Cognitive neuroscience lectures
Lecture 1 12-11-2020
Cognition is a very broad concept, it encompasses several different functions the brain allows us to
use in the world (visual perception, attention, memory etc.)
Neuroscience is measuring the biological/biochemical processes that occur in the brain that are
responsible for the cognitive processes (EEG, fMRI, physiology, gene expression etc.) to study
cognition.
History of neuroscience
Phrenology Franz Joseph Gall, different areas on the skull each represent a different skill
(enlargements or indentations of the skull). This was not scientific pseudoscience.
Modern neuroscience:
Cognitive neuroscience modern phrenology? In a sense, it maps a brain. It maps locations to
certain functions to certain areas. The difference is that they are defend by thorough
experimentation, multidisciplinary research. Not just the size of brain areas, also anatomy, structure,
effects of damage, development, activity, chemical elements, and simulate the brain.
Brain anatomy:
There are many different types of cells, connections, and neurotransmitters (chemical elements that
allow action potentials to transfer between synapses).
Brodmann was the first to map the cortex based on cell types (43x). More detailed maps followed
later. Each neuron type (so each BA) has a different function.
Structure says something about the function of the cell
Modulator cell body, affects for example how likely it is there is an action potential built up.
Neurons are different from normal cells because of the axons and dendrites. These transmit and
receive information through action potentials. The cells tend not to reproduce after birth. The
structure of the synapse connections does change.
Motor neuron may dendrites, quite long axon, myeline sheets to protect the information
Cognition (forebrain) pyramidal neurons
Coordination (cerebellum) purkinje cells, many dendrites that can collect a lot of information
Vision (retina) bipolar neurons, more simple and don’t have that many dendrites. They process
things progressively, they change and alter information in a simple way
Sensation (spinal cord) unipolar neuron, the axons goes straight from the dendrite to the
presenaptic terminals, the cell body is on top of it (does not modulate the AP directly)
Inhibitory cell dendrites spread out across the entire cell body.
Neural circuits enable complex computations, because there are so many different possibilities of
combinations.
Neural circuits are flexible and adapt (it allows us to learn). These changes can be long-term of short-
term. They all lead to more or less AP through biochemical changes at the synapses, receptors of
cellular membranes, or structural plasticity in synapses.
,In sum, there are many different cell types across and within brain areas, different computations,
different functions, and prone to change in several ways.
Methods used in cognitive neuroscience. These can be divided in categories:
- Neuroscientific manipulations:
o Accidental (tumor) or active (TMS)
Measure: performance on cognitive tasks
- Neuroscientific measures:
o Invasive (e.g. electrophysiology) or noninvasive (e.g. EEG)
Accidental manipulations are by definition invasive
Manipulation: multiple conditions in a cognitive task or group comparison
Damage to anatomy (clinical neuropsychology)
Strokes give valuable insights.
- Hemorrhagic stroke blood leaks into brain tissue
- Ischemic stroke clot stops blood supply to an area of the brain, resulting in cell death
Tumors or infections (or insects)
Neural degeneration (multiple sclerosis, Alzheimer’s)
Trauma (football)
Epilepsy & lesions
Genetic manifestations
these are all naturally occurring
Actively manipulating brain activity:
- Manipulating neural activity (perturbations)
o TMS sending out a strong magnetic pulse (targets large regions)
o tDCS put a direct current through the brain. This method is less spatially accurate,
but it can influence the activity of large populations of neurons
Measuring brain activity:
- electrophysiology, invasive old method, is invasive. You get an electrode and put it in the
brain. When you measure potentials, you know you hit the membrane of the cell. This way
you can measure the function of individual cells, it has a very high spatial resolution.
o action potentials (electrophysiology)
o local field potentials (electrophysiology)
- Electromagnetic fields at scalp (EEG/MEG). It allows to measure activity, noninvasive, from a
large group of neurons that is strong enough to pass the skull. This version can measure what
state someone is in. A more invasive version can or example, you can measure where a brain
tumor is or where an epileptic attack starts (too much activity).
- Blood oxygenation (fMRI; PET)
- Brain elements
o the measurement of neurotransmitters and hormones (measure it in blood or
manipulate by providing drugs, e.g. injection of testosterone)
o Food: choline acetylcholine. You can make sure that a person has a large amount
of a basic element the body needs to build nt’s
- Brain computation: more of a comparison method, not a manipulation or a measurement
method. making models of the brain (building neural networks) to improve applications
(training to recognize object in an image). Then they compare humans to models.
In sum, defining steps/networks in information processes by using neuroscientific methods. These
allow you to measure and manipulate the brain. Overview of methods:
,The least invasive are the least spatially accurate methods.
Thinking and the mind has a clear biological mechanism underlying it. The brain is not seen as a black
box liked it used to be seen.
EEG (electro-encephalography) continuous recording of brain activity. Electrodes are placed on
the scalp. The polarity change inside the neuron is recorded. Gel allows electrical signals to pass. An
amplifier strengthens the signal and prevents it from being distorted.
The change in voltage at each electrode is measured over time. From this you can extract a lot of
information.
EEG measures differences in voltage across the scalp. It basically reflects post-synaptic potentials
(PSP) that are created by differences in voltage at the level of the neuron. When a neuron (often
pyramidal cell) has an action potential, they create a different in voltage (positive at end, negative
closer to cell body). This differences also causes a kinetic field. Both inhibitory and excitatory PSP. It is
unclear which one you measure.
It reflects local field potential no single action potentials but a summation of many neurons.
Use EEG when measuring:
- Mass activity many neurons with the same alignment
- Synchronized activity not individual action potentials
- Close to scalp scalp & skull is not a good conductor smears out the signal
- No noise sources electronic devices artefacts in the data
EEG electrode layouts most cases use a 64 or 128 channel cap. You also have 32 and 256 channels.
Choosing the cap depends on the goals of the experiment.
32-64 enough for P100, N200, P3 etc. (when)
128 enough for localization (where)
Advantages EEG (ERP) vs MEG
EEG voltage potentials
- Measures brain states (frequencies) and temporal characteristics of brain processes (ERPs)
- Relatively cheap, measures more neurons
MEG (magnetoencephalography) measures changes in magnetic fields
- Similar measure, but better localization (less distortions by skull)
- Expensive, more spatially accurate
Know for exam: MEG has a little better spatial resolution than EEG and ERP.
Use EEG when you are interested in effects over time (sleep stages) at a high resolution (e.g., p100).
In some cases useful for the localization of neural loci ( e.g. epilepsy intracranial).
, Summary (dis)advantages:
- Great temporal resolution
- Spatial resolution is okay
o But the skull blocks information
o Combine EEG & MEG (or even with fMRI)
EEG measurements (can decide later on what type of analysis you are going to do)
- Brain states
o For example: close your eyes (no visual input).
o You see a change in amplitude and frequency when the eyes are opened.
o There are multiple states (which you can derive from different types of waves
Excited = actively processing information, planning something etc. lower amplitude
means that the voltage changes are smaller. Even further, you have Gamma activity
(extremely focused, frequencies even higher)
Sleep spindles are at relatively high frequency.
These are the different arousal stages, from delta to gamma
Slow waves (e.g., theta) low arousal
Fast waves (e.g., beta) high arousal
Gamma (32 Hz>) superlearning
Beta (16-31 Hz) processing information, analytical thinking
Alpha (8-12 Hz) eyes closed or very relaxed
Theta (4-8 Hz) sleep, REM (sawtooth patterns), dreaming, deep meditation
Delta (<4 Hz) deep dreamless sleep
IMPORTANT: always a mixture of different frequencies (multitude of waves). Only
alpha is very clear, but not always.
