Cutting edge brain imaging technique measures how babies respond to different faces

In a joint study between Birkbeck, University of London and University of Essex, researchers at the Centre for Brain and Cognitive Development, used an advanced brain-scanning technique with artificial intelligence, called Neuroadaptive Bayesian Optimisation (NBO), to see how infants' brains respond to familiar and unfamiliar faces. 

The study involved 61 infants aged 5 to 12 months. While the infants looked at faces - including their caregiver’s face, a stranger’s face, and faces that were a mix of both - researchers used electroencephalography (EEG) to track each child’s brain activity. A machine-learning algorithm then analysed the data in real time, identifying the face that produced the strongest brain response for each infant, which gave a strong indication of which face was most engaging for each infant.  

Interestingly, the study found no overall pattern across all infants, but 85% of the babies showed strong individual preferences for certain faces. This finding suggests that traditional studies might miss these personal patterns because they tend to look for group-wide trends rather than individual responses. 

Professor Emily Jones from Birkbeck's School of Psychological Sciences and principal investigator of the study commented:  

“NBO allows us to go beyond group-based approaches and explore individual neural profiles of infants. Our findings suggest that even within this critical developmental period, infants show unique patterns of attention to faces, rather than following a uniform developmental trajectory.” 

The study also reported lower drop-out rates of subjects (15%) compared to traditional infant studies (22%), further underscoring the potential benefits of using NBO for future research. This research is a step forward in understanding early brain development and could eventually help tailor learning experiences to each child’s interests and strengths.  

Dr Anna Gui, a co-leader on the project along with Dr Elena Throm said: “We now have a method that reliably shows us what each individual child is engaged with based on their brain activity. We hope this will help us understand how to personalise experiences from the early stages of life, leveraging each child’s strengths and interests”  

The study, titled “Combining Real-Time Neuroimaging with Machine Learning to Study Attention to Familiar Faces during Infancy: A Proof of Principle Study” was published in the journal Developmental Science.