There have been some extensive work with PET scans of the brain, I have seen comparative results between schizophrenic and normal patients, not seen any with "depression" but I am sure these exists (Google with either Pet Scan and Depression" or " Biomagnetic and depression", another method attempting to map actual in vivo processes in the brain).
well if you look for depression, you'll find lots of studies dealing with bipolar disorder ... and probably a better case can be made for organic causes there. but really, pet scans don't answer any of the questions raised earlier: e.g. pet scans are used also to map things like 'what happens in the brain when this grad student is solving a math problem?' in other words, there's no evidence that the brain activity is cause and not just a symptom or concrete manifestation of the problem.
