Development of a preliminary pediatric tracheal growth model from magnetic resonance images.

Abstract

To develop a growth model of the minimum cross-sectional area of the normal pediatric trachea with measurements from magnetic resonance images (MRIs) to supplement the clinical criteria used to determine if a child with tracheal stenosis needs surgery. Retrospective imaging review. A total of 81 patients were imaged for a variety of clinical reasons and declared to have normal tracheas fully visible in their T1 magnetic resonance image. Regression analysis was used to identify any contribution that age, gender, and z scores for height and weight have in predicting the minimum cross-sectional area of the trachea. The best-fit model for minimum cross-sectional area is: Area = -0.00451*age(4) + 0.177*age(3) - 2.05*age(2) + 12.6*age + 8.02 (area in mm(2) and age in years). Gender and z scores for height and weight did not provide any additional explanation of variance in tracheal size. Our study demonstrates the potential to create a growth model of the normal trachea based on cross-sectional area of the trachea using MRIs. Even with the relatively small number of patients used to build it, the model has demonstrated some ability to be used as an objective prediction tool when deciding a treatment path for a patient. With continued development of precise, objective measures to diagnose the severity of the tracheal stenosis, more patients can be given early and accurate prognosis and be treated appropriately.