Computational models of pulmonary mechanics in the preterm infant

Abstract: Respiratory distress in the newborn, a condition characterized by difficulty breathing, occurs in about 7% of newborns. Non-invasive ventilation has become the preferred method of respiratory support for these infants but often fails in extremely preterm infants (< 26 weeks gestation) despite surfactant treatment, leading to traditional mechanical ventilation interventions that may cause barotrauma and hyperoxia. This failure can manifest as symptoms of progressive lung collapse and thoracoabdominal asynchrony (TAA) that are evident clinically but for which root causes are not easily determined. Since most methods of providing ventilator support have been developed in adults and children then scaled for infants, there are very few treatments aimed primarily at preterm infants, and the inherent fragility of the demographic precludes extensive clinical evaluation. To help investigate symptoms of respiratory distress under various adverse clinical conditions, we developed a computational model of nonlinear lung mechanics in an idealized 1kg infant that can depict dynamic airflow, esophageal pressure, and thoracic volumes to compare to clinical measures. One study considers the effect of a stiff vs floppy chest wall and demonstrates variable rate of volume loss dependent upon noninvasive interventions. In a second study the chest wall is partitioned into rib cage and abdomen, and metrics of TAA are associated with active and passive mechanical risk factors with sensitivity analyses revealing their relative impact. Ongoing work includes examining the effect of continuous positive airway pressure on global parameter sensitivities, and application of parameter identification algorithms with sparsely available data.

Back to Search Research Symposium Abstracts