What is Affective BCI?

A brain-computer interface that detects, classifies, or modulates emotional states from neural signals, enabling technology that responds to the user's affective condition.

What category does Affective BCI belong to in BCI?

Affective BCI is a Applications concept in brain-computer interface science.

Affective BCI — BCI Glossary

An affective BCI (also called an emotional BCI or aBCI) is a system that detects the user's emotional or affective state from brain signals and uses that information to adapt the behavior of a device, application, or therapeutic intervention. Affective BCIs operate at the intersection of neuroscience, affective computing, and brain-computer interface technology.

Emotional Signal Detection

Emotional states produce measurable changes in brain activity:

EEG asymmetry: Frontal alpha asymmetry (greater left frontal alpha power during negative emotions, greater right during positive) is one of the most studied EEG correlates of emotion
Theta and gamma power: Changes in frontal theta and temporal gamma power correlate with emotional arousal and valence
Amygdala and prefrontal activity: Depth electrodes in epilepsy patients have recorded emotion-related activity from amygdala, insula, and prefrontal cortex
Autonomic correlates: While not neural signals per se, heart rate variability, skin conductance, and pupil dilation often complement neural emotion detection

Applications

Adaptive neurostimulation: Closed-loop DBS or cortical stimulation that detects worsening mood (in treatment-resistant depression) and adjusts stimulation parameters. Early clinical research at UCSF and Baylor has demonstrated feasibility.
Communication enrichment: Metzger et al. (2023) demonstrated that an ECoG-based speech BCI could decode not only words but also facial expressions and emotional prosody, enabling more expressive communication for people with paralysis.
Neurofeedback therapy: EEG-based neurofeedback for anxiety, PTSD, and depression trains users to modify their own brain activity patterns associated with negative emotional states.
Adaptive interfaces: Computer interfaces that adjust difficulty, pacing, or content based on detected frustration, engagement, or fatigue.

Challenges

Emotion detection from neural signals is inherently noisy and subjective. Emotional states are not discrete categories with clean neural signatures — they are continuous, multidimensional, context-dependent, and vary across individuals. Classification accuracy for emotion detection rarely exceeds 70-80% even in controlled laboratory settings, limiting practical utility. Cultural and individual variation in emotional expression further complicates universal affective BCI design.