Ah, so a meta-analysis of 168 phrenology, I mean MRI studies shows “structural” differences between different types of disorders. Well, if you include lacunar infarcts, hydrocephalus and such as neurological disorders, I bet there was a difference. When it comes to brain tumors, imaging is a valuable and lifesaving tool. However, when it comes to function, imaging is largely an expensive exercise in pseudoscience. “Structural” differences in populations are forever being proposed and deposed. Comparisons are based on small convenience samples. If there were normative sizes and shapes for different brain structures, we could measure deviations from the norm in clinical populations. But there are no such norms. So we are free to compare clinical populations to whatever sample we wish, get funding, and publish pretty false-color pictures “proving” our pet theories.