The Effect of Video Coding on Infant’s Fast Periodic Visual Stimulation Data Quality

Authors: Heather Wallace, Stefania Conte, Emma Kraan

SUNY Campus: Binghamton University

Presentation Type: Poster

Location: UU 111

Presentation #: 68

Timeslot: Session A 9:00-10:00 AM

Abstract: Fast periodic visual stimulation (FPVS) is a cognitive neuroscience tool used to collect information about brain activity in response to rapidly repeated visual stimuli through electroencephalogram (EEG) data acquisition. FPVS research conducted with young infants and toddlers poses the challenge of maintaining their focus on visual stimuli presentations throughout data collection. This study aims to determine whether completing a video coding of the infant’s looking behavior would result in significant improvement of data quality (i.e., higher SNR) compared with data without a video coding procedure. Infants 5.5 to 12.5 months of age were eligible for participation. Participants were divided into 3 groups based on age, 6 months (M = 181.7 days, SD = 7.5 days, n=12), 9 months (M = 273.8 days, SD = 11.4 days, n=12), and 12 months (M = 361.0 days, SD = 3.0 days, n=3). A total of 27 EEG acquisitions were included. Human faces and objects were presented as standard (6Hz) and deviant (1.2Hz) stimuli in an oddball paradigm. Video recordings of participant’s gaze behaviors during the experiment were coded to quantify the looking time per trial. EEG data quality will be compared for acquisitions with and without trials in which participants looked for less than 75% of time and analyzed as a function of participant’s age. This investigation purposes to provide guidelines for improving data quality of FPVS data acquisitions in infant populations.