TASK 3 - EEG Flashcards
EEG
= measure electrical signals outside of the brain
- signal of electrode is the difference between reference electrode and the active/measuring electrode
- -> electrodes (= small electrodes) are either glued to skin or included in a stretchable cap
- recording from each electrode reflects the electrical activity of the underlying brain region
- yields a continuous recording of the overall brain activity
neural source
- EEG not sensitive to action potentials (too fast, too small)
- EEG sensitive to post-synaptic potentials: slower, (= dipoles)
- for large groups of neurones to be measured the neurones need to have a) same input (excitatory or inhibitory) and b) same orientation (parallel)
- -> synchronised activity: activity of neurones add up
- -> de-synchronised activity: the activity of neurones cancel each other out (differing input or perpendicular)
a) electrodes detect the sum of positive and negative charges in the area
b) pyramidal cells (organised along cortical columns) are arranged in parallel + perpendicular to the cortex (in the gyri)
neural source
- dipole
= one end of neurone more positive and one end more negative
a) source = region of positive charge
b) sink = region of negative charge
- produce positive and negative deflection at different regions of the scalp
1. radial dipole: oriented perpendicular to the surface; directly measured in gyri
2. tangential dipole: oriented parallel to the scalp; sulci
within brain
- volume conduction
= ions of the same charge repel another –> repel other ions of the same charge –> wave of charge that travel
- -> principal: opposite attract and similar repel
- responsible for the propagation of the EEG signal within the brain
- signal from a large dipole may travel much further than signal from a small dipole (process continues further)
from brain to electrode
- capacitive conduction
= electrical signal must travel from the brain through the dura layers –> skull layers –> scalp –> to the electrode
- volume conduction can only happen within the brain
- -> relates to low spatial resolution: cannot differentiate where exactly the signal originated
- one electrode collects data from several locations –> can only infer about the locations (different reactions inside can lead to same information outside)
- spatial smearing = unclear borders of the locations
- -> inverse problem
capacitative conduction
- capacitor
= 2 pools of charges separated by an insulating layer (without layer charges would freely mix = neutral pool)
- within capacitors there are ions: volume conduction within capacitors –> goes until the insulating layer
- charge difference builds up across insulating layer (neg. ions push against one side, pos. ions accumulate on the other)
- how much charge builds up depends on properties of layer, size of charged pool, distance between pool of charge and layer
measurement of EEG
- design
head –> electrode –> amplifier + filter –> analog-digital (AD) converter –> computer –> disk + monitor
- presented on a scale with time on X and microvolts (amplitude) on Y
measurement of EEG
- output
- yields a continuous recording of the overall brain activity
- -> linear superposition: measured voltage is the weighted sum of electrical activity of multiple regions in brain (largest effect by regions near electrode)
- weights depend on distance from electrode, conductivity, source orientation
measurement of EEG
- topography
= colour plot = colour indicates the voltage that was measured at some location on the scalp across time
- -> get picture of distribution voltage across scalp
- map changes rapidly: every single image presents single moment in time)
- sensitive to dipole location, orientation and strength
noise (artifacts)
= artifacts = signals recorded in EEG that are not produced by the brain
a) physiological artifacts: ocular, muscle, cardiac activity (produce electricity
b) technical artifacts: cable movement, incorrect reference placement, body movements
- measured by signal-to-noise ratio (SNR) (signal/noise)
- pre-stimulus baseline period: needed to compare the amount of noise in both groups
–> if different, the observed effect post-stimulus may not be valid
TO AVOID NOISE
- signal must be large enough to be measurable (orientation not a problem)
- need electrode gel: for ions to pass through skull
- remove very slow + very fast voltage changes, as they are likely to be due to noise
- amplifier is needed: if amplified before introduction of external noise, this noise will have a smaller impact on the measurement (larger SNR)
inverse problem (localisation)
= inverse problem = only know final voltage pattern and must work back to determine which sources within brain produced this pattern
- cannot really determine location by looking at scalp distribution
a) single dipole: use scalp distribution to estimate location + orientation (unless deep in the brain or data is noisy)
b) multiple dipoles active: they sum together –> precision is lost but still works when the dipoles are far apart (capacitive conduction) - difficult to localise two dipoles of similar orientation and location
- -> the more dipoles are added, the more difficult it is
- need converging evidence
electrode placement
- 10-20 system = standardisation of locations
1. measure total distance between landmark points on the head (front-back and left-right dimension)
2. total distance is divided in number of segments of 10% and 20% of total - each electrode is then referred to by there neighbouring brain regions (F = frontal), whether it is left or right from midline (left (odd), right (even)) and distance from midline (close (1), far (7-8))
- needed to replicate EEG results
electrode placement
- reference electrode
- any output voltage is defined as coltage-difference between two locations between head-electrode and reference electrode
- should be placed close to the head but be silent
- -> as little brain activity as possible
- mastoid bone often used: skull is really thick right behind the ear
- right placement needed to reduce noise and correct recordings
state effects
the more excited the brain, the smaller and the faster the waves
oscillations
= recurring waveforms in the data pattern; have 3 features
1) frequency: amount of a full waveform in some period (1s)
- measured in Hertz: 1 Hz = 1 full waveform in 1s
2) amplitude: amount of volt from 0 to peak; describes the height of the peak
- measured in microvolts
3) phase: shift of the waveform on the time axis (where in the waveform the signal begins)
- measured in radians or degrees
