Development of Cortical Functional Networks in Healthy Infants
Abstract number :
1.198
Submission category :
3. Neurophysiology / 3G. Computational Analysis & Modeling of EEG
Year :
2021
Submission ID :
1825870
Source :
www.aesnet.org
Presentation date :
12/4/2021 12:00:00 PM
Published date :
Nov 22, 2021, 06:50 AM
Authors :
Derek Hu, BS - University of California, Irvine; Rachel Smith, PhD - Biomedical Engineering - University of California, Irvine; Rajsekar Rajaraman, MD - Pediatric Neurology - University of California, Los Angeles; Shaun Hussain, MD - Pediatric Neurology - University of California, Los Angeles; Daniel Shrey, MD - Neurology - Children's Hospital of Orange County; Amber Magers - Children's Hospital of Orange County; Clare Skora, MD - Neurology - Children's Hospital of Orange County; Nhi Tran - Children's Hospital of Orange County; Tammy Yuen, MD - Neurology - Children's Hospital of Orange County; Beth Lopour, PhD - Biomedical Engineering - University of California, Irvine
Rationale: Recent studies have demonstrated that functional connectivity networks (FCNs) may have utility as a biomarker of epileptic spasms (ES), a form of epileptic encephalopathy that occurs in infancy (Shrey et al. 2018; Smith et al. 2020; Hu et al. 2020). However, such studies are limited by the lack of information regarding changes in FCNs associated with age during healthy infant development. Here we quantify the evolution of amplitude and phase-based FCNs during healthy infancy, in comparison to infants with ES.
Methods: Clips of EEG during wakefulness and quiet sleep were retrospectively collected for 240 healthy infants aged 0-24 months; subjects were grouped by age in three-month intervals for analysis. For each subject, FCNs were generated for wakefulness and quiet sleep using two complementary methods: broadband cross-correlation (CC) and the weighted phase lag index (wPLI) in the delta, theta, and alpha frequency bands. Topological changes in FCNs were quantified using three different graph theory measures: (1) node strength, (2) clustering coefficient, and (3) path length. Connectivity differences as a function of age were then compared to 32 subjects with ES aged 0-24.4 months. Statistical differences were tested using a one-way ANOVA with Bonferroni correction for multiple comparisons.
Results: For healthy infants, we found that CC-based connectivity was significantly stronger in older infants compared to subjects 0-3 months old (p< 0.05). The theta and alpha-band wPLI connectivity strength also increased with age, with significantly stronger connectivity during sleep at 18-24 months, compared to 0-12 months (p< 0.05). During sleep, CC-based connectivity generally exhibited higher clustering coefficients and lower path lengths for subjects 12-24 months old, compared to subjects 0-9 months old, suggesting a shift toward a small world topology with age. In contrast, alpha-band wPLI during sleep showed a decrease in small worldness with age, with lower clustering coefficients and higher path lengths in subjects 18-24 months old, compared to those 0-12 months old (p< 0.05). The FCNs of ES subjects were significantly stronger than healthy subjects (student’s t-test, p< 0.05), with the differences more pronounced during sleep. This difference remained significant in the absence of age-matching in sleep, when all subjects aged 0-24 months were grouped together (Wilcoxon sign-rank test, n=32, p< 0.05).
Neurophysiology