¥Regression Model (General
Linear System)
_Simple Regression Model (one regressor): Y(t) = a0+a1t+b r(t)+e(t)
"Run 3dDeconvolve with regressor r(t), a time series
IRF
_Deconvolution and Regression Model (one stimulus with a
lag of p TRÕs):
"
Y(t) = a0+a1t+b0f(t)+b1f(t-TR)É+bpf(t-p*TR)+e(t)
"Run 3dDeconvolve with stimulus files
(containing 0Õs and 1Õs)
¥Model in Matrix Format: Y = Xb + e
_X: design
matrix - more rows (TRÕs) than columns (baseline parameters + beta weights).
_
a0 a1 b
a0 a1 b0 É bp
"
------------------
------------------------
q1
1 r(0)
1 p fp É f0
q1 2
r(1)
1 p+1 fp+1 É f1
q
. . .
. .
.
q1 N-1 r(N-1)
1 N-1 fN-1 É fN-p-1
_e: random
(system) error N(0, s2)