News for hackers. Stuff that matters.

A comment on 40 percent of fMRI signals do not correspond to actual brain activity.

Re:40 percent of fMRI signals do not correspond to actual br... by bschne on Tuesday December 16, 2025 @09:00AM (#46288557)
you're telling me the results of this paper were likely bs? --- https://www.sciencedirect.com/science/article/abs/pii/S10538...
[ Parent | on HN ]
Re:40 percent of fMRI signals do not correspond to actual br... by parpfish on Tuesday December 16, 2025 @09:10AM (#46288668)
The point of the salmon paper is to demonstrate to people “if you do your stats wrong, you’re going to think noise is real” and not “fmri is bs”
[ Parent | on HN ]
Re:40 percent of fMRI signals do not correspond to actual br... by prefrontal on Tuesday December 16, 2025 @10:58AM (#46290153)
As the first author on the salmon paper, yes, that was exactly our point. Researchers were capitalizing on chance in many cases as they failed to do effective corrections to the multiple comparisons problem. We argued with the dead fish that they should.
[ Parent | on HN ]
Re:40 percent of fMRI signals do not correspond to actual br... by rdgthree on Tuesday December 16, 2025 @01:28PM (#46292302)
Nothing to add to this conversation in particular, but just wanted to say - truly amazing paper. Well done!
[ Parent | on HN ]
Re:40 percent of fMRI signals do not correspond to actual br... by prefrontal on Tuesday December 16, 2025 @04:17PM (#46294611)
Many thanks! It was a ton of fun. Hard to beleive that we are coming up on 20 years since the data for the salmon was first collected...
[ Parent | on HN ]
Re:40 percent of fMRI signals do not correspond to actual br... by chuckadams on Tuesday December 16, 2025 @03:56PM (#46294334)
> We argued with the dead fish that they should.

Arguing with a dead fish may be a sign you're working too hard :)

[ Parent | on HN ]
Re:40 percent of fMRI signals do not correspond to actual br... by prefrontal on Tuesday December 16, 2025 @04:19PM (#46294642)
Yeah, it did prove to be a rather one-sided conversation... ;)
[ Parent | on HN ]
Re:40 percent of fMRI signals do not correspond to actual br... by chuckadams on Tuesday December 16, 2025 @04:24PM (#46294703)
Did you try tuning it? https://youtu.be/F2y92obnsc0
[ Parent | on HN ]
Re:40 percent of fMRI signals do not correspond to actual br... by kspacewalk2 on Tuesday December 16, 2025 @09:14AM (#46288715)
Curious what you find to be "bs" about the results of this paper? That statistical corrections are necessary when analysing fMRI scans to prevent spurious "activations" that are only there by chance?
[ Parent | on HN ]
Re:40 percent of fMRI signals do not correspond to actual br... by koolala on Tuesday December 16, 2025 @09:54AM (#46289263)
They were being sarcastic.
[ Parent | on HN ]
Re:40 percent of fMRI signals do not correspond to actual br... by fishnchips on Tuesday December 16, 2025 @09:01AM (#46288568)
Oh man you stole my thunder. I hoped to be the first to bring up the dead salmon.
[ Parent | on HN ]