Researchers out of Lawson Health Research Institute and Western University are shedding light on major advances in using brain imaging to classify mental illness.
The new study used machine learning to classify, with 92 per cent accuracy, which subtype of post-traumatic stress disorder people had based on MRI scans.
“We’ve known for a long time that people with post-traumatic stress can present quite differently, clinically,” researcher, professor, and psychiatrist Dr. Ruth Lanius told 980 CFPL.
“A group of them, about 70 per cent of them, have too much emotion, too much arousal. About 30 per cent of that group, actually, is very detached from their emotions, they’re very shut down.”
The study involved 181 participants, including those with the more common form of PTSD, those with the less common subtype, and individuals no history of PTSD.
“People are just lying in the MRI scanner, letting their minds wander, and from these scans we can predict with a 92 per cent accuracy who will have the PTSD subtype with too much emotion and who will have the subtype with too little emotion.”
The scans were inputted into a machine learning computer algorithm which was able to predict the diagnoses. Lanius believes the next steps are to use the data to personalize treatment, potentially predicting symptoms and even response to treatment.
“The field of psychiatry does not currently have objective biomarkers like those used to diagnose and understand other illnesses or diseases like cancer,” lead author Andrew Nicholson, PhD, wrote in a statement.
“By discovering and validating patterns of brain activity as biomarkers, we can bring objective measures to psychiatry and transform patient care.”
The study is published in Psychological Medicine.