Question about resampling and smoothing MRI

Hi,

I’m a computer science student and I do my internship in a medical imaging laboratory where I work on machine learning methods for predicting brain age from T1 MRI. I learn every day different things about medical imaging, but I’m still very novice in this field.

In a paper, authors talk about “smoothing MRI with a FWHM kernel of 8mm and resampling with spatial resolution of 8mm”. I understand that they apply a gaussian filter and they reduce MRI resolution, but I have two unanswered questions :

1. How do we reduce MRI resolution ? Is it correct to do this simply by using mean operator among neighbors ? That what they mean by saying “resampling” ?
2. Is it better to apply gaussian kernel before or after the resampling/reduction ?

Thanks in advance.