Noninvasive Arterial CO₂ Estimation for Neonatal Ventilation Using Gated Recurrent Neural Networks in the Preterm...

We present two Gated Recurrent Unit (GRU) neural networks for estimating the arterial partial pressure of carbon dioxide (PaCO₂) in mechanically ventilated neonates, using time series data of the end-tidal partial pressure of carbon dioxide (PetCO₂) and noninvasive respiratory and vital measurements. To train and validate these models, we use secondary data of preterm lambs. 1,182 arterial blood gas analyses (ABGs) from 58 animals are included. We compare our approaches to two clinical methods: (i) using PetCO₂ as a surrogate for PaCO₂ and (ii) adjusting PetCO₂ with a constant offset based on invasive ABG analysis. Estimation errors are evaluated by comparing model outputs with PaCO₂ measurements from ABGs via Bland-Altman analyses. Estimation by PetCO₂ and the GRU model using a univariate time series of PetCO₂ results the lowest accuracy. The invasive PetCO₂+Offset method shows improvement, while the highest accuracy is achieved with a GRU model that incorporates a noninvasive multivariate time series of PetCO₂ along with vital and respiratory measurements.