People with PTSD show a similar exaggerated amygdala response to fearful faces. Jorge Armony, the Canada Research Chair in Affective Neuroscience at McGill University in Montreal, is studying both PTSD patients and people who have recently experienced a traumatic event and may develop PTSD. Armony and his team want to see if they can use the amygdala signal and other factors to predict who is vulnerable to the disorder and who will be resistant to therapy. "After 6 to 12 months, some people recover -- what's the difference between people who recover and people who don't?" Armony asks.

While fMRI measures of unconscious processes are useful for studying populations of people with an illness, they're not yet precise enough to diagnose an individual with a particular disorder, says Armony. "We can say that [statistically] a person with PTSD will have an exaggerated amygdala response, but that doesn't mean that everyone will have it."

Hans Breiter, a neuroscientist at Harvard Medical School, one of the first researchers to study amygdala activity with fMRI in the mid-1990s, agrees that a more extensive evaluation of the neurological changes in psychiatric disorders is necessary before the technique can have clinical applications. "This approach is promising and is the right first step, but scientists will need to study larger numbers of people with fMRI to get a better sense of the variability in brain functions that underlie anxiety and depression," he says. "They may have very different [brain activity patterns] and may have very different therapeutic needs." He predicts those larger-scale studies will happen within the next five years.

Breiter and other scientists are optimistic that fMRI can one day be used to evaluate the benefits of therapy, but they say it's unclear what brain signals, conscious or unconscious ones, will be the most effective measure.

"The question still remains, how important are these subconscious phenomena?" says Cohen at Brown. "From a cognitive behavioral perspective, the conscious aspects of depression and anxiety are more important."

Both Etkin at Columbia and Armony at McGill are also using fMRI to study conscious processes, such as attention, in people with anxiety disorders; and they plan to examine how these different factors may be important in different anxiety-related diseases, such as depression and eating disorders.

"There's information processing going on in the brain that's completely outside of awareness, which previously we could only investigate with psychoanalysis," says Tom Insel, director of the National Institutes of Mental Health in Bethesda, MD. "Now you can track [those processes] with neuro-imaging -- a tool that may be much more compelling."

This is the second part in an occasional series exploring how new approaches to brain imaging could improve treatment for neurological and psychiatric disorders. The first installment, published on December 20, examined how patients can use real-time fMRI images of their own brains to control chronic pain.