New research published in Nature Scientific Reports claims to have found one new way to detect mental disorders thanks to the development of an AI. Specialists could detect certain patterns in mental illness much more quickly than is currently the case.
Today, the detection and diagnosis of mental disorders is a multi-stage process. In fact, in countries like the United States it is so slow that many patients can wait in queues for up to two years before having a first appointment with a psychiatrist specializing in the area. However, this new research, carried out by members of Georgia State University, could change things.
“We’ve built artificial intelligence models to interpret large amounts of information from magnetic resonance imaging (fMRI),” says Sergey Plis, associate professor of computer science and neuroscience at Georgia State University. In his statement, Plis says that while fMRIs are much more comprehensive examinations than a traditional blood test or MRI, “the sheer amount of data is much more difficult to interpret.”
In addition, they are not easy to obtain and usually cost much more than a blood test. However, thanks to the development of an AI model, we could see a big change soon. Something very similar to AI that allows us to detect breast cancer faster.
How this artificial intelligence works to detect mental disorders
AI is able to recognize early features of disorders such as ASD (Autism Spectrum Disorder), schizophrenia, and Alzheimer’s disease. Of course, while none of the three can be prevented at the moment (ASD, for example, is present even before a child is born), early diagnosis is essential for effective treatment.
But how exactly does it work? This AI was trained from scans made with MRI. This allows to measure brain activity dynamically, so changes in the blood flow of the organ are presented.
To train it, a database with more than 10,000 samples of people. This allowed the AI to record the images offered by the fMRI and link them to brain function. Likewise, the AI was offered a database with 1,200 samples of different patients, who presented any of the aforementioned disorders.
A promising future for patients with these disorders
Thanks to all this information, the AI was able to detect certain patterns in the behavior of the brain of patients who presented features of these psychiatric disorders. Additionally, it can pinpoint the exact time on an fMRI when the data most closely matched the disorder in question.
As for people with Alzheimer’s and schizophrenia, the study yields very positive conclusions. Unlike ASD, predicting when the first symptoms of these first two begin to appear can be essential in its early treatment.
In fact, Vince Calhoun, co-author of the study and founding director of the Georgia State TReNDS Center, says, “If we could find markers to predict Alzheimer’s risk in a 40-year-old, we might be able to do something about it.”
Schizophrenia, meanwhile, could be detected before it begins to change the structure of the brain. This step would be crucial if you want to prevent this type of disorder in the future.
Even if we know from other tests or family history that someone is at risk for a disorder like Alzheimer’s, we still can’t predict exactly when it will happen.
Brain imaging could shorten that window of time, picking up relevant patterns when they appear before clinical disease manifests.
Vince Calhoun, co-author of the study