P.074 Risk factors for perinatal arterial ischemic stroke (PAIS): A machine learning approach

Background: Perinatal arterial ischemic stroke (PAIS) is a leading cause of hemiparetic cerebral palsy. Multiple risk factors are associated with PAIS but studies limited by small sample sizes and complex interactions. Unbiased machine learning applied to larger datasets may enable the development robust predictive models. We aimed use identify compare these existing literature. Methods: Common data elements maternal, delivery, neonatal were collected from three perinatal registries one control over 7-year period. Inclusion criteria MRI-confirmed PAIS, term birth, idiopathic etiology. Random forest in combination feature selection was used develop model PAIS. Results: Total 2571 neonates included (527 cases, 2044 controls). Risk uniquely identified through infertility, miscarriage, primigravida, meconium. When compared, both literature-based maternal age, fetal tobacco exposure, intrapartum fever, low 5-minute APGAR. Conclusions: Machine offers novel, less biased method predictors pathophysiology. Our findings support known associations concepts placental disease difficult transition early screening for