’s roughly as shown in
Fig. 7

Question 10

The Medial Surface of the Cortex

Answer 10

• The two cerebral hemispheres are separated by a deep groove running down the midline, called the longitudinal fissure. A large part of the cortical surface is within this groove and is only visible in a midsagittal section that cuts between the two hemispheres (Fig. 8).
• The midsagittal cut transects the corpus collosum, the enormous fiber tract that connects the two cerebral hemispheres. Following the corpus collosum over much of its anterior and dorsal expanse is the cingulate sulcus. The cingulate gyrus lies between the cingulate sulcus and the corpus collosum.
• Like the insula, the cingulate gyrus is limbic cortex. The term “limbic” is vague and anachronistic, but still widely used. It generally refers to brain regions involved in connecting mental states, feelings, emotions, physiological responses and behavior.
• The insula and cingulate cortex are often simultaneously active when physiological responses, emotions
  and feelings are involved, for example when we feel pain.
• The parietooccipital sulcus separates the parietal and occipital lobes on the medial surface of the cortex. The calcarine sulcus is an important anatomical landmark since it runs through the center of primary visual cortex.

Question 11

The ventral surface of the cortex

Answer 11

• The** inferior temporal sulcus**, which was not visible in the lateral view of the temporal lobe is apparent running along the ventral surface of the temporal lobe (Fig. 9).
• The collateral sulcus and rhinal sulcus run more medially. These three sulci form the boundaries of the fusiform gyrus, which is important in visual processing.
• Medial to the rhinal sulcus is the parahippocampal gyrus.
• The hippocampus is buried within this region. At the anterior end of the rhinal sulcus and just medial to it is a distinctive triangular shaped bump called the uncus. Within the uncus is the amygdala, a cluster of nuclei, about the size of a pea, involved in emotions, especially fear, and their relation to learning and memory.

Question 12

What is phrenology and its contribution to to modern neuroscience?

Answer 12

Gall and his followers claimed that a broad range of human personality traits, such as benevolence, conscientiousness, and truthfulness, could be assessed in individuals by examining the patterns of bumps on their skulls, which reflected the bumps and contours of their underlying brains.

Despite its obvious (to us) absurdity, phrenology made an
important contribution to modern neuroscience: the idea that the cerebral cortex is organized into anatomically distinct regions, each specialized for carrying out a specific function.

Paul Broca and Carl Wenicke showed that patients with damage to localized areas of the left hemisphere showed deficits in specific aspects of speaking and understanding languaage.

Question 13

Organization of our cortex

Answer 13

• However, the cortex is not organized to meet our aesthetic or functional tastes. Its organization is a product of evolution by natural selection. So, instead of simply accepting that cortical organization “makes sense”, we might ask urselves what selection pressures would lead to the cortical organization we see.
• ie: The initial processing of visual information is carried out by primary visual cortex (V1), which is located on the medial surface of the occipital lobe, in the region surrounding the calcarine sulcus (Area 17). After this initial processing stage, the information is passed on to the region surrounding V1, referred to as V2 (area 18), for more complex processing. V1 neurons are mainly connected to other V1 neurons, whereas V2 neurons are mainly connected to other V2 neurons. This makes sense, because V1 and V2 are carrying out different specialized functions; however, it does not necessarily follow that V1 neurons and V2 neurons must be clustered together in two separate regions. The neurons could in principle be mixed together while maintaining separate connections (Fig. 5). Granted, the example on the right in Fig. 5 looks messy, but remember the brain doesn’t have to meet our aesthetic standards. There must be a more essential reason why different cortical functions are localized to different anatomical regions.

Question 14

Consider the cost of building + maintaning cerebral cortex: energy

Answer 14

• In the world in which our brains evolved, energy (i.e., food) was a scarce and highly valuable resource. Compared to the body’s other organs, the brain needs a disproportionately large amount of energy.
• It has, nevertheless, evolved to be highly energy efficient. Brains evolved in a world in which energy was a scarce and highly valuable resource, so their evolution has found ways to minimize energy consumption. We will focus on one:
  axon length. Axons are energetically expensive to build and maintain, so a good way to conserve energy is to make axons as short and as thin as possible.

Question 15

Consider the cost of building + maintaning cerebral cortex: space

Answer 15

• A second important cost is space. Your brain has around 100 billion neurons. They all take up space and they all must fit inside your head along with all the axons that connect them together.
• A good way to minimize the amount of space all these neurons and axons take up is to keep the axons as short and as thin as possible. Thus, constraints placed by both energy and space will tend to select for the shortest, thinnest possible axons

Question 16

Wiring costs

Answer 16

• There is substantial evidence that the cerebral cortex and neural circuits in general minimize these “wiring costs”. But this can’t be the entire explanation for the organization of the cortex.
• After all, if we were to take wiring minimization to its logical endpoint, we would conclude that each neuron in the cerebral cortex should be connected only to the neurons right next to it, since that would result in the shortest possible axons.
• A typical cortical neuron is in fact likely to be most strongly connected with nearby neurons; nevertheless, nothing but short-range connections would greatly limit the cortex’s computational power.
• Imagine, for example, that the front of the cortex needs to communicate with the back of the cortex. If all axons were short, messages sent from front to back would have to go through thousands of synapses. This would be slow and inefficient. An organism with a brain that was slow and inefficient would not be selected for and would not survive over an organism with a more computationally efficient brain.
• So, it seems that there ought to be at least some long-range
  connections. The brain appears to have evolved toward a balance between short range and long range connections and thus a balance between cost and computational efficiency.

Question 17

Graph Theory

Answer 17

• Graph theory is a mathematical field that examines the connectivity patterns of networks.
• In graph theory, the individual elements are called nodes and the connections between nodes are called edges.
• The path length is the number of edges required to get from one node to another.
• The degree of a node is the number of edges it has. So, for example, the path length from node A to node B is 1, whereas the path length from A to C is 2. Node A has a degree of 2, whereas node B has a degree of 3.

Question 18

Graph theory to examine the brain

Answer 18

• A node could correspond to an individual neuron, a column, or any other distinct cortical element, but usually when researchers refer to cortical nodes, they are referring to different regions of the cortex.
• Edges in this case correspond to the axons that connect these regions together.

Question 19

Lattice, complex and random topology

Answer 19

On the far left, each node is connected only to its nearest neighbor. This is equivalent to extreme wiring minimization. The cost of this lattice topology will be low because the axons will be short, but the computational efficiency, as measured by the long path lengths to get from one node to another, will also be low.

On the far right is a case in which each node has a random chance of being connected to any two other nodes in the network. This random topology will give the highest computational efficiency (that is, the lowest overall path length) but also the highest wiring costs, since there are a lot of long-range connections.

The cerebral cortex seems to have evolved to find a middle ground that strikes a balance between wiring costs and computational efficiency.

Question 20

Hypothethical model to describe how the cortex could be efficiently wired and computationally powerful

Answer 20

small world network
* Small world networks show a high degree of clustering and short path lengths.
* Social networks, for example, tend to be small world networks. My network of friends is clustered because people who are friends with me are also likely to be friends with each other. Social networks also typically have short path lengths. For example, you know me, I know Brenda Milner and Brenda Milner knew Wilder Penfield, so you are connected to Wilder Penfield by only three edges.
* Much functional and anatomical data are consistent with the hypothesis that the cortex is organized as a small world network.

Question 21

Small world network

Answer 21

• According to this model, groups of cortical neurons form nodes (e.g., primary visual cortex), nodes cluster together to form modules that carry out specific higher order functions (e.g., the network of visual regions that cause visual perception), while a smaller number of long-range connections enable anatomically separate modules to communicate directly with each other, integrating the computations in the individual modules into larger networks (Fig. 8).
• The key nodes that interconnect modules are called hubs. Hubs occupy central locations within the larger network and make many long-range connections between modules. From a computational point of view, the clustering of nodes into modules enables specialized processes to be carried out within localized cortical regions, whereas the hubs mediate the long-range connections that enable the large scale integration of cortical modules necessary for higher order cognitive processes.
• Functional anatomical studies suggest that regions in the frontal lobes and medial regions of the parietal lobes may be especially important hubs for integration of communication within the entire cortical network.

Question 21

How does this explain the organization of the cortex?

Answer 21

• A cortical column is a highly interconnected group of neurons carrying out a specific function. Because most neurons in a column are connected, wiring will be minimized if they are close together.
• Modeling studies indicate that wiring minimization leads naturally to the emergence of columns. The same logic applies to topographic maps.
• Take, for example, the somatotopic map of the body in primary somatic sensory cortex. The cortical neurons that represent the skin of your fingertips are likely to be co-activated when you feel an object with your hand, and they need to communicate with each other for you to perceive entire objects through touch. So, it makes sense, in terms of wiring minimization, that they are close together in primary somatic sensory cortex.
• In contrast, somatic sensations from your fingers and the back of your legs have less to do with each other and thus can be farther apart.
• The idea is that the competing selection pressures of wiring minimization and computational efficiency have led a cerebral cortex organized into highly interconnected and anatomically localized modules interconnected by sparser long-range connections.