That was our paper! We showed that you can get false positives (significant brain activity in this case) if fMRI if you don't use the proper statistical corrections. We did win an Ig Nobel for that work in 2012 - it was a ton of fun.