Using brain scans and machine learning, researchers have identified six types of depression. Such an approach paves the way for better screening and management of this complex disorder.
A founding study
Published in the journal Nature Medicine, this work involved the analysis of functional MRI scans (fMRI) of 801 subjects who had previously been diagnosed with anxiety or depressive disorders. The team examined their brain activity at rest and during a series of cognitive and emotional tests, paying particular attention to regions of the brain active. Using machine learning, they were able to identify six patterns of brain activity.
Secondly, 250 of the participants received, in a randomized manner, different common antidepressants, or followed behavioral therapy, and their response to the treatment was evaluated. This trial revealed distinct patterns based on “biotype ” of depressionas well as the most appropriate therapeutic approach.
For example, venlafaxine was found to be the most effective compound for one biotype characterized by generalized brain hyperactivity, while behavioral talk therapy was the best treatment for another form, characterized by high levels of activity in three regions associated with problem solving and depression.
« To our knowledge, this is the first time that different types of brain function disturbances have been highlighted for depression. », Estimates Leanne Williams, lead author of the study. “ This is essentially a demonstration of a personalized medicine approach to mental health, based on objective measures.. »
Implications for treatment-resistant forms of depression
It is estimated that around a third of people affected by depression suffer from a form that is resistant to conventional treatments.
The authors of the new study hope that their more precise approach will help improve the management of associated symptoms, and allow the identification of mechanisms that could be targeted by new compounds.