There are certain philosophical pitfalls this path has (I worry about misuse from unscrupulous practitioners, misinterpretation of results from those untrained in neuropsych, treating behaviors as hardwired, unchangeable brain states), but I think it's integral to the profession.
I think this is right on the money.
It would seem to me that much of this research would be predicated on beliefs such as "depression constitutes a disease process." However, my take on the literature is that this notion is primarily marketed by those who stand to make a profit from treating this "disease" through the use of drugs. Sure, there might be a very small subset of people who truly suffer from chronic depressed mood due to identifiable pathophysiology (which would itself likely be quite variable from individual to individual). But what we are being sold is that ALL do. I don't see much research to support this. I mean, has anyone seen the several recent meta-analyses that show anti-depressants to essentially be useless in all but the very most severe cases?
If what we are looking for in imaging/psychotherapy outcome research is universal "disease" processes that can be altered via some particular way of doing therapy, I doubt we'll find it...though some will claim that they have. In fact, every day free-standing neuroimaging clinics are bilking people out of thousands of dollars telling them that their MRI indicates "depression." As noted above, current science doesn't allow us to draw that conclusion.
Our brains are changing on a moment-by-moment basis constantly as we interact with the environment. Behaviors are repeated, and through principles of reinforcement/punishment, certain circuits become more resistant to modification. But, as noted above, they are modifiable. So, of course psychotherapy alters neurobiology. However, individual variability (in life history, genetic inheritance, epigenetic factors, etc.), I believe, is likely to preclude finding some "universal" way of treating "depression."
I think neuroimaging is invaluable for many functions, however, I'm not sure if this will be one of them.
Don't mean to hijack this thread, but what do others think??