Researchers led by Han Lu have developed a neurocomputational marker that predicts the future risk of depression in adolescents, according to a study published in the journal Science.
The study used the marker to predict future depression risk in a sample of 725 young people. Scientists said the tool reflects altered emotional processing and could clarify the link between depression and neuronal regularisation, serving as a tool for early detection and prevention of major depressive disorder.
To reach their conclusions, the research team analyzed neuroimaging data related to the visual processing of angry faces from a group of 1,332 adolescents. Weakened visual representations of these faces were associated with depressive symptoms within the group.
Brain modeling and emotional processing
Researchers also built a deep learning model to simulate how the brain encodes abstract emotional concepts, such as anger, which regulate visual information processing. By disrupting the model across different experiments, they concluded that weaker visual representations related to depression reflect an over-regularised process.
In an over-regularised process, an abstract belief about anger prevails over actual visual signals shown on a face. Scientists found that the neurocomputational footprint of this over-regularisation was linked to higher emotional symptoms in 19-year-old adolescents and predicted the development of emotional symptoms at age 23.
The study noted that this neurocomputational marker is associated with genetic risk for depression. Researchers added that the marker persists in patients who suffer from depression.
