Resting state EEG Flashcards
What is a dipole?
A small element of activity in the brain
What does dipole fitting solve?
The inverse problem - where the signal is coming from
How does dipole fitting work?
You estimate where you think the signal from your measured signal is coming from.
Often start with one dipole and the computer programme will generate the signal this dipole would create and see if this matched our measured signal
Adjust the position until we get the best fit
What happens if we can’t find the ‘best fit’ from our dipole?
Increase the number of dipoles by 1
What is the advantage of dipole fitting?
- Simple to calculate; modern computers do it instantly
- Can potentially model several simultaneous sources at once
- works well with simple sensory experiments
What are the disadvantages of dipole fitting?
- Not good for more complicated experiments; more susceptible to noise
- Need to have a good guess at where the activity is coming from
- Need to know no. of dipole sources in advance
What is a Distribution Current Model?
Try to work out a continuous distribution of current within the brain or a region which might be generating response
Assumes the weakest (lowest current) is the correct one
What paradigm is Garrido et al use that involved a distribution current model? What psychiatric disorder gets affected by this?
Mis-match negativity = play a continuous tone but then throw in a ‘oddball tone’ that evokes a stronger response in the frontal and temporal lobe
Schziophrenia
How does EEG fair compared to MEG in a distribution current model?
More distributed source ie not as good spatial res
But MEG might not be as good at finding the frontal sources
What are the advantages to Current Density modelling?
Don’t have to guess how many sources or where
Can be used in complex experiments
What are the disadvantages to current density modelling?
Computationally more demanding
Produces diffuse, weakly active sources that tend to be from surface of brain - hard to localise deep sources
How can we solve the problem of localising deep sources?
fMRI/MRI
How did dale et al (2000) use both fMRI and MEG to improve spatial localisation?
Had ptps do two language tasks in the MEG - hearing and repeating novel words and repeated words.
Use the Current Density model but source was widely spread
Did the same task in fMRI and combined with MEG on the computer to make a more precise and accurate model
What does beamforming localise compared to the other two methods?
Localises oscillations and detects increases and decreases in power in frequency bands such as alpha or theta
What’s the name for power increases or power decreases in the oscillatory bands?
Increases = Event-Related Syncronisation
Decreases = Event-Related Desyncronisation