Using Machine Learning Analysis to Assist in Differentiating between Necrotizing Enterocolitis and Spontaneous Intestinal Perforation: A Novel Predictive Analytic Tool

Machine learning models achieved 92-98% accuracy in differentiating necrotizing enterocolitis from spontaneous intestinal perforation in preterm neonates prior to surgery. Random forest and ridge logistic regression algorithms analyzed patient characteristics to predict disease etiology, offering a novel tool to guide surgical decision-making between peritoneal drainage and laparotomy.