Methods in cognitive neuroscience
Cogs 100, Summer 2025
What questions might we ask about the brain?
• Localization of function: Mapping the brain.
• The time course of mental processes: Timing the brain.
• It is very important to choose the right method for your
question!
Spatial vs. temporal resolution
• Spatial Resolution – the ability to distinguish changes
in a map across different spatial locations.
• Temporal Resolution – the ability to distinguish
changes in activity at a single location over time.
Spatial vs. temporal resolution
Image: [Link]
• Listening to a high note: auditory nerve res up to
1000 Hz.
• If the sound on a movie is off from the visual by
100ms, you will notice.
• Starting a car at a traf c light: 100ms.
• Eye movement to an unexpected stimulus: about
200ms.
Based on Baars & Gage, 2010.
fi
fi
Mapping the brain
Studying the effect of focal brain injury on
behaviour
Focal brain injury and behaviour
Lesion (injured, abnormal or dead tissue) Behaviour
?
(Natural) causes of brain lesions
• Focal
• Cerebrovascular disease (stroke)
• Brain tumour
• Hemorrhage
• Focal head injury.
• Diffuse
• Degenerative disease (e.g., PPA, Alzheimer’s disease.
• Hypoxia, poisoning etc.
Brain lesions
L L L
Note: Your right is this brain’s left.
Non-Focal Stroke A brain tumor –
Haematoma Glioblastoma Focal Stroke
Blood supply in the brain
Key artery to know
for acquired
language disorders:
Middle cerebral
artery
Stroke-related terminology
• Stroke: neurologic injury as a result of disease involving
blood vessels that disrupts blood ow to the brain.
• Ischemia: insuf ciency of blood supply
• Under ischemia tissue is deprived of oxygen and
glucose
• Ischemia prevents the removal of metabolites that can
potentially be toxic.
• When prolonged, neurons die (infarction).
fi
fl
A focal lesion in Broca’s area
Matching clinical classi cations with brain lesion
analysis
• Traditionally, neuropsychological patients’ behavioural
characteristics were studied.
• Their brain lesions were only studied post-mortem.
• Today we can study a patient’s lesion using neuroimaging
tools (e.g., CT scan, structural MRI).
fi
Neuroimaging
CT Scan
CT Scan
• Form of X-Ray
• Shows brain anatomy
• Useful for identifying tumors etc
• Resolution is not great
• Can be improved with dye injected into
MRI Scan
bloodstream
MRI Scan
• Creates image of the anatomy of brain
• Uses a really big magnet
• Detects differences in hydrogen (water)
concentration in different tissues
How do researchers perform studies of people
with focal lesions?
• Early research (and still today): Individual case studies.
• Gives an indication of what is possible, what variation might exist.
• Problems: not really an experiment (no control group).
• Problems with reliability of results (in the statistical sense).
• Group studies classify patients according to their characteristics (in
terms of behaviours and/or lesion site).
• This group can then be compared to control groups.
• E.g., age-matched controls, patients with no aphasia.
• Group results can be compared to chance performance.
• BUT – Groups are not typically homogeneous.
Example: comprehension of actives vs. passives in
Broca’s aphasia.
• Grodzinsky et al. (1999) pooled the results for
comprehension of active and passive sentences from
studies between 1980 and 1996.
• The authors classi ed patients as having Broca’s aphasia
according to lesion site and performance on
standardized test batteries.
fi
Actives vs. Passives
The woman dried the girl. Above chance performance
The girl was dried by the woman. Chance performance
Grodzinsky et al. (1999)
Graph of the # of patients at each
performance level for passives and
actives. The line for passives looks like
the distribution you might expect for a
coin flip.
Grodzinsky et al., 1999
Experimental tasks
• Behavioural tasks such as lexical decision task, sentence-picture judgment
task.
• Dependent variables can be:
• timed (e.g., reaction times in a lexical decision task).
• untimed (e.g., error rates in a sentence-picture judgment task).
• More recently methods such as recording of eye movements have been
used with patients.
Section summary
• Researchers have discovered a great deal by examining
people who have damage to a part of the brain.
• These studies have helped with localization of function.
• Experimental design is important even (or especially!) with
populations with disordered language.
Cytoarchitectonics
Cortical neurons
• The cortex is made up of different types of neurons.
• For example:
• Pyramidal cells
• apical dendrite and lateral dendrites
• Long axon
• Stellate cells
• Interneurons
• Short axon or no axon
Cortical layers
Neocortex can be divided
into 6 layers, from the outside
(next to the pia mater) to the
inside (just before the white
matter).
The layers are distinguishable
by the type of neuron and the
parts of neurons that they
contain.
Cortical layers
The differences in the
make-up of the cortical
layers correspond to
function.
Cytoarchitectonics
• Uses the changes in the structure of the 6 layers to
determine anatomical boundaries.
• The cell layers vary in thickness and density.
• Cells differ depending on their function.
• Anatomical borders -> functional borders
Method
• In order to study the layers, one must:
• Slice the brain post-mortem.
• Stain the slice of the brain.
• Mount the slice on a slide.
• Study the slice under a microscope for changes in lamina (layers).
Brodmann’s cytoarchitectonic maps
• Brodmann (1909) classi ed the brain into over 40
regions.
• Used tissue stains (such as the Nissl stain) to examine the cells at each
level of cortex.
• Brodmann hypothesized that the cytoarchitectonic differences between
regions re ected differences in function.
• He showed that area 17, the primary visual cortex, differed from other
adjacent areas.
fl
fi
Brodmann’s method
A stained sample, printed in Zilles and Amunts
(2010)
[Link]
[Link]
Brodmann’s map
Other cytoarchitectonic maps
Several other
cytoarchitectonic maps
have been published –
with many differences
between them!
Zilles and Amunts (2010)
Section summary
• The study of cell structure and distribution across cortical
layers can detect delineations of brain regions.
• The cell structure of a given part of cortex is related to its
function.
• Hypothesis: If two adjacent brain areas differ in
cytoarchitecture, they are functionally distinct as well.
(functional) Magnetic Resonance Imaging
MRI vs. fMRI
• MRI studies brain anatomy.
• Differentiates different tissues on the basis of their density/water
content.
• fMRI studies brain function.
• fMRI is based on the premise that blood ow in the brain is activity-
dependent.
• Greater activity -> more blood ow.
• More blood ow -> greater concentration of oxyhemoglobin
fl
fl
fl
fMRI
• Equipment
• The basis of the signal + terminology
• Key design concepts
• Analysis
fMRI
• Equipment
• The basis of the signal + terminology
• Key design concepts
• Analysis
Equipment for conducting an fMRI study
4T
magne
t
RF
gradient Coil
coil
(inside)
Magnet Gradient Coil RF Coil
Source: Joe Gati,
photos
fMRI
• Equipment
• The basis of the signal + terminology
• Key design concepts
• Analysis
How is the signal generated?
• Big magnetic eld
• protons (hydrogen molecules) in body become aligned to eld (Step 1: magnetization)
• RF (radio frequency) coil
• radio frequency pulse knocks protons over (Step 2: Excitation)
• as protons realign with eld, they emit energy that coil receives (like an antenna) (Step 3:
Relaxation)
• Gradient coils
• make it possible to encode spatial information of signal
• Signal depends on
• concentration of hydrogen in an area (anatomical MRI)
• amount of oxy- vs. deoxyhemoglobin in an area (functional MRI)
fi
fi
fi
MRI and fMRI signal
• MRI:
• Measures how quickly the protons realign with the main magnetic
eld (T1).
• Differences in T1 correspond to differences in tissue type.
• fMRI:
• Measures how quickly the protons give off energy as they recover to
equilibrium (T2).
• Indirectly affected by neural activity.
fi
Blood Oxygen Level Dependent signal
Blood Flow
↑neural activity ➔ ↑ blood flow ➔ ↑ oxyhemoglobin ➔ ↑ T2* • Blood is
sluggish
compared to
neuronal
ring.
• 2-3 sec to
rise; 4-6 sec
to peak.
Source: fMRIB Brief Introduction to fMRI
fi
A model of the hemodynamic response function
Peak
Rise
Undershoot
fMRI and cognitive activity
Basis of the signal:
Localized Local vaso- Decreased Increased MR
Increased
increase in dilation ratio: signal in a
Cognitive blood flow
neuronal (widening of deoxy-Hb specific brain
task to active
activity local blood oxy-Hb region
region
vessels)
Possible Advantages:
A. Segregates neurological natural
classes by neural dissociations
B. Puts cognitive hypotheses to the
neurological test
C. Helps mapping cognitive
operations onto brain regions
Temporal resolution of fMRI
Determining factors
• Sampling rate (how often you take functional volumes)
usually around 2 sec (this is determined by the
limitation of our equipment).
• Rate of the measured response of the neural tissue
(BOLD)
• BOLD response is sluggish, taking 2-3 seconds to rise above baseline
and 4-6 seconds to peak although neural activity occurs within ms.
• We don’t always know the value of these parameters (time to peak)
precisely.
Timing the brain
• Listening to a high note: auditory nerve res up to
1000 Hz.
• If the sound on a movie is off from the visual by
100ms, you will notice.
• Starting a car at a traf c light: 100ms.
• Eye movement to an unexpected stimulus: about
200ms.
Based on Baars & Gage, 2010.
fi
fi
Electroencephalography
(EEG)
Electroencephalography
• Equipment
• The basis of the signal
• Terminology (imaging in general)
• Key design concepts
• Analysis
Electroencephalography
• Equipment and measurement
• Terminology
• Key design concepts
• Analysis
EEG/ERP – The Equipment
Placement of electrodes
• Electrodes are named (roughly) by cerebral lobe.
• EEG voltages are always recorded with respect to a
reference.
EEG signal
From Steven J. Luck, An Introduction to the Event Related Potential Technique
Event-Related Potentials
What are ERPs?
ERPs are systematic voltage changes in the EEG signal
that reflect the brain’s processing of an event.
+
+
+
-
-
-
Electroencephalography
• Equipment and measurement
• Terminology
• Key design and analysis concepts
• Example
Event-Related Potentials: Key terms
• ERP peaks and components:
• Positive and negative de ections are
peaks.
• Ps for positive peaks, Ns for negative
peaks.
• Number indicates a peak’s position.
• Some people use P300 and others
P3.
• 300 refers to latency of the peak in
ms.
fl
Electroencephalography
• Equipment and measurement
• Terminology
• Key design and analysis concepts
• Example
ERP: Design
• Tasks are designed to minimize movements (e.g., facial
movements, eye movements).
• For language experiments, presentation is typically either
visual or auditory.
• The event of interest in the linguistic input triggers a
marker to be placed in the EEG signal (which is recorded
continuously).
ERP: Design and Analysis
• Because the differences in EEG are small, a large number
of trials are needed (might be 50 trials or more per
condition).
• What is typically presented in articles are grand averages.
• Averages across subjects.
• Each subject’s average re ects many trials.
• Statistical analyses often include location as a factor (left-
right lateralization, anterior-posterior).
• BUT be careful with interpreting localization information.
fl
Electroencephalography
• Equipment
• The basis of the signal
• Terminology
• Key design and analysis concepts
• Example
Sample experiment:
The ‘Oddball’ Paradigm
• Subjects presented (visually) with 80% Xs and 20% Os.
X X XO X X OXX
• The raw EEG was recorded, and marker codes were
sent to mark the signal at the presentation of stimuli.
• ERPs elicited by the Xs and Os were extracted.
• The segment of EEG surrounding the codes.
• Trials were then averaged.
The Oddball Paradigm
[Link]/~mjoao/SPM_EEGMEG.ppt
The Oddball Paradigm: Results
P3 wave following the infrequent Os
Section summary
• Techniques like EEG allow for high temporal resolution
studies of the brain.
• By examining Event-Related Potentials, we can learn
about the timing of the brain’s reaction to stimuli with
different properties.
• The oddball paradigm: when a visual stimulus is
relatively unexpected, a difference in brain reaction arises
on the order of 300ms.
Lecture summary
• In this part of the lecture, we examined a small number of
techniques for studying how the brain gives rise to
behaviour.
• Studying individuals and groups with classes of brain
lesions.
• Cytoarchitectonics.
• (f)MRI
• Electroencephalography/Event-Related Potentials
Take-home messages
• It is important to select the right method for the research
question.
• Is it a question of localization (mapping) or one of timing?
• Examining disorders following a brain lesion may answer
questions about localization of function an/or connections
within the brain. The question of temporal resolution is
irrelevant here (no time course).
• (f)MRI: High spatial resolution but low temporal resolution.
• EEG: High temporal resolution but low spatial resolution.