HomeTechnologyArtificial intelligenceDecoding Six Basic Emotions From Functional Brain Connectivity Patterns

Decoding Six Basic Emotions From Functional Brain Connectivity Patterns

Overview: Whole-brain functional connectivity patterns successfully classified six basic emotions from neutral expressions.

Source: Science China Press

Emotions are an important part of human intelligence. Identifying specific emotional categories from complex neural patterns (i.e., the neural decoding of emotional information) is an important point in current emotion research.

The categorical emotion models have suggested a range of basic emotion units (e.g., anger, disgust, fear, happiness, sadness, and surprise) that have specialized and independent neural circuits in the brain to support the expression of different emotional information.

As a result, different brain areas are specifically involved in processing specific basic emotions. In recent years, increasing evidence suggests that the representations of basic emotions may be supported by large-scale functional connectivity (FC) networks in the brain.

An article was recently published online titled “Decoding six basic emotions from brain functional connectivity patterns”. Science China Life Sciences by dr. Fang Fang’s group in the School of Psychological and Cognitive Sciences at Peking University.

This study analyzed the neural mechanism of emotional information represented by brain network patterns from a data-driven perspective. Using the sliding window technique and the random forest model, this study constructed the decoding model of emotional brain networks and provided evidence that functional connectivity patterns contain the representative information of basic emotions.

Prof. Fang’s team collected whole-brain fMRI data from human participants as they viewed pictures of faces expressing one of six basic emotions (anger, disgust, fear, happiness, sadness and surprise) or displaying neutral expressions.

They obtained FC patterns for each emotion in brain regions across the brain using the Harvard-Oxford Atlas, and applied multivariate pattern decoding to decode six basic emotions from the neutral expressions.

Must Read
Amazon’s AI image generation tool is how it plans to pull children into its hardware ecosystem

The results showed that the whole-brain FC patterns successfully classified the six basic emotions of neutral expressions. Analyzing the contribution ratio of each brain region during emotion identification further revealed the spatial distribution locations of the top 10 contributing brain nodes for each basic emotion.

Distributions of the top 10 contributing brain nodes for each basic emotion. Credits: Science China Press

This data-driven research method not only identified key regions previously relevant to the study of facial and emotion processing, such as the fusiform gyrus, the right amygdala under fear, but also identified some brain regions rarely reported to contribute to emotion representations, such as the supramarginal gyrus, supracalcarine cortex and other brain areas in the limbic system.

In addition, the brain network-based decoding model showed superior decoding performance over the traditional voxel-wise activation-based decoding model, both on the whole brain regions and on the top 10 contributing brain regions.

In conclusion, the results of this study further indicate that brain network patterns contain more useful information for emotion decoding than voxelwise activation patterns, and imply that there is great potential to study emotion recognition from the functional connectivity between brain regions.

About this emotion and neuroscience research news

Author: Press Office
Source: Science China Press
Contact: Press Service – Science China Press
Image: The image is credited to Science China Press

Original research: Closed access.
Decipher six basic emotions from brain functional connectivity patternsby Chunyu Liu et al. Science China Life Sciences


Abstract

Also see

This shows vegetarian kebabs

Decipher six basic emotions from brain functional connectivity patterns

Although distinctive neural and physiological states are suggested to underlie the six basic emotions, basic emotions are often indistinguishable from functional magnetic resonance imaging (fMRI) voxelwise activation (VA) patterns.

Must Read
Tracking ESG Goals Through the Supply Chain

Here we hypothesize that functional connectivity patterns (FC) in brain regions may contain emotion representation information beyond VA patterns.

We collected whole-brain fMRI data while human participants viewed pictures of faces expressing one of six basic emotions (i.e., anger, disgust, fear, happiness, sadness, and surprise) or displaying neutral expressions.

We obtained FC patterns for each emotion in brain regions across the brain and applied multivariate pattern decoding to decode emotions in the FC pattern representation space.

Our results showed that the whole-brain FC patterns successfully classified not only the six basic emotions of neutral expressions, but also every basic emotion of other emotions.

For each basic emotion, an emotion-representation network was identified that went beyond the classical brain regions for emotion processing. Finally, we showed that within the same brain regions, FC-based decoding consistently outperformed VA-based decoding.

Taken together, our findings revealed that FC patterns carry emotional information and argued for a greater focus on the contribution of FCs to emotion processing.

RELATED ARTICLES

LEAVE A REPLY

Please enter your comment!
Please enter your name here

Most Popular

Recent Comments