nullnullMEG Localisation in SPM
Rik Henson
(MRC CBU, Cambridge)
Karl Friston (UCL, London)
& other SPM authorsOverviewOverview
Standard epoching, thresholding, filtering…
Random Field Theory for Space-Time images
Distributed Source Localisation
OverviewOverview
Standard epoching, thresholding, filtering…
Random Field Theory for Space-Time images
Distributed Source Localisation
nullOverviewOverview
Standard epoching, thresholding, filtering…
Random Field Theory for Space-Time images
Distributed Source Localisation
nullSearching in Space-TimeSearching in Space-TimeRandom Field Theory is a method for correcting for multiple statistical comparisons with N-dimensional spaces (for parametric statistics, eg Z-, T-, F- statistics)
[ Where is there an effect in time, eg GFP (1D)? ]
Where is there an effect in time-frequency space (2D)?
Where is there an effect in time-sensor space (3D)?
[ Where is there an effect in time-source space (4D)? ]Worsley Et Al (1996). Human Brain Mapping, 4:58-73Where is an effect in time-frequency space (2D)?Where is an effect in time-frequency space (2D)?Kilner Et Al (2005) Neuroscience Letters 374, 174–178Where is an effect in time-sensor space (3D)?Where is an effect in time-sensor space (3D)?Reliable effects over EEG channels across subjects
(NB: MEG data would require some sensor-level realignment, or source localisation…)Henson Et Al (2008) NeuroimageSearching in Space-TimeSearching in Space-TimeExtended to 2D cortical mesh surface
RFT generally requires Gaussian smoothing, but exerts exact FWE control for sufficient smoothing
Nonparametric (permutation) methods of FWE control make fewer distributional assumptions (do not require smoothing), but do require exchangeabilityPantazis Et Al (2005) NeuroImage, 25:383-394OverviewOverview
Standard epoching, thresholding, filtering…
Random Field Theory for Space-Time images
Distributed Source Localisation…
nullDistributed Source LocalisationDistributed Source Localisation Weighted Minimum Norm & Bayesian equivalent
ReML estimation of hyperparameters (regularisation)
Model evidence and comparison
Spatiotemporal factorisation and power (e.g, induced)
Automatic Relevance Detection (hyperpriors)
Multiple Sparse Priors
A canonical cortical mesh (warped MNI template)
fMRI priorsWeighted Minimum Norm, RegularisationLinear system to be inverted: Since n