APS

2022 APS Annual Convention · 2022

Predicting Developmental Delay in Very Pre-Term Infants Using Machine Learning

Chicago, IL · May 2022

Poster · Methodology

  • Gozde Demirci
    CUNY Graduate Center
  • Chia-Ling Tsai
    Queens College, CUNY
  • Michael Flory
    NYS Institute for Basic Research in Developmental Disabilities
  • Ha Phan
    NYS Institute for Basic Research in Developmental Disabilities
  • Anne Gordon
    NYS Institute for Basic Research in Developmental Disabilities
  • Santosh Parab
    Richmond University Medical Center
  • Phyllis Kittler
    NYS Institute for Basic Research in Developmental Disabilities

Abstract

Our aim was to use machine learning for early identification of infants who would be delayed at 25 months. Initially, only predictors available at birth were used. ML produced models with high specificity/few false alarms and low selectivity/many misses. Adding subsequent infant assessments as predictors increased the models’ accuracy.

Other

← Poster Session XI