Can quantitative geometric measures and a computational fluid dynamic (CFD) model derived from medical imaging of children with subglottic stenosis (SGS) be effective diagnostic and treatment planning tools?
Background: Congenital SGS is identified when there is airway narrowing in the cricoid area without an endotracheal intubation history. Acquired SGS typically occurs in children with a previous intubation history, usually due to endotracheal tube trauma. A multidisciplinary care approach involves complex decision making from medical and surgical specialists. Computational 3D models could be used as a standardized, quantitative, predictive tool to estimate the effects of various medical and surgical interventions.
Explore this issue:
May 2016Study design: Retrospective chart and imaging review of computed tomography scans of 17 children with SGS in a tertiary care hospital from May 31, 2011, to November 7, 2013.
Synopsis: Four participants received surgical intervention, while 13 did not. All had geometric and CFD modeling, and 15 had polysomnograms completed. Study authors created the pediatric airway atlas (a compilation of airway geometries from radiographically normal children who received regional CT scans for reasons other than airway problems) in an effort to use population-based normative data for comparison to diseased subjects. Several geometric variables significantly distinguished those who received surgical intervention from those who did not: ASG (cross-sectional area of the subglottis), RASG (ratio of the cross-sectional area of the subglottis to the midtrachea); AS(ASG) (atlas score of the ASG), and the PR(ASG) (predicted percent relative reduction of the ASG). The two parameters most sensitive for discriminating the need for intervention were RASG and PR(ASG). PR(DSG) (predicted percent relative reduction of the hydraulic diameter of the subglottic airway) was most effective at discriminating intervention need. Statistically, sleep studies were poor discriminators for surgical intervention. Limitations included data representing a small, single-site, retrospective study; measurements that were performed in a static state; and potential selection bias.
Pages: 1 | 2 | Single Page