Feasibility of EEG-Based BCI for Infant Perceptual Development Monitoring

Although the use of Brain-Computer Interface (BCI) technology in pediatric populations, particularly infants, is still limited, this research explored its potential as a tool for monitoring infants' perceptual development. The study analyzed electroencephalogram (EEG) recordings to detect brain responses to looming stimuli in 100 infants at two developmental stages: pre-locomotor (ages 3–5 months) and crawling (ages 10–12 months). Since rapidly approaching objects often signal danger and trigger avoidance, this setup allowed investigation of the neural correlates of infants' perceptual responses to threats. Both age groups exhibited similar looming-related brain responses in the parieto-occipital cortex, with consistent inter-trial phase coherence and event-related synchronized oscillations in theta and alpha bands. Using these insights, the study developed a support vector machine (SVM) classifier to distinguish between epochs with and without looming-related brain responses. Trained on data from older infants, the classifier performed well on younger infants' data, achieving an accuracy of 70.5±18.5% based on signal amplitude. These findings confirm the potential of using EEG in BCI settings to monitor infants' perceptual development, suggesting a promising direction for future research and applications.