PCA to flatten unconstrained source kernel

I am attempting to flatten an unconstrained imaging kernel using PCA. However, after running the unconstrained to flat map option (selecting PCA), I noticed that the analysis returns the full results (flattened sources x time samples) as ImageGridAmp, as opposed to the ImagingKernel. Is there anyway to get the ImagingKernel instead?
Thank you for your help,

Hi Rebecca,

I was just looking at this function. The short answer is that unfortunately Brainstorm currently can't do that. But in theory it would be possible and it would save disk space. I'll keep that in mind if we end up modifying the function.