Prediction of Tidal Volume in Newborns Through a Novel Three-Dimensional Model: A Viability Study.

Abstract

Introduction: Although noninvasive monitoring through quantifying rib cage movement has been useful in serial assessment of pulmonary function in newborns, measuring tidal volume (TV) is commonly performed invasively. As it is the most basic measure of pulmonary function, expanding its assessment to a noninvasive measure can contribute to clinical findings and interpretations in neonatal clinical practice.

Objective: (1) Create a noninvasive measurement tool for TV for neonatal clinical use; (2) Evaluate the agreement between measured TV and predicted TV.

Methods: Observational study with healthy newborns. Predicted TV was calculated based on the newborns' weight using the mid-range of the volumes usually set during mechanical ventilation of term infants (5 mL/kg). For measured TV, newborns were filmed in supine position using a digital camera, and their chest circumferences were measured with a nonelastic tape. Body markers delimited the segmentation of the area of interest, and a MATLAB software routine established their relationship with the area in cm² and generated a representative image of the thoracic and abdominal areas in a cylindrical, elliptical-based format, as well as a three-dimensional model to represent quantitative data of measured TV. Paired t-test assessed the means of measured and predicted TV, Pearson's correlation assessed level of association, and Kappa coefficient assessed the agreement between them.

Results: Newborns' gestational ages ranged from 38 to 40 weeks and weights from 2190 to 4125 g. A total of 56 respiratory cycles were validated for analysis. The mean predicted TV according to weight was 5.06 mL/kg, and measured TV was 5 ml/kg (p = 0.31), with a correlation of 0.7 (p < 0.001) and a kappa coefficient of 0.39 (p = 0.01).

Conclusions: MATLAB software routine was a practical and easy-to-use tool to monitor noninvasive TV in a neonatal setting. There was no difference between predicted and estimated TV, with fair agreement between them.