They say, repeatedly and maddeningly in Amen's view, that he claims to diagnose psychiatric illnesses from the scans. He just as repeatedly insists that he never uses SPECT to "diagnose" anything, but looks at the scan within the context of the person's life and after a full psychiatric evaluation to get more information that might shed light on puzzling, difficult, or anomalous symptoms. "I've never said the image is 'the answer.' We use it when we don't know what's going on, when we think it might supply a piece of the puzzle--it's like radar that helps us better target the problem. The alternative to doing scans is blindly putting kids on meds or multiple meds without looking at their brains either before or afterward."
Finally, and most damagingly, they say that what Amen is doing is totally unsupported by the facts--that he hasn't remotely demonstrated the scientific validity or reliability of using scans as he does. They argue that it isn't yet possible to identify subtle distinctions in single subjects that reliably correlate with specific symptoms. "When a radiologist looks at a chest X-ray, he recognizes pneumonia, or an enlarged heart or liquid at the base of the lungs that correlate with disease because he's compared these abnormal scans to thousands of normals and other abnormals," says neuroscientist Bush. "Maybe in 50 years, when someone comes in looking as if he or she has a certain syndrome, we'll be able to refer them to a radiologist who'd be able to interpret that individual scan. But we're not yet, as a field, in a position to distinguish one brain scan from another. Right now, you can look at any individual scan and what you'll find is color blobs across a screen. What separates science from nonscience is the ability to pick out quantitative, statistically reliable, patterns [in those blobs] based on large numbers of scans--set certain criteria for a disorder, show it in 50 cases compared with 50 normals, for example. Amen isn't doing that. Essentially, all he's doing is 'blobology.'"
To this, Amen responds that factoring in the individual variability in each brain, rather than drawing general, statistical inferences from a large base of cases, is a strength of his method, not a weakness. Researchers look for signature features in a scan that will reliably correlate with specific DSM categories, which do not reflect the complexity and heterogeneity of psychiatric disorders as they're experienced by real people, Amen argues. " DSM diagnoses are artificially derived from symptom clusters," he says. "But they don't explain why people are having these symptoms--why one person's depression or anxiety or AD/HD may be entirely different from the same formal diagnosis in somebody else. Only by looking at each person's individual scan can you get some insight into the underlying physiological pattern in the brain, so we can target treatment specifically to what that person needs."