Here I have used 2 datasets: (1) visually presented houses vs faces, (2) retinotopic mapping task.
The figures below present the average change in adjusted coefficient of determination (R2) values for the voxels in the occipital and inferior temporal lobe. The adjusted coefficient of determination (aR2) was calculated at each voxel to compare the proportion of time series variance accounted for by each of the models while adjusting for the different numbers of regressors in each of the models. aR2 is defined as aR2 = 1 - (SS_err / SS_tot) (df_tot / df_err), where SS_err = (standard deviation of the residual errors)2, SS_tot = (standard deviation of time series)2, df_tot = number of degrees of freedom in the data - 1, and df_err = df_tot - 1 — number of degrees of freedom in each model. The resulting aR2 values were averaged over voxels within the reqion of interest and then over subjects.
The ROI for 'Houses vs. Faces' is the inferior temporal lobe and the ROI for 'Retinotopic mapping' is the occipital coretx (both defined by AAL over a normalised structural image).
The results indicate that, yes, it's worth regressing out noise data.
Regressors' legend: P - Pulse simple (filtered & downsampled to TR) PH - Pulse high frequency components (4 covariates) PL - Pulse: low frequency components (2 covariates) R - Respiratory simple (filtered & downsampled to TR) M - Movement (6 covariates)
Map of voxels's change in R2 when GLM included PL and R regressors.
Regressors' legend: P - Pulse simple (filtered & downsampled to TR) PH - Pulse high frequency components (4 covariates) PL - Pulse: low frequency components (2 covariates) R - Respiratory simple (filtered & downsampled to TR) M - Movement (6 covariates)
Map of voxels's change in R2 when GLM included PL and R regressors.
Changes in R2 distribution in the 'Houses vs. Faces' analysis
