Lecture 1: Introduction & EEG
Cognitive neuroscience is a combination of cognition and neuroscience. It seeks to create
biologically grounded models of cognitive functions (like visual & auditory perception,
memory etc.) and it wants to learn about these functions by using neuroscience methods
(like EEG/MEG) to study how these cognitive processes happen or what underlies them in
the brain. We use study methods to measure and manipulate the brain.
History of neuroscience:
Franz Joseph Gall: surface of the head depends on mental skills. Different areas on the skull
each represent a different skill.
è Phrenology: the skull can be measured and by measuring all these parts of the skull
could determine people’s cognitive skills but also their personality characteristics.
Modern neuroscience:
Cognitive neuroscience -> modern phrenology? Yes: functional differentiation of the brain.
No: functions are defined thorough experimentation. We do multidisciplinary research. We
don’t just look at the size of the brain areas, but we measure/look at: anatomy & structure,
effects of brain damage, development of areas & networks, neural activity, and chemical
elements in the brain. We can also model/ simulate brain processes.
1. Brain anatomy: our brain is made off many different types of cells, connections, and
neurotransmitters. Brodmann (Brodmann area’s) was the first to map the cortex
based on cell types (43x). More detailed maps followed later. Each neuron type has a
different function. Neurons who function together are often near each other. There
is this clear relation between the structure of the brain and the function.
A neuron has 4 main elements:
1. Dendrites: the input region of the neuron, where it receives information from the
other neurons that it’s connected to.
2. Presynaptic terminals: the output region of the neuron at the end of the axon,
where the neuron can release neurotransmitters into synapse and then
communicate with subsequent neurons.
3. Axon: important for the transfer of the electrical signal within the neuron from the
cell body to the presynaptic terminals
4. Cell body: modulator of the whole neuron.
Neurons look different from normal cells: Axon & dendrites are specialized structures to
transmit and receive information through action potentials. Neurons tend not to reproduce
after birth. What does change is the connections between the neurons. The neurons
themselves don’t change, but the connections do. Each cell type has its own specialized
function.
Other important cells in our brains are the glial cells which kind of support the neurons.
They do regulatory functions that are needed to make the neurons make properly. There are
two types of glial cells (schwann cells & oligodendrocyte) that create the myelin sheath
around the axons of the motor neurons. The myelin that allows faster transmition of the
signal trough the axons.
,2. Damage to anatomy: we can look at people that have damage in certain areas of the
brain and see what the effect is on their cognitive processing, on their behavior and
that can also inform us about how that damaged area may them relate to these
processes. One important cause of damage in larger population of cells are strokes.
There are two types of strokes: hemorrhagic stroke: hemorrhage/ blood leaks to
brain tissue. And ischemic stroke: clot stops blood supply to an area of the brain.
There are also other reasons for damage to the brain like tumors or infections,
trauma, epilepsy & lesions, genetic manifestations, and neural degeneration
(Multiple sclerosis, Alzheimer’s).
3. Development of anatomy: we can look at the evolution of anatomy and the
development with age and how that relates to changes in behavior and cognition.
We can look at brain size, relative size of specific parts of the brains, the relation with
cognitive development and decline.
4. Measuring brain activity: we can stick electrodes into the brain. By putting
electrodes either into specific cells or into the area between cells we can record
when neurons fire, so their action potentials, which can inform us about which cells
where are related or involved in which functions. We can do these for single neurons
by action potentials (electrophysiology) and by local field potentials. Here you are
measuring a group of neurons activity in a local field and then you can see how a
bigger group of neurons are functioning. We can also measure brain activity outside
of the head with EEG/MEG, that allows us to measure the electromagnetic fields
that are produced by groups of neurons in the brain but that we can measure outside
the brain at the scalp level. We can also stimulate/ manipulate brain activity by
TMS/tDCS, manipulating neural activity. it allows us from outside of the head to
modulate activity inside the head and see what happens to how we think and
behave. We can also look at the blood flow in the brain, blood oxygenation by fMRI;
PET; fNIRS. Studying the blood flow tells us about which brain regions are active in a
different way. We can use MRI- scanners to measure DTI, white matter connections,
which tells us about the connections between groups of neurons. We can also
measure indirect reflections of brain activity, muscles glands, & hormones via the
pupil etc.
5. Brain elements: we can also study neurotransmitters in the brain. Hormones can be
measured from the blood. We can also manipulate these measures. We can also
manipulate hormone levels and neurotransmitter levels by injecting them or giving
them to the participants. For example: pharmacology: injection of testosterone. And
then see if it changes your ability for spatial navigation. We can also have participants
eat specific food supplements: choline -> acetylcholine. We know that these can have
an indirect impact on neurotransmitter levels in their brain.
6. Brain computation: making models of the brain functions to improve applications
(Facebook, google) and understanding cognitive processes. We see that these models
when they are trained on visual information that they can learn in a similar way, they
also learn to extract features first and more and more complex combinations of
, information in higher levels of the model -> object classifier. In some cases, these
models can really help us inform but also mimic our human cognitive system.
EEG: a test that measures electrical activity in the brain/ at the scalp. It measures
differences in voltage across the scalp. It reflects post- synaptic potentials (PSP): difference
in voltage along dendrites. Basically, the axon leads to a change in voltage membrane
potential which leads to current flow in the cell but also to opposite current flow outside the
cell and that flow in the space between the dendrites leads to an electrical field that we can
measure outside of the brain. Both inhibitory and excitatory PSP! EEG reflects local field
potential -> no single action potentials but a summation of many neurons.
Both brain areas that are close to each other are more likely to be involved in similar
cognitive processes. Also because of volume conduction, electrodes that are close to each
other will most likely have a shared signal that they are picking up that doesn’t reach the
other side of the brain.
A measurement is good when: there is mass activity: many neurons with the same
alignment, when there is a synchronized activity: no individual action potentials, when it’s
close to the scalp: scalp & skull is not a good conductor -> smears out the signal, and when
there are no noise sources, if it’s a clean measure and it only measures what we want to
measure: electronic devices -> artefacts in data
EEG electrode layouts: 32-64 electrodes= enough for ERPs such as P100, N200, P3.
128 electrodes= enough for localization.
EEG measures voltage potentials
• Measures brain states (frequencies) and temporal characteristics of brain processes
(Event Related Potentials).
• EEG is much more sensitive to neurons in the gyri (cortical peaks) than neurons in the
sulci. Measures mainly activity in gyri.
• Relative cheap, measures more neurons
MEG measures magnetic field
• Similar measure, but
• Better localization (less distortions by skull), measure deeper sources in the brain
better.
• MEG is much more sensitive to neurons in the sulci (cortical valley’s) than neurons in
the gyri. Measures only activity from sulci.
• Expensive
EEG is used when you’re interested in brain states or effects over time (sleep stages) at a
high resolution (P100). In some cases, useful for the localization of neural loci (epilepsy).
The temporal resolution of EEG is great, the spatial resolution is okay. But there’s volume
conduction and the skull blocks information. So ideally, we combine EEG with MEG.
EEG waves. Basal brain states: there are many different types of waves. Each frequency
reflects a different arousal state.
, Brain waves & arousal
• Fast waves (beta/ gamma) = high arousal
• Slow waves (delta/theta/alpha) = low arousal
Theta/ beta ratio
• Lower ratio= more beta, more arousal
• Higher ratio= more theta, less arousal, more sleepiness
Brain responses to stimuli
Event- related potentials -> ERP reflects fine changes in potentials in response to cognitive
events, stimuli. We measure amplitudes & latencies of peaks and throughs and that gives us
information about the amount and timing of the activity.
Brain source localization
Scalp distribution: distribution of a peak (or trough) across the scalp. At the scalp we have a
summation of potentials, but we try to model where this is coming from in the brain.
Inverse problem: there are always multiple models possible to explain the same scalp
distribution. A solution is to combine it with methods who have a better spatial resolution.
Combine with fMRI, combine EEG & MEG.
EEG as a Brain Computer Interface (BCI): we can use the information that we measure to
help people who can’t communicate to communicate for example. Using their brain signals
rather than their actual ability to use their muscles or their mouth. We can also use it as
neuro feedback training.
Lecture 2: MRI & visual perception -> look at pp
Functional MRI is the main method we use to locate and characterize the responses of the
human brain. Both structural and functional MRI are important ways to understand what
changes in the brain during disease might underlie the symptoms of the disease.
MRI relies on interactions between a magnetic field and tissues of the body. It measures
(magnetic) properties of tissues. The tissues of the body, the meat, may not seem to have
obvious magnetic properties: they are not attracted to magnets. But they react differently to
magnetic fields depending on their structure, for example the amount of water they contain,
and this differs between different tissues within the body. MRI being an imaging method
records from many points in space, allowing reconstruction of an image inside of the brain.
fMRI also records at many points in time -> examines how tissue properties change over
time.
Structural/anatomical MRI images the type tissue depending on magnetic properties. F.e.,
gray and white matter. This allows us to analyze how these tissues change in health and
disease. F.e., gray matter volume increases during development and decreases with ageing
Many neural degeneration diseases and strokes are associated with changes to gray or white
matter. MRI is also common in surgical planning. F.e., to localize structures for operations,
like cancers and the important organs nearby. -> images tissue types.
Cognitive neuroscience is a combination of cognition and neuroscience. It seeks to create
biologically grounded models of cognitive functions (like visual & auditory perception,
memory etc.) and it wants to learn about these functions by using neuroscience methods
(like EEG/MEG) to study how these cognitive processes happen or what underlies them in
the brain. We use study methods to measure and manipulate the brain.
History of neuroscience:
Franz Joseph Gall: surface of the head depends on mental skills. Different areas on the skull
each represent a different skill.
è Phrenology: the skull can be measured and by measuring all these parts of the skull
could determine people’s cognitive skills but also their personality characteristics.
Modern neuroscience:
Cognitive neuroscience -> modern phrenology? Yes: functional differentiation of the brain.
No: functions are defined thorough experimentation. We do multidisciplinary research. We
don’t just look at the size of the brain areas, but we measure/look at: anatomy & structure,
effects of brain damage, development of areas & networks, neural activity, and chemical
elements in the brain. We can also model/ simulate brain processes.
1. Brain anatomy: our brain is made off many different types of cells, connections, and
neurotransmitters. Brodmann (Brodmann area’s) was the first to map the cortex
based on cell types (43x). More detailed maps followed later. Each neuron type has a
different function. Neurons who function together are often near each other. There
is this clear relation between the structure of the brain and the function.
A neuron has 4 main elements:
1. Dendrites: the input region of the neuron, where it receives information from the
other neurons that it’s connected to.
2. Presynaptic terminals: the output region of the neuron at the end of the axon,
where the neuron can release neurotransmitters into synapse and then
communicate with subsequent neurons.
3. Axon: important for the transfer of the electrical signal within the neuron from the
cell body to the presynaptic terminals
4. Cell body: modulator of the whole neuron.
Neurons look different from normal cells: Axon & dendrites are specialized structures to
transmit and receive information through action potentials. Neurons tend not to reproduce
after birth. What does change is the connections between the neurons. The neurons
themselves don’t change, but the connections do. Each cell type has its own specialized
function.
Other important cells in our brains are the glial cells which kind of support the neurons.
They do regulatory functions that are needed to make the neurons make properly. There are
two types of glial cells (schwann cells & oligodendrocyte) that create the myelin sheath
around the axons of the motor neurons. The myelin that allows faster transmition of the
signal trough the axons.
,2. Damage to anatomy: we can look at people that have damage in certain areas of the
brain and see what the effect is on their cognitive processing, on their behavior and
that can also inform us about how that damaged area may them relate to these
processes. One important cause of damage in larger population of cells are strokes.
There are two types of strokes: hemorrhagic stroke: hemorrhage/ blood leaks to
brain tissue. And ischemic stroke: clot stops blood supply to an area of the brain.
There are also other reasons for damage to the brain like tumors or infections,
trauma, epilepsy & lesions, genetic manifestations, and neural degeneration
(Multiple sclerosis, Alzheimer’s).
3. Development of anatomy: we can look at the evolution of anatomy and the
development with age and how that relates to changes in behavior and cognition.
We can look at brain size, relative size of specific parts of the brains, the relation with
cognitive development and decline.
4. Measuring brain activity: we can stick electrodes into the brain. By putting
electrodes either into specific cells or into the area between cells we can record
when neurons fire, so their action potentials, which can inform us about which cells
where are related or involved in which functions. We can do these for single neurons
by action potentials (electrophysiology) and by local field potentials. Here you are
measuring a group of neurons activity in a local field and then you can see how a
bigger group of neurons are functioning. We can also measure brain activity outside
of the head with EEG/MEG, that allows us to measure the electromagnetic fields
that are produced by groups of neurons in the brain but that we can measure outside
the brain at the scalp level. We can also stimulate/ manipulate brain activity by
TMS/tDCS, manipulating neural activity. it allows us from outside of the head to
modulate activity inside the head and see what happens to how we think and
behave. We can also look at the blood flow in the brain, blood oxygenation by fMRI;
PET; fNIRS. Studying the blood flow tells us about which brain regions are active in a
different way. We can use MRI- scanners to measure DTI, white matter connections,
which tells us about the connections between groups of neurons. We can also
measure indirect reflections of brain activity, muscles glands, & hormones via the
pupil etc.
5. Brain elements: we can also study neurotransmitters in the brain. Hormones can be
measured from the blood. We can also manipulate these measures. We can also
manipulate hormone levels and neurotransmitter levels by injecting them or giving
them to the participants. For example: pharmacology: injection of testosterone. And
then see if it changes your ability for spatial navigation. We can also have participants
eat specific food supplements: choline -> acetylcholine. We know that these can have
an indirect impact on neurotransmitter levels in their brain.
6. Brain computation: making models of the brain functions to improve applications
(Facebook, google) and understanding cognitive processes. We see that these models
when they are trained on visual information that they can learn in a similar way, they
also learn to extract features first and more and more complex combinations of
, information in higher levels of the model -> object classifier. In some cases, these
models can really help us inform but also mimic our human cognitive system.
EEG: a test that measures electrical activity in the brain/ at the scalp. It measures
differences in voltage across the scalp. It reflects post- synaptic potentials (PSP): difference
in voltage along dendrites. Basically, the axon leads to a change in voltage membrane
potential which leads to current flow in the cell but also to opposite current flow outside the
cell and that flow in the space between the dendrites leads to an electrical field that we can
measure outside of the brain. Both inhibitory and excitatory PSP! EEG reflects local field
potential -> no single action potentials but a summation of many neurons.
Both brain areas that are close to each other are more likely to be involved in similar
cognitive processes. Also because of volume conduction, electrodes that are close to each
other will most likely have a shared signal that they are picking up that doesn’t reach the
other side of the brain.
A measurement is good when: there is mass activity: many neurons with the same
alignment, when there is a synchronized activity: no individual action potentials, when it’s
close to the scalp: scalp & skull is not a good conductor -> smears out the signal, and when
there are no noise sources, if it’s a clean measure and it only measures what we want to
measure: electronic devices -> artefacts in data
EEG electrode layouts: 32-64 electrodes= enough for ERPs such as P100, N200, P3.
128 electrodes= enough for localization.
EEG measures voltage potentials
• Measures brain states (frequencies) and temporal characteristics of brain processes
(Event Related Potentials).
• EEG is much more sensitive to neurons in the gyri (cortical peaks) than neurons in the
sulci. Measures mainly activity in gyri.
• Relative cheap, measures more neurons
MEG measures magnetic field
• Similar measure, but
• Better localization (less distortions by skull), measure deeper sources in the brain
better.
• MEG is much more sensitive to neurons in the sulci (cortical valley’s) than neurons in
the gyri. Measures only activity from sulci.
• Expensive
EEG is used when you’re interested in brain states or effects over time (sleep stages) at a
high resolution (P100). In some cases, useful for the localization of neural loci (epilepsy).
The temporal resolution of EEG is great, the spatial resolution is okay. But there’s volume
conduction and the skull blocks information. So ideally, we combine EEG with MEG.
EEG waves. Basal brain states: there are many different types of waves. Each frequency
reflects a different arousal state.
, Brain waves & arousal
• Fast waves (beta/ gamma) = high arousal
• Slow waves (delta/theta/alpha) = low arousal
Theta/ beta ratio
• Lower ratio= more beta, more arousal
• Higher ratio= more theta, less arousal, more sleepiness
Brain responses to stimuli
Event- related potentials -> ERP reflects fine changes in potentials in response to cognitive
events, stimuli. We measure amplitudes & latencies of peaks and throughs and that gives us
information about the amount and timing of the activity.
Brain source localization
Scalp distribution: distribution of a peak (or trough) across the scalp. At the scalp we have a
summation of potentials, but we try to model where this is coming from in the brain.
Inverse problem: there are always multiple models possible to explain the same scalp
distribution. A solution is to combine it with methods who have a better spatial resolution.
Combine with fMRI, combine EEG & MEG.
EEG as a Brain Computer Interface (BCI): we can use the information that we measure to
help people who can’t communicate to communicate for example. Using their brain signals
rather than their actual ability to use their muscles or their mouth. We can also use it as
neuro feedback training.
Lecture 2: MRI & visual perception -> look at pp
Functional MRI is the main method we use to locate and characterize the responses of the
human brain. Both structural and functional MRI are important ways to understand what
changes in the brain during disease might underlie the symptoms of the disease.
MRI relies on interactions between a magnetic field and tissues of the body. It measures
(magnetic) properties of tissues. The tissues of the body, the meat, may not seem to have
obvious magnetic properties: they are not attracted to magnets. But they react differently to
magnetic fields depending on their structure, for example the amount of water they contain,
and this differs between different tissues within the body. MRI being an imaging method
records from many points in space, allowing reconstruction of an image inside of the brain.
fMRI also records at many points in time -> examines how tissue properties change over
time.
Structural/anatomical MRI images the type tissue depending on magnetic properties. F.e.,
gray and white matter. This allows us to analyze how these tissues change in health and
disease. F.e., gray matter volume increases during development and decreases with ageing
Many neural degeneration diseases and strokes are associated with changes to gray or white
matter. MRI is also common in surgical planning. F.e., to localize structures for operations,
like cancers and the important organs nearby. -> images tissue types.
