Structural MRI is even more abused, where people find "differences" between 2 groups with ridiculously small sample sizes.