Next: Results
Up: Methods
Previous: Temporal
- Spatially stationary (SS) anisotropic spatial AR(1) (i.e.
), compared with a spatial non-stationary (SN)
anisotropic spatial AR
- Different spatial AR priors:
- Noninformative (NI)
- Automatic relevance determination (ARD)
- Spatial MRF with adaptive smoothness
The data used is an audio-visual dataset using echo planar images
(EPI) acquired using a 3 Tesla system with TR=3 seconds, time to
echo (TE) = 30ms, in-plane resolution 4mm and slice thickness 7mm.
The first 4 scans were discarded to leave scans and the
data was motion corrected using MCFLIRT (31) and
high-pass filtered as described in (43). The stimuli
were a reversing checkerboard boxcar stimulus (30 seconds on, 30
seconds off), and an auditory boxcar stimulus (45 seconds on. 45
seconds off). The full signal (HRF) model is used in combination
with each of the noise models just described.
The different models were fitted to one slice of the
data (slice 8 of 20).