Machine learning-based quantitative analysis of barium enema and clinical features for early diagnosis of short-segment Hirschsprung disease in neonate

This study demonstrates that machine learning algorithms combining clinical data with quantitative colon shape analysis from barium enemas can accurately diagnose short-segment Hirschsprung disease in neonates. The models achieved 86% accuracy and outperformed traditional radiologist interpretation, offering a promising tool for early detection in this challenging patient population.