Please enable JavaScript.
Coggle requires JavaScript to display documents.
Behavioural and cognitive neuroscience lecture 2 (EEG (Measurement (place…
Behavioural and cognitive neuroscience lecture 2
EEG
Measurement
place electrodes on scalp
measure neural activity
place electrode underneath left eye to pick up eye movements
eye movements cause very strong artefacts
electrode under ear to give neutral reference point
conductive gel applied
need to measure scalp in order to position cap in correct position to allow estimation of relative position of signal
electrodes pick up fluctuations of signal originating from cortical neurons
cheap and easy to conduct
non-invasive
neurophysiology
signals are amplified and then recorded
arrangement of electrodes is always the same
Action potential between dendrites of neurons
when this occurs neurons become a dipole (axon is different charge to cell body)
single cell signal not strong enough, need lots of neurons to be spatially alligned to generate signal
pooled activity comes from pyramid cells
orientation of neurons determines signal and some orientations lead to no signal
get no signal from sulcus only gyrus that is close to the scalp
EEG signal is the sum of signals from neurons measured
Event-related potentials
single trial EEG too noisy for meaningful interpretation
should average across several trials
EVP is average signal that is time locked to event of interest
different aspects of signal =components
variation is averaged out
Advantages
good temporal resolution
get data at time cognitive process is unfolding
disadvantages
spatial resolution bad
hard to tell where in brain signal comes from
inverse problem
if you know the source of the signal, can go back and make a comment about the signal being received by not the other way around
Multivariate pattern analysis
Multi-variate decoding
why use?
signal might contain more info than single ERPS
no ERP component is known for a particular problem
no specific hypothesis about location
might be interested in when info becomes available
generalisations of representations can be represented across time and content
Look at each channel signal across time
look at a particular time window of interest
temporal generalisation
differential bran activity recorded at each time point
classifier trained at each time point
3.classifer tested on its ability to generalise across time points
can see if classifier can retrain
if it does then event is happening for a long time
generally look at just a subsection of electrodes which are chosen based on hypothesis
MPA uses full info
Temporalspatio pattern
n features= n time points x n channels
average of time and spatial configuration of signal
spatial decoding
average across time for each channel
temporal decoding
for a single channel look at info across time
research examples
decoding upcoming decision errors
Participants asked to perform flanker task as fast as possible
sorted through time windows moving towards time of event
roughly 100ms there is activity before mistake indicating error has occured
thought that motor decision made and then realise you are wrong
asked if they thought picture was related to present or future
to see if this relate to decisions
decoding emotion regulation sucess
participants instructed via symbol what emotion regulation strategy to use
want to detect success based on distance between anticipation phase and implementation phase
didn't matter for distraction
did matter during implementation stage
did matter for reappriasal
interesting because you cannot apply reappraisal before you see the stimulus
decoding semantic processing of stimuli
can show people reward stimuli to trigger quicker reward related activity
presented stimuli passively in background to task they were asked to perform