Predicting future depression – new approaches

Do you ever worry about your mental health and wonder if you might develop depression? Recent research by Klaas Enno Stephan and his research team suggests that we may soon be able to identify signs of impending depression in advance, meaning significant implications for mental health treatment and prevention.

Stephan and his colleagues used a large set of fMRI data and found that combining different models and connectivity analyses could accurately predict who was at risk for depression. Over three years they studied participants of whom 50% did and 50% did not develop depressive episodes. The team examined combinations of 8 connectivity features and 17 classifiers to detect connections predictive for depressions.
Their findings suggests that in the future, fMRI could be used to help detect individuals at risk of depression early on, which is critically important for prevention and early interventions. However, combining fMRI with other data modalities may be necessary to achieve higher accuracies of clinical utility. Overall, the study offers hope for improved mental health care.


Predicting Future Depressive Episodes from Resting-State fMRI with Generative Embedding; Herman Galioulline, Stefan Frässle, Sam Harrison, Inês Pereira, Jakob Heinzle, Klaas Enno Stephan
medRxiv 2022.11.17.22281138; doi:

Klaas Enno Stephan’s page:

Image: Herman Galioulline