Serum Bilirubin Prediction for Neonates using Segmentation-Guided Neural Networks

Across the world, millions of newborns suffer from severe neonatal jaundice, a condition that can cause neurological damage and death. Traditional methods for assessing jaundice are financially inaccessible in developing countries, and current machine learning approaches suffer from ease-of-use issues. Our results indicate a segmentation deep learning model can adapt to environmental irregularities in smartphone imagery of blood based bilirubin extraction test strips; segmentations fed into an artificial neural network predict total serum bilirubin levels that strongly correlate with state-of-art measurements (r = 0.87). Our inexpensive system can enable widespread proliferation of neonatal jaundice screening in low-middle income countries and reduce fatalities.