We all have slightly different brains. Different sizes, different shapes, different in every region. So comparing one person's brain to a crowd only tells you so much.
The most powerful thing I can do is compare a brain to itself. Scan a patient once, scan the same patient again months later, and measure what changed.
If a region has gone from its normal size to markedly smaller, that is not noise and it is not natural variation. Structures do not shrink like that on their own in a short window.
This is the same signal professional sports medicine has been chasing for years. Not a single snapshot, but a trajectory. Not the patient against their peers, but the patient against who they used to be.
A single image can raise a question. Two images across time can answer it.
Whenever the stakes are high, I would rather measure the change than argue about the snapshot.