PowerPoint Presentation - Serial Correlation in AFNI

–14–

Outline of SPM and FSL Approaches

• SPM5 and SPM2

H Estimate fixed ARMA(1,1) (more precisely, AR(1)+white noise) model for all “voxels of interest” (pass an OLSQ F-test)

oBy averaging estimated auto-covariance matrix from OLSQ residuals over these voxels

oSPM assumes AR parameter a » 0.2, and approximates ARMA(1,1) correlations via linear Taylor series, to make correlation parameter estimation easier to program

H Use GLSQ (same for each voxel) to solve for b s

oSPM99: Use OLSQ and adjusts DOF downwards to allow for serial correlation

• FSL and FMRIstat (similar, but differ in important details at several points)

HUse OLSQ to get first-pass residuals; use these to estimate each voxel’s auto-correlation matrix; smooth these matrices spatially (FSL & FMRIstat vary here)

HEstimate AR(1) parameter for each voxel separately from smoothed matrices

HUse GLSQ (different for each voxel) to solve for b s

• All these programs use a non-REML method to estimate serial correlation parameter(s) from the OLSQ residual auto-correlation matrix, and then adjust these estimates to reduce the bias thus introduced