Believe it or not, there really is a difference between healthy brains and “mentally ill” brains. Unfortunately, it’s also true that some medications are over-prescribed, and some illnesses are over-diagnosed.
A new study has revealed that MRI scans can be used to diagnose mental disorders like depression, bipolar disorder, schizophrenia, ADHD, and Tourettes, using a machine learning algorithm.
There is only a 1 in 10 million chance that the results of the study would occur by chance.
The study was the result of a collaboration between the Columbia College of Physicians and Surgeons, the New York State Psychiatric Institute, Columbia University, and Yale University.
The results suggest that, at some point in the future, you will be able to get an objective diagnosis for a mental illness from a brain scan, rather than relying on the subjective preconceptions that you or your therapist may have. It is closer to basing diagnosis on how the brain works, rather than a list of symptoms chosen by committee.
Previous attempts to diagnose mental disorders by brain scan were less successful because they focused in on single parts of the brain. Ravi Bansal and the team suggested that we should be approaching the subject more like fingerprint identification. Fingerprints aren’t identified by individual lines, but by patterns.
The scans, and the algorithm used to compare them, looked at patterns in the shape of the brain. The results were remarkably consistent.
The differences between people who had been previously diagnosed as having a disorder or being healthy were staggering. The probability that we would see these results by chance is only one in ten million.
The algorithm accurately diagnosed people with the correct disorder between 89 and 100 percent of the time. There was one exception for people who had been diagnosed with a mild risk of depression, in which case the algorithm diagnosed them correctly 74 percent of the time. This would be expected, of course, since these people were already low risk.
What is perhaps more interesting about this technology is that it allows us to identify mental illness based on brain scans, even though we don’t know exactly how these mental illnesses work. The fact that the algorithm requires an analysis of patterns throughout the entire brain certainly lends credibility to the idea that mental illness is a whole brain phenomenon in most circumstances, not an issue related to single parts of the brain.
In any case, it’s increasingly clear that mental illness is in fact a disorder of the brain. We can’t rule out the possibility that “emotional trauma” and environment play a part, but the result is indisputably an effect of the brain itself, not just “the mind.”
In addition to diagnosis, this technology offers a promising lens into understanding how these diseases work in the first place. Scientists still know remarkably little about how the brain works, and how it can go wrong. By measuring the differences in brain structure between people with various disorders, we may come to understand ourselves in ways that weren’t possible before.
The Other Side of Normal: How Biology Is Providing the Clues to Unlock the Secrets of Normal and Abnormal Behavior