Ð18Ð
¥ Assume a fixed shape h(t ) for the HRF
H e.g., h(t ) = t 8.6 exp(-t/0.547) [MS Cohen, 1997]
H Convolved with stimulus timing (e.g., AFNI program waver), get ideal response function r(t )
¥ Assume a form for the baseline
H e.g., a + b×t  for a constant plus a linear trend
¥ In each voxel, fit data Z(t ) to a curve of the form
¥         Z(t ) È a + b×t + b× r(t )
¥ a, b, b are unknown parameters to be calculated in each voxel
¥ a,b are ÒnuisanceÓ parameters
¥ b is amplitude of r(t ) in data = Òhow muchÓ BOLD
Simple Regression Models