EEG - Principles, ERPs, Time-Domain

Question 1

Who recorded the first EEG?
What did they discover?

Answer 1

• Hans Berger in 1929
• discovered a low-frequency rhythm (alpha rhythm, 8-12 Hz) while performing mental calculations

Question 2

What does EEG actually measure?
What is the general assumption behind it?

Answer 2

• electric potential: electric activity from one electrode relative to reference electrode
• assumption: EEG measures the summed up activity of thousands of parallel oriented, neighboring and synchronously active pyramidal cells

Question 3

What does the EEG signal represent?

Answer 3

• modulations in extracellular currents due to postsynaptic potentials of pyramidal cells
• PSPs in apical dendrites create a dipole towards the cell soma (negativity at the scalp)

Question 4

What are pyramidal neurons?

Answer 4

• located in cortex layers 5 and 6
• arranged in parallel
• oriented perpendicular to the cortex surface
• responsible for most computations in cortex

Question 5

Do all neurons contribute equally to the EEG signal?

Answer 5

No. The EEG signal is more sensitive to
- perpendicularly oriented neurons (gyri > sulci)
- neurons closer to the scalp (predominantly cortical/pyramidal neurons; deep structures (base of the cortical gyrus, mesial walls of the major lobes, hippocampus, thalamus, and brain stem) do not directly contribute to the signal)

Question 6

Which factors influence the polarity and amplitude of the EEG signal?

Answer 6

location of synaptic activity
- positive charge when excitatory input near soma
- negative charge when excitatory input at apical dendrites
neural synchrony
- higher synchrony in neuron firing leads to higher amplitude
- epileptic seizure is characterized by very high synchrony across the entire brain

Question 7

What are pros and cons of EEG?

Answer 7

pros:
- high temporal resolution (ms range)
- non-invasive
- relatively small and cheap
- lightweight, portable

cons:
- low spatial resolution (20 mm)
- lengthy setup

Question 8

How are electrodes mounted, placed and named?

Answer 8

• mounted in electrode cap following 10-20 system (percentages)
• placement independent of head size and comparable across labs
• naming after brain lobes (frontal, parietal, temporal, occipital) and central sulcus

Question 9

Which artifacts (non-brain signals) in the raw EEG are visible to the naked eye?

Answer 9

• eye movements artifacts (e.g. saccades, blinks)
• muscle artifacts (e.g. clenching of teeth, talking = high-frequency)
• heartbeat
• slow drifts (caused e.g. by sweating)
• blocking

Question 10

What is the only brain signal being visible to the naked eye in raw EEG?

Answer 10

alpha rhythm of the visual cortex when eyes are closed

Question 11

Which 3 measures are taken in EEG preprocessing?

Answer 11

• subtracting systematic artefacts
• filtering frequencies (high-pass (0.01 Hz), low-pass (40 Hz) and band-pass filters)
• rejection of data segments with excessive noise (e.g. movement artifacts)

Question 12

What are the 2 signals of interest in EEG?

Answer 12

spontaneous oscillations
- occur in specific frequency bands (“brain rhythms”)
- modulated by e.g. sleep-wake-cycle, mental and physical activity
event-related potentials (ERPs)
- short in duration (100-700ms)
- time-locked to an internal/external event
- modulated by many internal and external factors

Question 13

How is the signal-to-noise ratio in EEG and what follows from it?

Answer 13

• bad
• a high number of trials is needed in order to reduce noise by averaging across trials
• amplitude of the noise goes down by a factor of sqrt(K) in an average across K trials

Question 14

ERPs

What do Pe, Ne, ERN and LRP stand for?

Answer 14

• Pe: positive deflection related to erroneous decisions
• Ne: negative deflection related to erroneous decisions
• ERN: error-related negativity (other notation for Ne)
• LRP: lateralized readiness potential reflecting movement preparation

Question 15

What are the components of an ERP and what do they relate to?

Answer 15

early components (exogenous)
- mainly automatic sensory responses
- highly influenced by physial stimulus properties
- clinical use (e.g. test integrity of sensory pathways)
- can be modulated by emotion, mood, attention, reward etc.
later components (endogenous)
- reflect internal, higher-order processing
- not strongly dependent on physical stimulus properties
- influenced by task, strategy, emotional processing, etc.

Question 16

Example of early component: P1

Answer 16

• peaks at around 100ms after onset of visual stimulus
• reflects perceptual processing
• generated in visual cortex
• attention allocation leads to amplification of sensory input

Question 17

Example of late component: P3/P300

Answer 17

• typically observed in oddball paradigm
• infrequent target stimuli are randomly presented in a background of frequent standard stimuli
• subjects are instructed to respond mentally or physically to the target, and not respond otherwise

Question 18

The spectral power distribution in EEG typically follows which law?

Answer 18

• 1/f law
• the higher the frequency, the lower its proportion (or amplitude/synchrony?) in the signal
• however, certain brain rhythms exist which are characterized by slightly higher power in particular frequency bands

Question 19

What are the different brain rhythms?
What are they associated with?

Answer 19

• delta (0.5-4 Hz): infants, sleeping adults, abnormality in waking adults, deep sleep
• theta (4-8 Hz): children, sleeping adults, awake & drowsiness, light sleep stages
• alpha (8-13 Hz): occipitally, awake, eyes closed, mental inactivity, physical relaxation
• beta (13-30 Hz): frontally and parietally, sharp spike-waves, light sleep stages
• gamma (31-100 Hz): binding of consciousness, unity of perception