Lecture notes with Cognitive Neuroscience at Leiden University - 2015/2016

Lecture 1: History and methods

While the historical background is mentioned in the literature, it will not be the focus of the exam. Rather, the methods of cognitive neuroscience are more important.

Cognitive neuroscience is a broad field that targets explaining behavior in neural terms.

The study of the relationship between the brain and the mind, which is at the base of cognitive neuroscience, dates back to ancient Greece. According to Descartes (17^th century), the mind and the body are separate entities, which interact in a specific structure in the brain (which is not important for this course). This interaction is problematic, and this was addressed by Damasio, who studied traumatic injuries and claimed that emotion arises from bodily states.

In cognitive neuroscience, there is a difference between mental representation (simulation of the outside world by cognition) and neural representation (a physical phenomenon). For this field, it is important to record the human brain in action, to underlie the relationship between structures of the brain to their function.

Reductionism is the over-simplification of phenomena, studying them physically (e.g. what structures are activated during emotional states?) to ignore the bigger question (what is the relationship between body and mind?). We tend to reduce our questions from high mental states to physical terms.

The building blocks of cognition (see slide 8) describe the process that occurs between perceiving a stimulus to producing a response.

Localization in the Brain

There were many attempts to localize functions in the brain, and the idea of localization was adopted and abandoned several times throughout history. It began with Gall's phernology, which assumed strict localization of personality, which was then disputed by Flourens who claimed that the brain cannot function as separate units. Later, Broca and Wernicke showed that specific lesions lead to lingual deficits, and finally, Brodmann used cytoarchitecture to map different regions in the brain according to certain cell structure and co-activation.

Pioneers of Neuropsychology

From the incorrect idea of Golgi that all neurons are connected to each other, Ramon y Cajal developed evidence that neurons are the base of the brain, and that they communicate by electrical signals. Sherrington updated this idea when he showed that synapses communicate chemically.

Until the 19^th century, the brain was a complete mystery. At the late 50s several psychologists gave rise to cognitive psychology. Fr example, Chomsky showed that to pick up language, a person needs a mental representation of the words in mind. Miller, in addition, addressed the importance of short-term memory. In the 60s, electrophysiology changed things, and in the 70s it was possible to relate brain activity to stimuli. In the 80s, brain imaging developed, and improved in the 90s. Nowadays, they are being improved all the time.

All the different methods used in cognitive neuroscience can be seen in three dimensions. Temporal resolution is the integrity the method has to the time of the activity. Spatial resolution is the precision in which the method identifies the structures in the brain. Finally, a method is assessed by how invasive it is. Look at slide 15 for description on these dimensions for each method.

Electrophysiology

• works by changing resting potentials (doesn't mean there is activity. Could be synchrony of graded potentials, for example) or single cell recordings of action potential.

• Single cell recording: almost exclusively done in animals. during this procedure, an electrode is implanted near or in an axon, recording the number of action potentials a neuron produces in a second. It is very invasive. This helped localizing some functions in the brain. Problematic for psychology.

• Electro-encephalography (EEG): placed in predetermined organization, included in an international standard. The pyramidal cells in the cortex are arranged in a line, and their synchrony in gradual increase/decrease of activity is recorded. Event-related potentials are based on EEG recordings, correlating an event to an EEG signal, averaged over repetitions. As a result, several peaks are easily interpreted. However, localization is difficult. Examples for such ERPs can be found on slide 22. ERP has good temporal resolution, but poor spatial resolution. As evidenced in research, certain ERP patterns can be related to cognitive functioning, but those are difficult to identify (as there are many electrical activities at a given moment).

Neuroimaging techniques

PET was the first imaging technique, and it was based on glucose use. fMRI is even more detailed, measuring hemoglobin that is not carrying oxygen in the brain (so oxygen was used at this area). Diffusion tension imaging is a method for 3D brain images. None of the imaging techniques is ever perfect.

• Magnetoencephalography (MEG) is using a magnetic field placed over the head to record electrical current. It mostly records a vertical signal, while in EEG most of the signal recorded is parallel to the skull. Therefore they are complementary to each other. The vertical signal detected results in missed activity in gyri. There is less use of EEG in research. While EEG shows event-related potentials, MEG is used to detect event-related fields.

• Structural MRI is used to create a 3D image of the brain. It has a very high spatial resolution. It can be used in different magnetic strength, with very strong signals resulting in unpleasant side-effects. When the magnetic field is activated, atoms able to react to it get in line (move at the same pase), and the MRI measures the time each atom takes to get back to normal, translating this time to the type of atom, thus recording the structures of brain by type of atoms.

• Diffusion tensor imaging (DTI) detects fatty tissues (myelin sheath) by measuring water molecules diffusion in them.

• Positron Emission Tomography (PET) marks glucose with radioactive substance. When this glucose is used in the brain, it releases two gamma photons that are then detected, and the cell between them is assumed to have used them. This has a serious limitation, since it takes about half a minute to record an activity.

• Functional MRI uses hemoglobin not carrying oxygen to infer brain activity (BOLD response). It happens in a certain pattern: initial dip of oxygen in the structure (which is the actual use of oxygen), followed by an overcompensation which is detected by the MRI machine and lasts for multiple seconds, and then another, deeper dip, which is still unclear for researchers. When studying brain functions, a contrast is made between two stimuli to tell the difference in brain structures involved.

Anatomical methods

MRI uses fat/water ratio in specific areas to spot myelinated axons. TMS identifies lesions in the brain.

• In reverse engineering, a researcher could remove a part of the brain and then infers the effect on bodily functions and behavior. Alternatively, those can be studied following a natural or elicited damage, or even by temporary damage induced in TMS, by a magnetic current. However, this has several limitations. It could be ethically or practically problematic. One lesion could affect people differently.

• In Transcranial magnetic stimulation (TMS), strong magnetic current is applied to a single location in the neocortex, inducing electrical currents. It is very safe, and has several applications in therapy.

• Transcranial direct current stimulation (tDCS) changed neuronal resting potentials, making them more or less likely to occur. It could have several clinical applications.

• Deep brain stimulations is used in several cases, including treating Parkinson's disease. It is also used in research, but only if the person has to go through brain surgery, because wires are implanted deep in the brain.

Biochemical methods

• Several methods use neurotransmitters, hormones, and other chemical agents to measure cognition. These can be seen in the saliva, blood, and genetics.

To conclude, several methods are used for comparing situations, since the brain is always active and cannot be at zero. Each method has its limitations, and interpreters should always be critical in their assessment.

Lecture 2: Structure and Function of the Nervous System

We are looking at the things that are especially important for human cognition. The other structures are less important for this course.

Brain Cells and Their Communication

For an Illustration of the Neural structure look at slide 3. The dendrites are the input sites, and there is usually one axon, which is the output site. The Endoplasmic reticulum and Golgi apparatus are there for transport purposes, and the mitochondria are for energy creation. The myelin sheaths are not a part of the cell, but are created by glia cells, which circle around the axon to provide isolation.

There are three types of glia cells:

1. Astrocytes connect the blood vessels with the brain, are capable of transporting what the brain needs (e.g. glucose and oxygen), and are an important part of the blood-brain-barrier (BBB). The BBB provides a protection of the brain from large molecules and harmful materials, since the brain is sensitive for toxic

2. Oligodendrocyte and Schwann cells are the ones providing myelin sheath, the first in the central nervous system and the Schwann cells in the central nervous system.

3. Microglia are used to clean and repair damage in the brain by removing toxic material (similarly to phagocytes in the immune system).

The membrane is not permeable for any material. Most materials enter or leave the cell through different proteins. Some proteins let certain ions transport through them (ports). The neuron usually has a resting potential of -70mV inside it, and when an input enters the cell there will be a fluctuation in it. If small, it would fade, but a large enough fluctuation will spread out and influence other parts of the membrane. When this fluctuation arrives the axon hillock, and the threshold is passed (-45mV), a port will open and Na+ ions will enter the cell, neutralizing the polarity difference. This will cause the potential in the axon to spike and become positive, and then decrease quickly to a state where no other action potential will occur. This is called the refractory period. The number of spikes of electrical current can indicate the activity of the cell.

Creating a current in the axon hillock artificially will result in a fading current along the axon, which does not happen naturally. Myelin sheaths will cause the action potential to 'hop' from one nod of Ranvier to the next, making a faster signal from an axon hillock to its terminal. This is needed in large axons (in large animals). this 'hopping' is called saltatory conduction.

Axons communicate with each other in a chemical manner, by using neurotransmitters. The presynaptic neuron releases those chemical messengers that are then recepted by the dendrites of the next axon (the postsynaptic dendrites). Those receptors than cause ports on the postsynaptic neuron to open and an excitatory or an inhibitory postsynaptic potential (EPSP/IPSP) is created, increasing or decreasing the chance of an action potential in the neuron.

Neurons communicate at synapses, using chemical messengers called neurotransmitters, released as a result of an influx of Ca²⁺. These neurotransmitters are later bound to the next neuron's receptors. There are multiple classifications of neurotransmitters:

1. Amino acids: Glutamate and GABA support more local connectivity between cells (glutamate is excitatory and GABA inhibitory), while ACh is spread throughout the body.

2. Biogenic amines: not primarily caused locally, but stem from certain nuclei in the brainstem, from where they project to the rest of the brain. One example of these is catecholamines, for example norepinephrine (NE; important for arousal and awareness), dopamine (DA; higher level cognitive processes), epinephrine (EP; important for the stress response). NE is created in the brainstem, but it can also be created in the adrenal cortex, where it contributes to the fight or flight response. DA and NE start in the superior brainstem (DA in the substiantia nigra and the ventral tegmental area and NE from the locus coeruleus), and from there they project to the entire brain. DA is mainly projected to the basal ganglia, important for motivation-driven behavior and reward sensation. NE is also spread through the brain, but more to the direction of the limbic system and less to the basal ganglia.

3. Neuropeptides: large chains of amino acids. Hormones are an example of these (oxytocin, endorphins).Oxytocin is important for bonding behaviors and endorphins are important for pain reduction, and is associated with addiction. Neuropeptides are more likely to work in distant locations.

Electrical Synapses

When two neurons are very close together (almost never happens in humans), an almost direct connection is formed between them, creating a faster connection between them. Those two neurons share the cytoplasm while using two different cell nuclei.

Large Scale Brain Anatomy

A diagram of relations of the main divisions of the nervous system (NS) can be found in slide 16.

The nervous system includes the central (CNS) and the peripheral NS (PNS). The autonomic NS, which is a part of the PNS includes the sympathetic nervous system (SNS; energized by NE), which prepares your body to perform an action, e.g. increasing blood supply and oxygen and decreasing digestion. The parasympathetic NS brings the body to a state of relaxation, performing the opposite of the SNS by using Acetylcholine (ACh).

The CNS includes the spinal cord and the brain. You could distinguish 5 areas in the spinal cord, named after the bones around them, but they are not important for psychology. In psychology the important things are the butterfly-shaped gray matter (neural nucleus) and white matter around it (myelinated axons). The brain is constructed from bottom up.

The brainstem is constructed from the medulla (including 12 cranial nerves that serve basic processes in the face and throat for example), which is vital for life, the pons, which connect the cerebral cortex directly to the cerebellum, and the midbrain, that emits DA, NE, and other neurotransmitters.

Above the brainstem are several subcortical structures. Diencephalon is the thalamus and hypothalamus. The hypothalamus is in charge of homeostasis in the boy, using the pituitary gland as well. The thalamus is more about regulating the brain, acting as a relay station between the sensors to the cerebral cortex, and back between the cerebral cortex to the muscles. Within the thalamus we can differentiate different several other parts, some of them are important (we will see in the next lectures).

The telencephalon also belongs to the cerebral cortex, including also the amygdala and the hippocampus (belonging to the limbic system, important for emotion and memory), and the basal ganglia, important for motor functions. The cerebral cortex can be subdivided to lobes, and gray versus white matter. The two hemispheres in the brain are connected by the corpus callosum. There is a fixed location for chambers containing the cerebral spinal fluid, and those are the ventricles.

We can make different distinctions about locations: layers (e.g., 6 layers in the neocortex), columns (there are functional units to be identified combining layers and columns creating a box), gray/white matter, and shapes (gyri and sulci).

Development of Cortex

The meaningful part for psychology is described in slides 19-20. In the embryonic stage of pregnancy (3-7m), cells are not yet differentiated (called precursors). Those can split up, proliferate, and can later be differentiated. Their location and function is determined by the moment in which they start migrating. They start in the base of the brain (ependymal layer), and develop to the superficial layers of the brain. It is hypothesized that this happens with the help of radial glia cells that guide the migrating cells to their destined locations. They also support the connection between one area in the cortex to another.

Brain Stem

An illustration of the brain stem is found on slide 21. The most important structures are the reticular formation, substantia nigra, superior colliculus and collective name: cranial nerves.

The thalamus is important for transfer of information between the brain and the rest of the body. The hypothalamus tells the pituitary glands to exert certain hormones, thus it regulates it. The cerebellum and the basal ganglia are important for motor control. Finally, the limbic system is primarily the hippocampus, the amygdala and the cingulate cortex, and it is important for creating, regulating and remembering emotions. The amygdala is very sensitive to level of arousal.

The cortex can be differentiated by different gyri and sulci, and those are named mostly after the lobe in which they are in, and by their relative location in the brain. The primary motor cortex is found anteriorly to the central sulcus, and it is specialized to certain parts of the body, creating a map of motor neurons in the brain. The same goes for the somatosensory cortex, which is right posteriorly to the primary motor cortex. The association cortexes integrate multiple senses from the different primary sensory cortexes.

Lecture 3: Sensation and Perception, Object Recognition, and Attention

Note: the lecturer skipped many things because of lack of time. Look at the lecture slides and their notes for a bigger picture.

We study sensation and perception, object recognition, and attention in this order because we start with the simple matters and slowly build up to the higher order concepts. First we should clarify the central terms in these chapters (5-7). Transduction is when a physical form of energy is translated into the language of nervous system. Sensation is when we are aware of the presence of an object, without necessarily knowing what that is. It could also be seen as an activation of a sensor organ or sensory cortex, without necessary awareness. Perception is an extra layer over sensation. It allows us to discriminate it, apply judgment over it, and a sort of interpretation (this is a yellow large car, therefore it is an ambulance). Recognition is when there is a relationship between the object in front of us and a representation of it in our memory.

Many times we use patient studies to get insight about the neural mechanisms of perception. One type of deficiency is agnosia, which is a failure of recognition. There are three types of visual agnosias that were important for the discussion of vision and perception:

1. Apperceptive agnosia is the failure of assimilation of your own knowledge to the object in front of you. For example, when hats are presented in proper lighting and are spaced from each other, the patient will differentiate them, but under darkness, it would be difficult to see the difference. A friend with a new haircut might not be recognized.

2. Integrative agnosia is the failure to infer the boundaries between objects and infer which boundary belongs to which object. The patient can't see the whole as its components, because the borders between them are not clear.

3. Finally, associative agnosia is when the patient fails to associate background knowledge with the object in front of him/her. For example, the patient will recognize a hammer, but would not know what it is for.

In some patients, the impairment is selective for one type of category of objects (either animate/inanimate, living/non-living).

Sensation

Sensation begins with transduction, different for different senses.

• Hearing: hairs in the cochlea are sensitive for different frequencies of sounds causing them to vibrate. In the outer part they are hard and respond to low frequencies, and deep in the cochlea they are limber and respond to high frequencies. Their ordered responsiveness is considered tonotopical (the frequencies of sounds correspond with the locations of the hair responsive for them). The cochlea represents pitch (location of the activity), loudness (the intensity of that resonance), and timbre, which is the information about the secondary frequencies coming with the main primary frequency (this is dictated by the combination of pitch and loudness). This tonotopic representation also exists in the primary auditory cortex. However, it is not neat (slide 10). It is sensitive to 20Hz.-20,000Hz, and any tone out of this spectrum is inaudible. In auditory sensation, the anterior part is in charge of recognition of sounds while posterior parts of the location of this sound.

• Olfaction: a chemical sense (gustation as well), the only system that doesn't go through the thalamus. Every other system goes through the thalamus and then to their primary cortexes. The primary olfactory cortex was originally thought to be about the sense of smell, but studies of sniffing showed that activity there is actually about the act of sniffing and the sense of smell. The passive experience of smell is found in the orbitofrontal cortex. Some people have a problem with this, saying that smell is never passive because it is something that we do, just like touching.

• Gustation: papillae are engulfing the taste buds in the tongue. The taste cells underlying those taste buds initiate an action potential that eventually leads (through the thalamus) to the primary gustatory cortex in the insula. One study tried to dissociate our motivation to eat more of the pleasantness we feel about food. They noticed that motivation drops faster than pleasure, showing the difference between them. We also find separating evidence in neural studies. The orbitofrontal cortex has increased activity with eating- interpreted as aversion increased with eating (decreased pleasure), while the primary gustatory cortex has decreased activity- satiety (decreased motivation).

• Somatosensation: the perception of mechanisms affecting the body. Note: learn the name of the different cells. Meissner's corpuscles adapts fast, responsible for light touch, Merkel's cells adapt more slowly, targeting long-lasting touch, Pacinian corpuscles are responsible for sensing vibration or deep pressure, Raffini's corpuscles are important for temperature sensation, and nociceptors are free nerve endings that sense pain.

• Vision: the most extensively studied sense. Relatively easier to study and most prominent. When light hits the back of our eyeball (retina) and activates photoreceptors, having photoreceptors that release a protein compound that initiates action potentials. The closer you are to the fovea (center of retina), the more focused vision is. This is where almost all of the cones exist. Cones come in three types: short (blue-like), medium (green), and long (red). Those are connected to bipolar cells, which are connected to the optic nerve, which relays to the thalamus. There are 260 million photoreceptors, while we have 2 million bipolar cells, meaning many photoreceptors share bipolar cells. Rods in the periphery of the retina are color-blind, and converge from many to one bipolar cells so that their input is summed up to create a signal- less is needed, while cones take a lot more to create an action potential, and fewer of them share a bipolar cell. In this case, the lower later of the cells know what is going on in the upper layers, which enables an acute picture. However, since there are more rods sharing cells in the periphery, there is less acuity there. This is why the best resolution for vision is when we focus the picture on our fovea.

Perception: Vision

We perceive color by the dominant wavelength. Brightness corresponds to the intensity of this wavelength, and the saturation of it is the purity of it. The primary projection pathway provides 90% of the visual information. There are two tracts leaving the eye (nasally and temporally), and they project through the optic chiasm contralaterally and ipsilaterally, respectively. Conscious visual experience including the color and recognition of the object happens in the ventral pathway, and the dorsal is specialized for vision for action (how far you should reach, the firmness of the grasp). Evidence for this difference is provided by studies in humans and monkeys. In monkeys, for example, we can indicate perception of an object by eye movements, but this is also important for vision for action. The patient D.F. is a lady who had a brain lesion in the ventral pathway and some of the primary visual cortex. She gained some vision back, but could not name objects. When asked to imitate the orientation of a cylindrical object, she failed, but when she was asked to enter the card to it, she succeeded (slide 24). She had visual agnosia, an inability to recognize an object by vision. Optic ataxia, on the other hand, is an inability to orient oneself in the environment (problem with the dorsal pathway).

The region V4, a part of the ventral pathway, is responsible of colors and shapes. The deeper we go in the pathway, the cells become more selective about the features they need to be activated (color and shape). Vision could be lost unilaterally or bilaterally. The MT area (V5 in humans) responds to motion, based on two features: direction and speed. The cell that is responsive to dots moving upwards will not respond to dots moving downwards. It is also possible that this area is a byproduct, which requires more research. Therefore researchers decided to stimulate specific cells in this area to see what happens. Indeed, stimulating one neuron was followed by an upward movement illusion. Motion perception is only lost when it is damaged bilaterally.

Deficits in Visual Perception: Scotomas

Selectivity of cells causes a deficiency very specific to a certain movement if one of those cells is damaged this is called scotoma.

Multimodal perception

When the McGurk effect occurs, the person's lips determine what the person is hearing (video in slide 33). This is merely one of many examples for multimodal processing, where information is processed better coming from different modalities, and in which some modalities are dominant and could influence the perception of other modalities.

There are neurons in our parietal cortex that are specialized for sensation in a specific spot. If there is a visual stimulus placed on the same spot, the same neuron will be activated. Unitary sensory cortexes communicate with each other, so the visual cortex is not only devoted to vision. This can explain the McGurk effect. Another example for multimodality is synesthesia.

Perceptual reorganization

In blind patients, fMRI studies showed that their primary visual cortex was activated during brail reading. In patients that were blinded for five days and taught to discriminate tactile stimuli, their visual cortex was also activated during tactile stimulation. These examples are both evidence for neuronal plasticity in the adult brain.

Object Recognition

This persists despite sensory variance. Even if an elephant is pink we will recognize this is an elephant. We can also identify an object from different viewpoints. The features of the objects could be described as its different components or as its functional features. The latter are in both the ventral and the dorsal pathways, while the former is traced to the ventral pathway.

In certain patients with categorical problems in recognition- this could happen because animate objects are more detailed and require more knowledge from the semantic memory. In addition, there is a special characteristic of faces- directed at the fusiform area. A large variety of faces activate this area, and damage to it selectively damages the ability to perceive faces (right damage mostly). This is called prosopagnosia. The right side damage is a hint here. Face perception is one type of holistic damage. In contrast, analytic perception is the emphasis on the components versus the parts. It seems that the fusiform face area is about the holistic view of faces. So is there an area important for faces' analytic perception?

Paying attention to the face in a task requiring judgment could help us find an answer to this question. Locked in patients seem to be in a coma but are not (1/20 comatose patients): one way to communicate with them is to ask them to imagine something that has a specific neural stamp (imagine that you are playing tennis), detected by fMRI. This enables communication with those patients.

Attentional Resources

The anterior areas of the brain are more related to top-down processes.

Bálint's syndrome is an impairment of attention, where the patients can only focus on one thing at a time. When you present two objects, they recognize one. They are not able to disengage what they are paying attention too. This could be seen as a problem engaging in the periphery (being conscious on one thing at a time). Unilateral spatial neglect is results from a damage to the right cortex, resulting in neglecting one side of space (e.g. forgets to shave one side).this is different from Bálint's syndrome because they could experience their left side, but it requires more effort. They perform extinction- when there is something on their right side, they ignore their left side. Otherwise they could notice their left side.

Paying attention modulates visual processing as early as the thalamus- which goes against the idea that sensation is at the beginning, and attention comes later.

Inhibition of return is when attention is reflexively directed at one side of space and then for a second we don't see anything there. After that the visual cortex inhibits that side of location (there was nothing important there, so I shouldn't address it). The reason for this is still unknown.

Lecture 4: Action Control and Executive Functions

This lecture focuses on the Prefrontal cortex (PFC) and the motor areas.

Action Control

Despite the differences between your dominant and non-dominant hand writing, there are a lot of similarities between them, showing common control of both of them. Each executed action is complex and has a rich repertoire of representations in the brain.

When we look at the hierarchy of motor control (look at slide 4), we begin with muscles themselves, followed by motor neurons connecting to the spinal cord (releasing acetylcholine to contract neurons) but in this spinal cord we can see many patterns of control. Higher there are dedicated systems specializing in controlling actions that are crucial for life (blinking, swallowing), which are in the brainstem. The psychological meaningful areas are next: thalamus as the relay station, controlled by the basal ganglia and the cerebellum. Higher we can find the primary motor cortex, directly connected to the spinal cord, but also indirectly, through the cerebellum and the basal ganglia. The primary motor cortex (PMC) is the command giver. The supplementary motor area (SMA), which is more lateral, and premotor cortex are in charge of coordination. The PFC is important here for action control because it represents a goal people have. But the PFC is not enough: it collaborates with posterior areas (parietal lobe for example). So there are many areas involved in any small movement. Of course, when it comes to reflexes, fewer areas are involved, because it is only controlled by the spinal cord.

The spinal cord

In the middle there is the gray matter, representing cell bodies. Efferent neurons leave the spinal cord from its ventral side, releasing ACh to contract muscles. Then affective tract is found dorsally provides information about the location of the hand (feedback) for example, from the muscle back to the spinal cord. The power of the spinal cord is illustrated by a headless chicken continuing running with no brain control. So apparently we don’t need the efferent and afferent pathways from and to the brain respectively to create simple movements like walking.

Primary motor cortex (M1)

There is a somatotropin representation in the PMC, with contralateral control. The primary motor cortex receives information from the premotor cortex, the supplementary motor cortex, the parietal cortex, and the thalamus. It sends information to the spinal cord, the cerebellum, the basal ganglia, and the brainstem.

Different parts in the PMC will be activated representing different directions (this is called direction codes). Activity of several cells can indicate a specific direction. This is an intention that can be recorded, even when the person is paralyzed and cannot perform the motor action. In robotic arms, larger brain areas are used to show intention of movements that can be then translated to movements by the robotic arm. This is a technique of brain computer interfacing. ECoG is the part recording the information used for this kind of prosthetic limbs. Forward copy (efferent copy) is a message sent to other parts of the brain, preventing them from being surprised of sudden movements.

Cortex, Basal Ganglia, Cerebellum, and Thalamus

In slide 9 there is a picture of the basal ganglia. The striatum is sending the commands, and those pass through the globus pallidus, and the subthalamic nucleus has an opportunity to send direct commands later. The substantia nigra is a dopamine releasing nucleus which is very important for movements. An organogram is the description of interconnection between different brain areas. Tan organogram for motor control can be found at slide 10. It describes three pathways of action control.

1. The first pathway is called the internal loop. It uses internal information unrelated to the environment to create motor commands (like talking). The PFC, along with the parietal association areas send messages to the basal ganglia directly and indirectly through the SMA. The SMA also sends messages to the PMC, which receives indirect messages from the basal ganglia as well, through the thalamus. The PMC later sends messages to the spinal cord, and from there a command is sent to the muscles.

2. The second pathway is the external loop. The cerebellum and the premotor use external information to adjust commands to fit the movements to the environment. The cerebellum also uses the thalamus to relay its messages to the PMC.

3. The third pathway is the feedback loop, which helps updating the information needed for the external loop. It uses information from the visual and auditory cortexes being sent to the PFC and association areas. The spinal cord also sends information to the PFC and association areas, but also to the cerebellum and directly to the PMC.

Basal ganglia

An organogram of motor control by the basal ganglia is illustrated in slide 11. Dopamine is released from the substantia nigra, reaching the striatum to two types of receptors: excitatory and inhibitory. It influences movements directly, releasing dopamine, but there is also an indirect route passing through the globus pallidus, later controlling the cortex. This route is very slow. Both routes protect the body from impulsive behavior. We can see the direct route as giving the 'go' command, and the indirect route giving the 'no go' command. In some situation the PFC gives commands to the subthalamic nucleus, creating a shortcut to the indirect route.

In Huntington's disease, multiple indirect route parts are damaged. The direct route then dominates, making behavior impulsive. One symptom is chorea, which is a sudden, abrupt movement.

In Parkinson's disease, the direct route is inferior, resulting from a loss of dopamine producing cells in the striatum. This causes difficulties in initiation or switching of motions. People are also interestingly losing cognitive flexibility.

Supplementary Motor Area

The SMA uses the corpus callosum for collaboration of the two sides of the body for order and timing. When a person is trying to perform two actions at the same time, one action with each hand, it is possible, but switching between hands becomes difficult. This is because of the SMA. It is resource demanding, so when a task becomes too difficult the default would be parallel symmetric activity. When it is damaged, the two hands are acting in two unrelated manners.

Cerebellum

There is a distinction between different nuclei in the cerebellum, and between the different systems in it. While the central part of the cerebellum is associated with balance (spino cerebellum), the lateral areas are associated with our limb movements. It is organized ipsilaterally, in contrast to the cortex which is organized contralaterally. It is still unknown how exactly it controls bodily functions. We only know that there are very strong connections between the cells of the cerebellum.

The primary functions of the cerebellum are integration, control, and timing of current movements. The superior part is in charge with vision and proprioception, the medial nuclei send descending neurons to the brainstem and the spinal cord, the lateral part sends ascending axons to the thalamus, and from there to the motor cortex, the neo cerebellum, which is inferior to that, receives information about moto precision from the PMC and the sensory and association areas, and the most inferior part is the vestibule cerebellum, and it is in charge of stability of visual images and maintaining balance.

Lateral Premotor Cortex

In the lateral premotor cortex that collaborates with the cerebellum, there are areas for specific types of basic movements: reaching, grabbing, and pointing. This is one of the first areas in which mirror neurons were indicated. If someone is performing an action, the same cell groups will be activated as if the person is viewing someone else performing the same movement. This helps us understanding other people, empathizing with them, and imitating, learning new actions. Later they were found in other areas.

The preparation for action occurs multiple milliseconds before the action is performed, and this can be seen in the readiness potential in the motor cortex.in Libet's famous experiment about this subject, he measured the readiness potential in the motor cortex of his patients and asked them to press a button, while indicating the precise clock position when their choice to press the button was made. He noted that the motor activation was made 300ms before the actual intention. He concluded that the person's awareness of a choice occurs only after the motor command was given.

Posterior Cortex

Several posterior areas also influence whether a person will be able to perform a motor action. For example, there are two pathways that are meaningful for motor functions. The ventral route, also known as the 'what' pathway, is important for integration of visual information. If it is damaged, objects will not be recognized, but the person will be able to act on them. This is known as visual agnosia. The dorsal route is known as the 'how/where' pathway, and it is important for motor commands over visual information. In visual ataxia, damage to that pathway will lead to difficulties reaching objects. Apraxia is an inability to perform an action. It could happen because of inability to perceive, resulting from damage to the dorsal route, or it could result from damage to the motor cortex.

Cognitive Control

The PFC is involved when there is a competing tendency over possible reward. It is also important for goal maintenance. The PFC makes use of its interconnectivity with the rest of the cortex to make decisions. It is the most connected part in the brain. To perform decision making, it has to make use of information coming from distant resources. It can be used for different purposes, including working memory, integration, and rule execution.

The ventromedial area is involved in decision making, emotional regulation, and identity (personality relating to goal). The lateral PFC is more related to cognitive, rational functions. The PFD is the most sensitive part in the brain. It is slow to mature in children (the dorsolateral PFC only matures after age 20), and loses neuronal connections and white matter in older age (60+). It is also very sensitive to intoxication by alcohol and drugs, high stress, low levels of oxygen and glucose, fatigue, and current workload. These all lead to lack of flexibility (ventromedial PFC). Damage to the PFC could lead to default behavior rather than goal-oriented one. Problems associated with inferior performance of the PFC are impulsivity, and rigidity of behavior. The ventromedial PFC is related to the ego type of performance, setting goals for oneself. Miller and Cohen described it nicely: The PFC serves as a representation of goals, and based on those goals it is determined what will be inhibited and what will be executed. Goal-directed actions are not always prominent. They could become automatized after a while Controlled processes require more resources, but they allow more flexibility than automatized processes. It is hard to change the way you do something that you have done so many times. The distinction between controlled processes and automatic ones is a matter of time and repetition.

The involvement of the PFC in the working memory is illustrated by a study done in monkeys in a delay match to sample task. They taught the monkeys to fixate on a centered X and remember the location of target being presented on the screen. Later, the monkey has to move their eyes to the direction of the location. The dorsolateral PFC is activated even when there is no information present, pointing to activation of the working memory. This was also demonstrated in humans many times. One example is of task switching. Previous trials in a task will reduce performance on a changed task.

Another function of the PFC is response inhibition, which is illustrated by a baseball player preventing himself from hitting a ball if it has not arrived yet until the very last moment, so he can hit it accurately. This kind of response inhibition is indicated in the right ventrolateral PFC. Inhibition is also related to stopping an automated response or ignoring a distractor to perform a task accurately (selective attention). This kind of a conflict between a target and distractors can be overcome by experience. After performing an error, the person is immediately aware of it, which leads to a strong negative potential that is related to fixing that error. This also happens when people are confronted with negative feedback. This increases that chances that the next trial will be correct. This negative potential arises at the anterior cingulate cortex. This area is connected to dissonance between what a person wanted to do to what they did. This indicates that more control is needed to avoid more errors. The anterior cingulate cortex receives feedback about the action performed and uses the dorsolateral PFC to increase cognitive control. An organogram of the PFC and action control is depicted in slide 33.

There are three dimensions in the PFC:

1. Dorsal/ventral: in dorsal, more manipulation of working memory, while in the ventral area of the PFC there is more maintenance of the working memory. This is similar to the distinction between the where and what pathways.

2. Anterior/posterior: while the anterior part of the PFC is involved in higher, abstract processing in the brain involved in coordination of complex tasks, the posterior is involved in lower processing and simple commands.

3. Medial/lateral: emotional actions are controlled in the medial area of the PFC, while more external, rational processing is done in the lateral parts.

Illustrative Tasks of Cognitive Control

There are several tasks used to diagnose problems in the PFC:

• Tower of Hanoi: people have to move a pile of discs, one by one, so that no larger disc is put on top of a smaller one. This is a task of planning ahead and working memory by mental simulation. The number of steps required for this and the number of errors are indicative of these abilities.

• Wisconsin Card Sorting Task: there are sets of cards that could be sorted by shape, number, or color. The examinee is not told the rule, but only if he did it correctly or not. After managing to find the rule and sorting correctly several times, the rule changes. This is a test of perseveration.

PFC and IQ

There are two types of intelligence: crystallized and fluid. Crystallized intelligence I about knowledge acquired from experience. There is a weak correlation between this and PFC activity. However, fluid intelligence, which s known as inductive reasoning, is an ability to solve problems, and this is tightly related to PFC functioning.

Lecture 5: Memory

Memory is a central aspect of cognition, connected to any other cognitive process. It is tightly connected to learning, behavior, decisions, basic functions like understanding language, speaking, recognizing emotions, and social interactions. It is crucial for our perception of ourselves.

Encoding is the most basic process of memory formation. It starts with acquisition, which is the formation of the representation of the memory, and the memory is then consolidated. This is the process of stabilizing the memory. Storage puts the memory in long-term memory, and retrieval uses memory from the long term memory. There are three types of memory which vary in time: sensory memory is very shortly stored (less than one second), short-term memory (STM) lasts between a few seconds to a minute, and long term memory (LTM), in which representations are stored for as long as years. Information stored in the long term memory is not currently active unless it is retrieved.

Sensory Memory

Sensory memory is active in different modalities, such as echoic memory of audition or iconic memory of vision. It decays rapidly, but we can access it for a little while. It has high capacity, with a short sensory representation which is not always perceived. Iconic memory can be tested with a short (50ms) presentation of an array of letters. People are able to repeat some letters after a delay, although they do not consciously perceive them. This could work up to a delay of half a second. The brief representation of visual information is important for identifying changes in the scene. Without a continuous representation of the visual scene it is difficult to perceive change (change blindness).

Echoic memory decays a little slower. In the oddball task there is a series of identical tones with some deviant tones. EEG scans in the auditory cortex show that there is a greater negative potential for mismatched tones in this paradigm. This potential fades away slowly (about 10 seconds).

Short-Term Memory (STM)

The STM has a longer time course, but its capacity is limited. It is about what is going on now. There are different stores for different modalities of STM. Phonological STM, for example, differs from visuospatial STM.

Phonological STM is assessed by span task, in which a person has to repeat a series of digits. There are different beliefs about how limited the STM is. Miller believes that the average amount of items a person can hold in STM at a moment is believed to be 7/-+2. Items could be meaningful chunks- numbers, words etc. On the other hand, Cowan believes it is actually lower, and that a higher number could be achieved by rehearsal. Indeed, when this rehearsal is prevented, the limited capacity decreases. Phonological properties of the item also count. Span length decreases if the item has many syllables and if the items are very similar to each other, thereby interfering each other's maintenance in the STM.

Visuospatial memory is assessed by the Corsi block-tapping test, in which the examinee participant has to repeat the taps an examiner is performing, in the same order. This test has a spatial-sequential component, so it is not purely visual. Another test is the visual-pattern test, in which the examinee is shown a matrix of colored squares for about 3 seconds, and then he or she has to reproduce it. Performance on the two tests does not correlate perfectly, and they relate to different lesions.

According to the Modal model, which was followed for a long time in the cognitive psychology domain, there are three sequential stages of memory (sensory, STM, LTM), and information could get lost in any of them. In order for a sensory memory trace to enter the STM, it has to be attended. To arrive to the LTM it has to be rehearsed. In any step the information could decay or be interfered. Some aspects of this model do not seem to be correct, since there are patients with good STM who are lacking LTM, and vice versa. This suggests that LTM could exist without STM. This limitation lead to the development of the idea of working memory (WM).

Working Memory

The WM replaces the STM. The limited amount of information there is not only stored, but is actively used, monitored and manipulated. Another new aspect to this stage of memory is that information arrives not only from the sensory memory, but also from the LTM. There are several models that describe the WM.

Baddeley and Hitch came up with a model consisting of three components: visuospatial sketchpad, the central executive, and the phonological loop. The visuospatial sketchpad and the phonological loop are both used to store information from the sensory areas, and the central executive manipulates it. Baddeley later revised this model by adding the episodic buffer, which uses personal information from the LTM to the equation. In addition, the visual semantics component adds internal information to the visuospatial sketchpad, and language component does the same to the phonological loop. The central executive supervises over input from multiple modalities, and they are all integrated to create one story. This maintenance and manipulation of information is related to activity in the frontal cortex, especially the lateral prefrontal cortex. Evidence for this comes from several sources. The lateral PFC is active during working memory tasks, single cell recording in monkeys show activity in this region in these kind of tasks (the monkey has to move his eyes to a location of a target on a screen after a delay of its presentation), and working-memory related problems arise after lesions to this area. It seems to be involved in retrieving LTMs and maintaining them in the WM. The dorsolateral PFC is also implied to be involved in WM, in tasks requiring manipulation of information. The storage of information is related to more posterior areas, in areas involved in information processing from different modalities.

Long-Term Memory

Long term memory can be divided into several subtypes. There is a clear diagram of these types in slide 18. The most basic separation is between declarative memory and nondeclarative memory. Declarative memory is explicit, and it can be farther divided into episodic memory (of personal events) and semantic memory (facts that we cannot relate to a certain event). Nondeclarative memory, on the other hand, is implicit, and can be farther divided into procedural memory (which relates to physical skills we can perform without thinking about them), perceptual representation system (priming perceptual information), classical conditioning (basic associations between different types of stimuli that can be seen in bodily reactions as well, even unconsciously), and non-associative learning (even simpler, including habituation and sensitization, which are observable in our behavior but not important for cognitive neuroscience). Different memory types are indicated in different brain activity. Declarative memory is seen in the medial temporal lobe, the middle diencephalon, and the PFC, procedural memory is seen in the basal ganglia and the cerebellum, perceptual representation is seen in the perceptual and association neocortex, classical conditioning is seen in skeletal muscles, and non-associative learning is seen in the reflex pathway.

Amnesia

H.M. is the most famous amnesic patient. He suffered from severe epilepsy and had both of his medial-temporal lobes removed, including the hippocampus. He was unable to learn any new information after his surgery, and lost 10 years of his previous memory- severe case of declarative memory. He was studied broadly. The hippocampus is about 4-5 cm in humans, and the amygdala is right in front of it. It important for remembering emotional information. Also important for memory are the hypothalamus and mammillary bodies. Lesions to these areas could result neurosurgery (like in the case of H.M), stroke, head injury, viral infection and nutritional deficiencies, like in Korsakoff's syndrome. In H.M., an MRI was performed 40 years after the surgery. There was a lesion that was originally believed to be 8 cm, but then discovered to actually be about 5 cm long. There was still a lot of damage.

Hippocampal damage leads to loss of declarative memories, especially episodic memory. In amnesic patients we look at the amount of time lost to their memories. Usually, anterograde amnesia is much more severely impaired. In hippocampal damage, the Ribot gradient is a phenomenon seen in retrograde amnesic patients, in which older memories are better retained. This suggests that the hippocampus is not solely responsible for the storage of memories. But several types of memories are unharmed in hippocampal damage. Short-term memory and working memory seem to be intact. Implicit memories are often intact as well, and this can be seen for example in procedural memory and non-associative memory, since amnesic patients can still perform old motor tasks and even learn new ones. Classical conditioning also seems to be intact. This is an illustration of how different memories are separately represented in the brain.

Semantic memory seems to be impaired. Information that the person have learned in the last few years is usually not remembered. This is still debatable, because some patients were able to learn new languages after their amnesia. It might have been because of spared tissues, so this area is still unclear.

Functions of the Hippocampus

Damage of hippocampal tissue leads to loss of declarative memory. This is typically explained by impaired consolidation. Consolidation is the process of stabilizing memories over time. This occurs through strengthening of synaptic connections, but this happens everywhere in the brain. Another bizarre thing about the LTM traces is their temporal gradient. These two issues are addressed in different explanations:

• Consolidation Theory: consolidation is a gradual process. During this process the memories become less and less dependent on the hippocampus, and more reliant on the neocortex. Therefore the hippocampus's role is temporary (although for a long time). Initially, different components of a memory are represented throughout the cortex in parallel in their specialized regions, and the hippocampus creates bonds between them over time. Therefore it connects different modules of a memory. After a long time, the hippocampus is not needed to maintain those connections, and then the memory is independent of it.

• It is about the difficulty of memories. Childhood memories are easier and more primary, and therefore it is easier to remember them. Some studies tried matching difficulty level across different time period and there was still a time gradient there, so this is not a satisfactory explanation.

• Later memories have not been properly encoded yet, therefore they cannot be retrieved. This could count for several amnesic patients, but sudden-offset amnesia is more difficult to explain. There is still a temporal gradient there.

• Older memories are not episodic anymore, but rather semantic, since they were rehearsed so many times.

• Multiple trace theory: each time we remember an event, a new trace of memory is created for it. Memories that have been retrieved many times are therefore less likely to be vulnerable to brain damage. This implies that a complete removal of the hippocampus will delete all memories, and partial damage will allow salving some memory traces of old memories in spared tissues. Unlike the consolidation theory, this theory implies that the hippocampus has a permanent role in LTM. This is based on amnesic patients with hippocampal damage that did not show the Ribot gradient.

This theory was later refined. The hippocampus appears to be crucial depending on the type of memory involved. If the memory involves contextual details, it is dependent on the hippocampus. Otherwise, the memory is more semantic and cortex-dependent. Every time we retrieve a memory trace, the details of the trace differ if it is dependent on the hippocampus. This does not happen for semantic memories. So according to this theory, a memory is transformed over time rather than transferred. It becomes more generalized and free of context.

Another key-role of the hippocampus is contextual, as we could see in the multiple trace theory, and specifically spatial context. Recorded hippocampal activity showed that activity in the hippocampus is selective for certain locations in the environment. With a lesion in the hippocampus, a rat dropped in a pool of water is unable to orient towards the platform it can leave from. It can only find it quickly from a location it has been dropped in several times before. This is an illustration of the cognitive-map theory¸ stating that the hippocampus has a spatial map of the external environment.

Functions of the Prefrontal Cortex in Memory

The ventrolateral PFC is important for retrieving and maintaining information in the WM. The dorsolateral PFC on the other hand is important for manipulation and monitoring the information in the WM. The manner in which information is encoded determines how well it will be retrieved later.

Ventrolateral Prefrontal Cortex

The level of processing is important to determine this. The deepest level of processing is semantic level, in which a person elaborates on the information being processed. There seems to be a distinction between the left PFC, which is more important for language, and the right PFC, which is more important for faces and nonverbal information. I subsequent memory paradigm, participants memorize words one by one while their brain is being scanned in an fMRI method. The participants then need to indicate from a list of words which ones they had recognized from the memorization process. This paradigm shows difference in fMRI activity during the encoding of information that will be remembered to one that will be forgotten.

Dorsolateral Prefrontal Cortex

The dorsolateral PFC is especially active when the retrieval demands are high, when the person is asked about context rather than content, and when there is high level of conflict/uncertainty. Tip of the tongue states also activate the dorsolateral PFC, especially the right one.

Evaluating Retrieved Memories

The PFC is responsible of monitoring where the memory is coming from: is it imagined, or happening in reality? The dorsolateral PFC is important for this kind of contextual information of the memory trace.

Curiosity

This part of the lecture is not mentioned in the book, but the lecturer said it is important. The more curious we are about a topic, the more likely we are to remember information about it. There are two types of curiosity: perceptual curiosity is basic and it exists in animals as well. It is evoked from unexpected, ambiguous stimuli. Epistemic curiosity, on the other hand, is the desire for conceptual knowledge. In a study that compared the free recall of participants in different states of induced perceptual curiosity showed that relieved curiosity boosted the score on free recall. This curiosity was indicated in hippocampal activity. In a study that looked into epistemic curiosity, participants were asked a question, and then asked how curious they were about the answer, and in between were showed a picture of a face. Participants later recognized those faces better if they were at a curious state, and they were also better able to recall the answer the question they were curious about. This kind of curiosity was correlated with activity in the hippocampus, and the relief of curiosity was correlated with activity in the nucleus accumbens and in the substantia nigra and the ventral tegmental area. Therefore is seems that learning following curiosity is reinforced by the reward system, thus improving the memory trace.

Lecture 6: Emotions

There are three components to emotions: physiological reaction, behavioral response, and an affect associated with the emotional event. Describing emotions could be difficult. Some people are higher on emotional granularity, which is the individual's ability to describe the specific emotion they are feeling. In order to study emotions effectively, there have to be some agreement. However, different researchers talk about emotions in different ways. Some talk about basic categories of emotions, some talk about complex emotions (such as embarrassment), and some talk about dimensions of emotions.

Basic Emotions

This is a discrete approach, claiming that emotions do not overlap with each other. It was formed by Ekman, who traveled the world and recorded facial expressions cross-culturally. He found six basic emotions are happiness, sadness, anger, fear, disgust, and surprise (and neutral is the 7^th emotion), and that they are inborn and are present universally. He was inspired by Darwin, who noticed that different animals express emotions in a similar way.

Expression of Emotions

Facial expressions on themselves are not always straightforward. The context of the facial expression, what the person is saying, the tone of voice, and the person's body language give a lot more information about the actual emotion. Another thing that helps clarifying facial expressions is body dynamic (movements).

Criticism on Basic Emotions

Several researchers claim it is difficult to determine only six emotions. They are not always clearly distinct from each other. In addition, some more complex emotions are not covered by those basic emotions. It might be because they are longer lasting or have no clear facial expression identified with them (can you express jealousy with your face?). Moreover, those emotions are not associated with one brain area (neither do the basic emotions though).

As mentioned before, the approach of basic emotions is discrete, and an alternative approach is dimensional. In clinical practice, there are both dimensional and categorical (discrete) tools for diagnosing psychopathology (like depression; there is the DSM-V which is categorical, and the Hamilton Depression Scale, which is dimensional). There are benefits and drawbacks for both.

Emotional Dimensions

Emotional dimensions allow to differentiate between different levels of a certain emotion (how happy are you?). Emotions can be differentiated by valence (pleasant versus unpleasant), arousal (intensity), approach (or withdrawal). Positive emotions usually evoke approach behavior, while negative emotions evoke avoidance behavior, but this is not always the case. Dark tourism is an example for the contrary. In dark tourism, people are attracted to visit a negative or threatening situation. This is not unique to humans. The reasons for this are still unclear: empathy? Change of hierarchy? Learning opportunity?

Theories of Emotions

While there is a consensus among researchers about the components of emotions (physiological response, behavioral response, and an affect), theories of emotions differ on the causation order between the components, the exact experience of emotions, and about the sources of emotions.

• James-Lange theory: an emotional event causes a physiological reaction, and only after this does the awareness come (my heart is beating, I must be afraid).

• Cannon-Bard: how can it be possible that different emotions trigger different bodily reactions? The physiological reaction and the awareness should be parallel. He added that the thalamus is responsible for the perception of the stimulus, and the hypothalamus is responsible for the expression of the emotion.

• Appraisal theory: a person first assesses whether the stimulus is beneficial or harmful (cognitive appraisal comes first), and only then does the person have a physiological response. Following that is the subjective feeling of the emotion, and finally the action in response to the emotion.

• Singer-Schachter: the perception of the situation preceding the physiological reaction influences its cognitive component that will come after the perception of physiological reaction. The interpretation is what matters here.

• Constructivist theories: the cultural and linguistic factors are very influential. Emphasis within the language and the cultural norms are needed to understand the emotions of the individual. The emotions are constricted to the social structure of the individual. Therefore, a perception of an emotional event is followed by a physiological response, which combined with the past experiences relevant for that event, lead to the experience of emotions.

• Evolutionary psychology approach: emotions can be understood in terms of their contribution to success in survival. They put into action a set of cognitive programs that set actions proper to the situation. The perception of a stimulus triggers an ancient program that creates an emotion appropriate for the situation.

Almost all why questions have an evolutionary answer. We have emotions because they tell us what is important for our survival. It is important to perform cross-species research to find these answers.

• LeDoux' high road and low road: there are two routes of processing emotions. The low route is fast and dirty and it is sometimes inaccurate in the response it makes. It goes through the amygdala. The high route is slow and more accurate. There are different distinctions regarding the brain areas concerned here, and there is still no consensus about them.

Limbic System

The limbic system is a little of an outdated term, there is no single brain area that performs a single cognitive function. However, when it is mentioned in the literature, it usually involves some areas that are specialized in processing emotions: the amygdala, which is important for salient biologically relevant external information, including emotions, the insula (disgust), the orbitofrontal cortex (OFC), which is important for making decisions, and the anterior cingulate cortex (ACC), which is important for empathy and pain processing.

Amygdala

The amygdala is a nucleus in the medial temporal lobe. It consists of many different nuclei, which are difficult to separate in fMRI studies, namely the cortical nucleus, the anterior nucleus, the lateral nucleus, the central nucleus, the medial nucleus, and the basal nucleus. In the 1950s', Klüver and Bucy damaged the amygdala of monkeys, and those monkeys lacked fear (this is now known as the Klüver-Bucy syndrome). Those monkeys even approached snakes again after being bitten, showing they were unable to learn fear. This shows how amygdala is important for emotion processing. It is also interconnected to many different areas.

SM is a patient with an amygdala lesion. She was not emotionless, but lacked any fear response. Her lack of fear response puts her at risk, since she is unable to learn to avoid danger. She lacks fear conditioning.

Interaction between Emotions and other Cognitive Processes

This might be the second of importance point of discussion in the emotional field. Fear, for example, can be learned both implicitly and explicitly.

Fear Conditioning

The case of little Albert illustrates classical fear conditioning. When a rabbit, which was not feared by Albert at first, was paired with a loud sound several times, the rabbit itself became scary. In fear conditioning, the repetitive pairing of a conditioned stimulus, which is neutral at first, with an unconditioned stimulus, which is naturally scary, turns an unconditioned response to the condition stimulus, which is neutral, to a conditioned response, where the conditioned stimulus becomes scary as well.

Damage to the amygdala impairs the conditioned response. The lateral nucleus of the amygdala, which gets input from several brain areas, is also connected to the superior dorsal lateral amygdala, which has cells that change the connection between the unconditioned stimulus to the conditioned stimulus. This creates an association between the conditioned stimulus to the conditioned response.

As mentioned before, there are two visual pathways for a stimulus to be processes in the brain. The normal visual pathway is salient when there are no emotional stimuli in the scene. It involves object recognition in the ventral stream and assessment of location and size in the dorsal stream. This is the high route, which is slow and accurate. In emotional situations, however, the low route is salient. It is unconscious and fast. The visual input goes from the eyes to the amygdala, which allows an immediate action (fight or flee), before the stimulus gets into the visual cortex and into awareness. An illustration of the two routes can be found at slide 54.

The low, fast route is unconscious. This shows the implicit role the amygdala has in fear conditioning. Amygdala lesioned people can report the conditioned pair of stimuli, but do not fear it. They could predict that a shock will follow light, for example, and they have the natural response to the shock, but they will not try to avoid it and will not show fear. On the other hand, there are hippocampus lesioned patients who acquire the conditioned fear response, but they don't know why. This shows a double association: the amygdala is important for the acquisition and expression of implicit emotions, and the hippocampus is important for explicit memory of the emotional properties of the stimulus. Those two areas are closely connected. In several instances, the fear response is not based on experience, but rather on explicit warnings.

Explicit Learning of Fear

Researchers were interested in the possible role of amygdala in explicit learning of fear. To study this, Phelps measured skin conductance responses in amygdala damaged patients after explicitly telling them they will receive a shock after seeing a blue square. Those patients showed no startle response at all, in contrast to control patients, but they knew that a shock might be coming (they never got the shock). This lead to the conclusion that the amygdala is important for explicitly learned fear as well. The hippocampus is still important for the explicit representation of the implicit fear activated by the amygdala. Might it be possible that amygdala is important for creating explicit memories in the hippocampus? In declarative memory, emotional memory is remembered a lot better than neutral one. This might imply that the amygdala has a role in improving memory.

In the Morris Water Maze, if you arouse a rat before throwing it in the water, it will find the platform faster. If you lesion the amygdala of the rat, it will still be able to learn the task, but the arousal will not help it anymore. This enhancement of learning by arousal can also occur after the learning phase, when the memory is being consolidated. This was shown by lesioning the amygdala right after the learning, which eliminates the effect of arousal. These findings illustrate that the amygdala modulates the learning process done by the hippocampus, and enhances consolidation of the memory in the hippocampus. It was also found that the amygdala interacts with the hippocampus when information is being encoded in memory. The right amygdala seems crucial for arousal-modulation of memories, more so than the left amygdala. Higher activation in the amygdala implies stronger memory of the event. On the other hand, excessive stress impairs memories.

The attentional blink paradigm

This is a paradigm often used in emotion research. A string of letters and letters is presented quickly. It is more difficult to recognize the second target (letter) if it is presented within 500ms of presentation of the first target. However, this is not the case when the target is an emotional word. Emotional content can override this attentional blink effect. This is mediated by the amygdala, since amygdala lesioned patients do not have this override.

Damasio's Definition of Emotion

Damasio connects emotions to decisions. According to him, emotions are important for making decisions about survival. Emotion is an internal process putting movement into action, internally or externally. It consists of a set of alternations in the body, which aim at avoiding danger or taking an opportunity. Feeling can be seen as the process of perceiving what is going on in the body during an emotion. Damasio called this hypothesis the somatic marker hypothesis. He stresses the importance of bodily sensations during an emotion, which are important for making choices, and those bodily sensations form the somatic marker. An important brain structure for emotion perception is the insular cortex. It is not the most important part of the brain for emotions, but it is active through several different emotions. There is also interplay between the amygdala and the orbitofrontal cortex, which is important for remembering the situation causing the somatic marker. An illustration of this is found on slide 79. The perception of the stimulus leads to thoughts and evaluations of it (cognitive component). Those thoughts and evaluations are performed in the PFC and the amygdala, and they lead to bodily responses, which lead to the way they are perceived (in the somatosensory cortex. The thoughts and evaluations are also influencing the manner in which the somatic marker is perceived.

This theory was also demonstrated in the Iowa Gambling task. Participants were able to choose cards from four different decks. There were two bad decks, where they could win 100$, but lose per 10 cards 1250$. On the good deck, they could win 50$ and lose 250$. Healthy subjects learned and chose from the good decks. Patients with amygdala lesions or orbitofrontal lesions kept on choosing the risky decks without learning. In addition, while some normal subjects took risky decisions and showed skin conductance response (showing stressful response), this was not the case in patients with amygdala or orbitofrontal cortex lesions. This was also tested in ventromedial PFC lesioned patients. Those patients showed lower skin conductance response under the disadvantaged punishment situations (choosing the bad deck and losing money), but also normal skin conductance under the reward situations and the punishment in the advantageous situations. Both damages can increase risk taking, but they are not the same. The OFC is responsible of the decision making, while the amygdala is more in charge of the somatic marker of the emotion on the decision. This is illustrated in slide 85.

To conclude, the amygdala is involved in in implicit fear conditioning, in explicit emotional learning, in recognizing salient social responses, and in attention to threatening and highly positive events. It is closely interacting with the hippocampus and the OFC.

Lecture 7: Social Cognition

Emotions are usually connected to social situations.

Role of Orbitofrontal cortex

The famous case of Phineas Cage, who got his orbitofrontal cortex injured in an accident, illustrated the role of orbitofrontal cortex in many social functions, namely goal-directed behavior, social inhibition, tolerance of frustration, aggression, empathy, and emotional warmth. He was no longer the same person. He formed the beginning of the research about brain areas and personality.

The prefrontal cortex consists of different parts. The ventrolateral part, along with the dorsolateral and the paracingulate cortex are involved in self-referential processing, while the ventromedial PFC and the orbitofrontal cortex are involved in decision making.

Learning about social behavior

We can do this by many means: looking at the dysfunctions related to brain lesions, psychiatric disorders involved in social impairments, looking at developmental perspectives, and using the comparative view.

Autism

Individuals suffering from autism have social and communicational problems. Autism develops in a very early developmental stage. It involves also difficulties in theory of mind (mind blindness).

One finding in research about autism is that children with autism can recognize emotions, recognize stimuli, and discriminate lexical stimuli, but cannot differentiate the positive versus negative facial expressions. Another study found that children with autism failed to create eye contact in social interactions. Autistic individuals are also less likely to make eye contact. While control subjects focus on eyes of another person, autistic individuals focus more on the mouth.

Neural correlated of autism are difficult to find. A meta-analysis found that social-related areas in brain of patients suffering from autism are less activated. It seems that they are less interested in social stimuli. There might be less connection between the hemispheres (smaller corpus callosum).

This teaches us that the human brain is social by default. One study compared default mode networks with social activated areas in the brain and found they overlap a lot. In fact, in black macaque, the social group the monkey is living in influences the size of the social brain activity. This makes sense from evolutionary point of view.

Theory of Mind

A theory of mind helps us to predict other people's emotions and behavior and empathize with them. Autistic people do not read the theory of mind of other people. This can be tested. For example, the Sally-Anne test in children. This test is passed by healthy 5-year-old children. This is a test to see if the individual is able to predict another person's false belief, even if it's different than their own. Autistic individuals normally do not pass this test. The Second order Theory of Mind test tests the same thing with a more complex story. In addition, the mirror test has been used to test animals' theory of minds. This is a test of self-awareness. While it is not a direct measurement of theory of mind, self-awareness is very important for social interaction and therefore they are comparable. The nose of the animal is being painted and a mirror is put in front of them. If the animal touches the mark on their nose, it suggests they understand the reflection is the animal itself. This is performed and passed by 2 year-old children as well. However, non-western children do not pass this test even at the age of 6. Findings in the lab do not always reflect reality, for example, while a dog that usually shows understanding of human behavior, this is not found in a lab.

Eye contact

Eye contact is very important for social interactions. Making eye contact, for example, is difficult for autistic individuals. Focusing on the eyes is natural for most humans; especially considering the amount of eye-white we have in our eyes. This assists us in following other people's gaze. The pupil dilation of our eyes also provides a lot of important information (arousal, light, friendliness). It is possible that this information we gather (without conscious prior knowledge) from other people's pupils is created through mimicking the other person. This is a natural reaction in humans, and through this mimicry we can empathize and understand what the other person is feeling. This is consistent with the idea of the facial feedback hypothesis, suggesting that the face we express gives us feedback on the way we are feeling.

Pupil mimicry is the synchronization of a person's pupils with the one's they are talking to. Is it the same in chimpanzees as well? In one study comparing humans and chimpanzees' eye mimicry, it was found that human participants synchronized their pupils with other human pupils, and chimpanzees synchronized their pupils with other chimpanzees. Eye mimicry seems to be important for the creation of trust. The trust game is a paradigm used to investigate this hypothesis: one participant invests in another, then the investment triples itself, and the invested participant can choose whether or not to give money back to the investor. The first observation was that participants invested more money in partners whose pupils enlarged, in contrast to partners with constricted pupil size. In addition, the pupil size was mimicked by the participants. Interestingly, this mimicry of dilated pupils was stronger during interactions with ingroup versus outgroup partners. In addition, mimicry of constricted eye pupils was stronger with the outgroup. The question is: can the small increase in trust be explained by pupil dilation mimicry? It seems so- but only during interactions with the ingroup. The same eye mimicry experiment was done with participants being scanned in an fMRI scanner. It was found that the temporoparietal junction (TPJ) is active when people have eye contact, it is somehow related to eye mimicry (still not clear how), and it is implicated in social decisions about trust.

Temporoparietal Junction

Another paradigm studying eye mimicry is hyperscanning. Hyperscanning is when two participants are being scanned at the same time during a social interaction between them and the synchronization between the two brains is being looked at. The area that was most densely correlated was the right TPJ, showing its important social attention function (joint attention). It is also important for social cognitive functions.

Joint Attention

Join attention is an important precursor of theory of mind. There are different types of joint attention, prominent at different developmental stages. The dyadic form is when an infant looks at an adult (2 months onwards). Next, 7 month-olds follow gaze of other people. Then they follow the direction they point, from 9 months onwards. This is helpful in learning language, developing theory of mind (missing in autistic children), and identifying other people's intentions. As mentioned, this ability is very difficult for autistic individuals, and it correlates with language abilities and social abilities. It was even implied as a test for autism.

The brain activation patterns comparing normal healthy controls to autistic children in joint attention shows importance of the dorsomedial PFC (the area that was correlated with group size in macaques) and the posterior superior temporal sulcus, a region found in many emotion studies, also activated during emotional changes in people. Across species, theory of mind can be found in some great apes, in dogs, horses, and several other species.

Perception of the Self

Can there be other people without a self-concept? Ich un Du is a book written by Buber claiming that there is no self without the other. We are constantly comparing ourselves to others, in terms of fairness but also in terms of personality, social roles, cultures, and situations. Leon Festinger, the theorist behind social comparison theory and cognitive dissonance, stressed the importance of comparing ourselves to others. This could be seen as conducting an experiment with a control condition. We can look upwards and downwards for a comparison, both could be motivating. Looking upwards could motivate people to do better, but the same can occur by looking downwards for comparison.

The self-reference effect is an illustration of the importance of perception of the self. When this effect takes place, people remember information best when it is related to themselves. This illustrates again that we have a social brain. When we learn something new, we integrate it into our existing knowledge, and the social brain is activated all the time (during daydreaming, ruminating, mind wandering, etc.). Therefore, we have most knowledge about ourselves, and self-relevant information can be integrated better.

Relevant Brain Areas

The mPFC seems to be important for this. Damage to this area is damaged, the self-reference effect is abolished. The temporal-parietal junction shares many components with mPFC. They are both social components are important for empathy and theory of mind, but the frontal area is more important for making decisions while the TPJ is more relevant for agency, multiple sensory integration, depersonalization disorders (in deactivation), and even out of body experiences.

Lecture 8: Consciousness, free will, and the law

Consciousness and free will are two fields in cognitive neuroscience that are difficult to study. The law is also interacting with those two.

Consciousness

The first reason of the difficulty in studying consciousness is the problematic definition of it. Some researchers claim there is nothing to investigate here. Others have been distinguishing functions of consciousness. Damasio, for example, claimed that consciousness is not wakefulness, but more than that. There are actually core consciousness (awareness of the present) and extended consciousness (narrative of ourselves), according to him.

There are several brain regions that might be related to consciousness. The medulla oblongata is very important for wakefulness, and damage to it could cause a range of problems from mild sleeping problems to a coma. The reticular activating system is also important for wakefulness. The thalamus is important as well, integrating all the physical sensory input, and the most important part for the extended consciousness is the cortex.

Pinker, on the other hand, distinguished three components to consciousness: sentience (being aware of surroundings, qualia), access awareness (what we can tell about the surroundings), and self-knowledge.

Sentience

Sentience is closely related to the hard problem of consciousness: how do subjective experiences arise from biological processes? This is also closely related to the concept of qualia, which is the subjective experience of an event. Most scientists adopt the view of Physicalism, which states there is only physical matter in the world. However, it is still very difficult to see how subjective experiences would arise from such physical matter. The problem with qualia is that we cannot really express them as well as we experience them. There is also no way for one person to know the other person is experiencing the same thing. One thought experiment illustrating this problem is the Philosophic Zombie. Could there be a creature that looks exactly like humans do, talk and behave like us, and have no consciousness? We can also consider a zombie in more modern terms: a robot that does the exact same things as a human does.

Another thought experiment illustrating the problem of qualia is Mary the Color Scientist. She was raised in a black-and white room her entire life, and she studied her entire life everything about colors. Later she is released from the room into the colorful world. The question is, how will she react? Will she experience something that was not known to her before? If so, then we have to admit qualia are not merely physical.

The concepts of access awareness and self-knowledge are a lot easier to study.

Access awareness

Access awareness is the ability to report the experiences we go through, both externally and internally. Although most of the times sentience and access awareness have the same content, in some cases they do not. Perception can also occur subliminally, without our awareness, but this information can still be used in our access awareness. At times, unconsciously processes visual information is not accessed by awareness. For example, primed stimuli can influence our opinions and emotions. The case of blindsight patients illustrate the phenomenon where people use information they do not perceive.

Blindsight

Blindsight patients are cortically blind, but they are sometimes capable of responding to visual information without consciously seeing them. They respond above chance the correct choice in detection tasks. They are also able to navigate a route with obstacles, without any problem. When sked why they do not walk in a straight line, they would confabulate and make up a reason.

There are several theories that explain the capability of Blindsight patients to act on unperceived stimuli:

• There is some spared tissue in V1 where blindsight patients could use to compensate for the lost tissue. This explanation is unlikely, because blindsight might happen to patients with large-scale lesions to V1.

• The lateral geniculate nucleus (LGN) in the thalamus might be able to extract location and distance information before that information is projected to V1. This might also be possible, but more complex information can also be extracted, like emotional information, so there must be further information. On the other hand, blindsight hardly occurs with associated damage to the LGN.

Self-Knowledge

Some people describe consciousness as the stream of thought they hear in their mind, telling a story. This is what we would call the interpreter, making sense of our thoughts. On characteristic of this interpreter is that the only information it will use is what it receives, and logical explanations will follow later. In is illustrated in Capgras syndrome, which is a disorder where the patient believes their loved ones are actually someone else trying to impersonate their loved ones. They see the right person, but believe it is someone else. The same goes for Confabulation, as in the case of blindsight patients making up reasons for why they do not walk in a straight line. This shows how the interpreter makes the story fitting the evidence it receives.

Free Will

There is a problem of definition here as well. Many researchers claim there is no real free will, but we still feel like we experience it. However, there is some evidence showing that free will does not exist.

• Newton: everything is made of matter, and all particles are moving in a predicted manner. Therefore, we can predict anything, even brain (which is also made of particles) activity. This is a deterministic view. However, our brain is a chaotic system. There are infinite predictions to be made in this kind of a theory, and even a small measurement error would lead to a completely inaccurate determination. This does not mean there is free will or that our brains are not deterministic, only that it is very difficult to prove it.

• Quantum theory: a probabilistic rather than deterministic view. This view states that in fact the world is not deterministic, so our brain cannot be deterministic either. This view allows the possibility of free will, but not many scientists agree with this view.

• Emergence: there are multiple layers for each organism. Humans are made of atoms and molecules. Does it mean that investigating those explain something about the person itself, their personality and thoughts? This does not make any sense. We are also made of cells (a level higher). Can we draw conclusions about our personality from those cells? How about the activity of the cells? This can go on and on to higher levels of hierarchy. This view is trying to illustrate how each layer has its own law of organization, and that it does not make sense to generalize laws from a lower layer to the next one, although this is exactly what most scientists do.

• Libet: the delay between displaying a stimulus and the person's awareness of the stimulus is about 500ms. However, we are not aware of the delay, because of backward referral made by our brains to compensate for it. In one experiment he conducted, subjects were looking at a rotating clock-hand and were asked to press a button at will and report the location of the clock-hand when they made the decision to press the button. This experiment revealed that their motor cortex was activated 350ms before the reported urge. Therefore, free will according to him could be merely our post-hoc interpretation of our action. However, he claims there is a free won't. There is enough time between our awareness of the intention and the action to be taken, to override it and stop it from happening.

The Law

There are two issues with the law involving consciousness: reliability of eye witnesses and accountability of criminal offenders.

Eyewitness Testimony

Eye witnesses are merely humans. Their interpreter makes sense of information that is given to it, even if it is false, and they might believe their false story with their whole hearts. They might be misled to believe a story that is suggested to them by interrogators, for example. Children are easily misled with their memory. Therefore, eyewitnesses are notoriously unreliable.

Accountability

While someone's actions are indeed caused by neuronal states influenced by environmental and genetic factors, their responsibility cannot be diminished. If a person has a normally functioning brain, the person is being held responsibility for their action. Otherwise, the sentence is usually relieved.

However, it is difficult to evaluate whether a brain is functional or not. Imaging does not provide a clear indication of psychopathology yet. We can identify functions in the brain over a mean of images, but each individual image is different. Imaging the same person over time will also produce different results. In addition, People with brain abnormality caused by a lesion, for example, usually do not commit any crime when there is a police officer present. So can we determine they do not have a free will?

Punishment

There are three types of punishments: retributive justice punishes the offender as revenge. Utilitarian justice provides protection of society from the offender by removing the offender from the society and rehabilitating him/her. Finally, restorative justice is when the offender compensates for his/her crimes with beneficial actions. People tend to prefer utilitarian justice when asked to choose, but when they get the power to do it, they chose retributive justice. This can be explained by the evolutionary point of view. We need to cooperate in order to survive. Cooperating is bad for the individual, but it is better for the group. Dawkins claims that people tend to do what is best for them, which is bad for the group, and the group in turn punishes them for that.

This can be seen in the ultimatum game. Partner A is given 20€ and decides how to split them up between himself and partner B. partner B can then decide whether to accept or reject the offer. If the offer is rejected, nobody gets any money. Rationally, it would be better to accept the offer no matter how bad it is, but people tend to like punishing so much that bad offers are being rejected, just to punish partner A. The dorsolateral prefrontal cortex seems important for this urge to punish, and inhibiting its activity with TMS will result in less punishing, accompanied with the sensation of unfairness.

Source

These lecture notes are based on the subject Cognitive neuroscience from the year 2015-2016