Newborn Cardiovascular Digital Twins: Personalised Identifiable Computational Models of the Neonatal Circulation

<p>Introduction: Preterm babies have an increased lifetime cardiovascular risk. We aimed to collect ultrasound data on the developing cardiovascular system in healthy late preterm and term newborn and to develop newborn cardiovascular digital twins: personalisable identifiable computational models of the neonatal cardiovascular system. Methods: Single-centre, prospective, observational cohort study recruited healthy term (≥37+0 weeks’ gestation) and late preterm (34+0–36+6 weeks’ gestation) babies. Ultrasound data were collected within 48 h of birth and 3 to 6 weeks later. These data inform personalised 0D closed-loop bond graph models that simulate blood pressure and flows in the newborn cardiovascular system, and model identifiability was demonstrated. Results: Fifteen term and 10 preterm babies were assessed at 2 days old, and 12 (80%) term and 7 (70%) preterm babies underwent a second assessment at median post-menstrual age 295 days. We found no evidence that late preterm birth is associated with structural cardiovascular changes in early life. Although our term and preterm models had similar lower body resistances at birth, by 3 to 6 weeks of age, the preterm group showed evidence of greater vascular resistance. Conclusion: These newborn cardiovascular digital twins suggest a possible mechanism of cardiovascular remodelling in late preterm neonates. </p>