lyman.glm.prewhiten_image_data

lyman.glm.prewhiten_image_data(ts_img, mask_img, X, smooth_fwhm=5)

Estimate autocorrelation and transform data and design for OLS.

Parameters

ts_imgnibabel image

4D image with fMRI data. If using autocorrelation smoothing, the affine must have correct information about voxel size.

mask_imgnibabel image

3D image with mask defining voxels to include in model. Also used to constrain the autocorrelation estimate smoothing.

Xn_tp x n_ev array

Design matrix array. Should have zero mean and no constant.

smooth_fwhmfloat

Size (in mm) of the smoothing kernel for smoothing the autocorrelation estimates. Requires that the time series image affine has correct information about voxel size.

Returns

WYn_tp x n_vox array

Prewhitened time series data for voxels in mask.

WXn_tp x n_ev x n_vox array

Prewhitened design matrix for voxels in mask.