Brain Organizes Emotions as a Structured Map Using Hippocampal-Prefrontal Circuits

Neuroscientists at Emory University have determined that the human brain organizes complex emotional experiences into a structured, navigable internal map, suggesting related feelings are spatially clustered while distinct ones are separated within this neural landscape. This framework offers a new perspective on how emotional processing, interpretation, and regulation occur. The findings, published in Nature Communications on January 26, 2026, detail how specific brain signal patterns recorded via functional Magnetic Resonance Imaging (fMRI) correspond directly to participants' reported emotional shifts.

The investigation utilized fMRI to record signals from the hippocampus and the ventromedial prefrontal cortex (vmPFC) as subjects viewed 14 evocative film clips over 2.5 hours across four scanning sessions. This process revealed a functional division of labor in charting this affective space. The hippocampus, central to memory formation, showed clear signatures for emotion categories in a structured hierarchy of nodes. Specifically, the interior sections distinguished broad contrasts, such as separating generally 'good' from 'bad' experiences, while the posterior sections encoded finer distinctions. For example, brain activity patterns consistently placed emotions like anger and fear in close spatial relation, contrasting sharply with the more distant positioning of happiness and excitement.

The ventromedial prefrontal cortex (vmPFC), a frontal area critical for value assessment, tracked the overall placement of these broader emotional categories within a two-dimensional affective space defined by valence and arousal. This dual-level organization—from general direction to specific states—suggests a sophisticated, hierarchical system for emotion knowledge. The ability to make these fine-grained distinctions, known as emotion granularity, has been linked to superior mental health outcomes, whereas conditions like depression are associated with a 'compressed' map lacking differentiation.

To test if this map-like structure emerges from standard learning, the research team employed computational modeling using the Tolman-Eichenbaum Machine (TEM), an artificial hippocampus designed to learn relational structures. Researchers trained the TEM on an abstract emotion grid derived from the film-viewing data. The model generated patterns mirroring the human brain scans, providing evidence that the brain may construct its emotional organization gradually over time through learning. Subjective emotional timelines were gathered from a separate cohort of 44 raters who marked their moment-by-moment feelings while viewing the same film segments, allowing for comparison with brain activity.

Study co-author Philip Kragel, an Assistant Professor of Psychology at Emory University, noted the research probes how the brain compresses these complex emotional experiences. The findings suggest that emotions operate less chaotically and more like an organized system, which could influence therapeutic approaches for mental health conditions marked by emotional confusion or poor regulation. Future research, according to Kragel, will explore how this internal emotional layout shifts across different cultures, developmental stages, or in the presence of various mental illnesses. The work integrates fMRI data, pattern recognition, and AI simulation to provide an objective, scientific understanding of the mechanisms underlying subjective emotional experience.