Evaluating Head Models for Cortical Source Localization of the Face-Sensitive N290 Component in Infants

Area of Science:

• Neuroscience
• Developmental Neuroscience
• Cognitive Neuroscience

Background:

• Accurate cortical source localization of event-related potentials (ERPs) relies on participant-specific structural magnetic resonance imaging (MRI).
• Limited MRI accessibility poses a challenge for developmental studies, particularly with infants.

Purpose of the Study:

• To compare the accuracy of source localization for infant N290 event-related potentials (ERPs) using participant-specific head models versus various substitute head models.
• To determine the optimal substitute head model for developmental ERP research when participant-specific MRIs are unavailable.

Main Methods:

• Measured N290 ERP responses to faces versus toys in 36 infants (4.5-12 months).
• Constructed substitute head models: individual-based (close/far measures), age-appropriate average template, and age-inappropriate average template.
• Compared source localization results from substitute models against participant-specific models.

Main Results:

• Participant-specific models showed greater N290 responses to faces than toys in the middle fusiform gyrus (mFG).
• Individual-based 'close' substitute models best preserved these face-specific activations and showed the highest fit with participant-specific models.
• Average templates, including age-appropriate ones, did not consistently support the stimulus effect.

Conclusions:

• Individual-based head models with closely matched measures are the most effective substitute for participant-specific MRIs in infant ERP source localization.
• Reliance on average templates may obscure crucial findings in developmental face-processing research.
• Optimized head modeling is critical for advancing developmental neuroscience research.