The Digester

Using film‑based fMRI data and computational modeling, researchers found that the hippocampus and ventromedial prefrontal cortex encode emotion categories and their relationships in a map-like structure, offering a neurocomputational account of how humans organize abstract emotion knowledge.

Researchers used the Emo-FilM fMRI dataset of film-driven emotion ratings and predictive models to link reported feelings to brain activity.
Self-reported emotions could be decoded from activity in the hippocampus and ventromedial prefrontal cortex.
The hippocampus contained more information about distinct emotion categories while the vmPFC carried more relational and long-term transition information.
Within the hippocampus, anterior regions represented broad valence categories like good or bad and posterior regions represented finer-grained emotion concepts.
A computational model called the Tolman-Eichenbaum Machine learned an abstract emotion graph from the data and reproduced map-like relations between emotion concepts.
Researchers plan to test how this emotion map varies across mental health conditions, cultures and development.

Brain organizes emotions on a map in hippocampus and prefrontal cortex

Articles