Artificial intelligence in the prevention of respiratory distress syndrome

Abstract

Malformations in fetal development affect the health of the product and the mother. Preventing illnesses during gestation, allows for a healthy and dignified life at birth. Data from INEGI in its report on Fetal Death Statistics (EFD) 2022, show that in Mexico there is an average of 72.2 fetal deaths per 100,000 women of childbearing age. Of these deaths, 25,041 deaths were registered during the year 2022. 5950 deaths correspond to gestational disorders, respiratory or cardiovascular disorders and congenital malformations. Lung defects result in induced abortion or Respiratory Distress Syndrome (RDS). RDS can be prevented by clinical studies and radiological criteria. Identification of abnormal developments using digital analysis of lung images during pregnancy can help identify a defect. We propose ONE classification tool using deep learning in a multiclass categorization of bronchopulmonary sequestration, cystic malformations and diaphragmatic hernia, where there is a risk of defect appreciation and thus misjudgment leading to complications or death. Resulting in a model accuracy of 88.88%, out of a set of 42 two-dimensional sonograms.