Accurate prediction of 2-year-old neurodevelopmental outcomes in mild HIE with neonatal white matter microstructure.

Motivation: Hypoxic-ischemic encephalopathy (HIE) is a major cause of neonatal mortality and long-term disability. Mild HIE remains understudied, despite adverse outcomes, and lacks effective methods for predicting neurodevelopmental outcomes. Goal(s): We aim to use neonatal diffusion MRI(dMRI) to evaluate white matter (WM) microstructure alterations in mild and moderate HIE, and to predict their outcomes at age 2. Approach: We use neonatal dMRI to investigate WM microstructure of mild HIE at birth and develop ensemble models for prediction of their outcomes at 2. Results: Mild HIE subjects show reduced FA in key WM tracts with notable language deficits at 2. Ensemble models accurately predict outcomes. Impact: This study establishes that diffusion metrics can identify white matter tract alterations in mild HIE at birth, and ML offers reliable prediction of future neurodevelopmental outcomes expected at age 2. Early identification enables tailored interventions, benefiting patient outcomes significantly.