The same photograph printed on matte and glossy paper looks different: some details come out sharper, others softer. Brain maps work the same way. The smoothing method, that is, the way you blur out noise, changes what a researcher ends up seeing.
MRI measures brain tissue properties as numbers: how much myelin, iron or water sits in each image cube (a voxel). But a single cube may hold a mix of gray matter, white matter and fluid, and on top of that noise creeps into the numbers. To keep noise from getting in the way, the map is smoothed: each cube's value is averaged with its neighbors. The question is which neighbors to take.
Scientists took data from 138 healthy people aged 19 to 75 and built maps of four parameters. Each map was smoothed three ways, and they checked where the values change with age. It turned out that the first two methods give almost identical results, while SUSAN finds fewer significant changes but sees them better at the boundaries between tissues. Until now it was unclear how much the choice of smoothing mattered, and now it is clear: it can hide subtle age-related changes or, conversely, bring them out. So if a researcher is looking for changes right at the gray-white matter boundary, SUSAN is the way to go; if the overall picture deep inside the tissue is what matters, one of the first two methods will do.
On MTsat maps in white matter, tissue-weighted smoothing found as many as 93 060 significant voxels, while SUSAN found only 63 340, even though both methods are considered "tissue-specific."
Not published on arXiv. Here are its page, the PDF and the formalization: you can check it yourself, not only read our account.