WIP 73 - Association between automated neonatal EEG background analysis in the first 6 hours of life and MRI in term and late preterm neonates with a history of perinatal depression at birth

Publication Number: WIP 73.7590

Emily S. Taketa, University of California, San Francisco, School of Medicine, San Francisco, CA, United States; Marie-Coralie Cornet, UCSF Benioff Children's Hospital San Francisco, San Francisco, CA, United States

Background: Evaluating early neonatal brain function can inform clinical management of neonates requiring resuscitation at birth and guide conversations with families. Electroencephalogram (EEG) is a validated tool for assessing brain function. However, interpreting EEG background is time-consuming and depends on the institutional resources and interpreter. These limitations could be mitigated with the use of the Brain Stage of the Newborn (BSN), an open-access and automated EEG background quantification tool. However, how BSN results within the first 6 hours after birth relate to brain injury on MRI is poorly understood.

Objective: To describe the association between early BSN and brain injury on MRI in neonates with perinatal depression at birth.

Design/Methods: This retrospective cohort study includes all neonates born greater than 35 weeks of gestation and admitted to a single center level IV intensive care nursery from 2015-2024 with perinatal depression at birth who received an EEG within the first 6 hours of life and had an MRI performed within the first month of life. Perinatal depression was defined as either 1) an Apgar score at 5 minutes less than 5 or 2) intubation in the delivery room or 3) acidosis at birth, evidenced by a pH less than 7 or base excess less than -13 in the first hour after birth. EEG were uploaded in BABACloud, a secure HIPPA-compliant open access software and BSN values extracted. BSN ranges from 0 (no brain activity) to 100 (normal brain activity). MRIs were classified as the presence or absence of HI-related brain injury. We also extracted the presence of other brain injury and brain malformation. Using univariate logistic regression, we will assess the association between median and maximal BSN in the first 6 hours and brain injury on MRI. This study was approved by UCSF IRB (1002694) as exempt for informed consent. Data will be collected from UCSF electronic health records and stored on secure REDcap database. Data abstraction and analysis will be completed by Feburary 2025 and the poster preparation will be completed by March 2025.