Date Available


Year of Publication


Degree Name

Master of Science in Biomedical Engineering

Document Type



Graduate School


Biomedical Engineering

First Advisor

Dr. Abhijit Patwardhan


Feeding problems in preterm neonates stem from complications of early delivery. Attainment of independent feeding is a prerequisite for Neonatal Intensive Care Unit (NICU) discharges. Some quantitative studies of infant feeding involve excessive amounts of time for data processing. Multivariate spectral analysis was used to minimize time for investigation of variability in these rhythms. Auto and Cross-spectral parameters of the rhythms were correlated with Gestational Age (GA), Postmenstrual Age (PMA), Birthweight (BW), Days of Life (DOL), and Time Since First Nipple feeding (TSFN). Auto-spectral analysis showed 25.55% increase in Bandwidth of suck (bw-su) for a 2-week increase in GA (DOL fixed) and 8.99% increase in bw-su for a 10-day increase in DOL (GA fixed). Crossspectral analysis showed a decrease of 0.158Hz of Bandwidth of Suck-Swallow (bw-SS) for a 2-week increase in GA for GA later than 28 weeks. For GA earlier than 28 weeks, peak coherence decreased by 0.774 for a 2-week increase in GA (PMA fixed) and decreased by 0.126 for a 2-week increase in PMA (GA fixed). The method describes the progression of feeding rhythms through correlations with clinical indexes, thus providing clinicians with an understanding of the development of infant feeding and helps predict long-term developmental outcomes.