Newborn Resuscitation
Newborn resuscitation research focuses on improving the effectiveness and efficiency of life-saving interventions for newborns experiencing birth asphyxia. Current efforts leverage computer vision techniques, including convolutional neural networks and vision-language models, to analyze video recordings of resuscitation events, automatically identifying key actions (e.g., ventilation, suction) and provider attention patterns. This automated analysis promises to enhance training, optimize resuscitation protocols, and ultimately improve neonatal survival rates by providing objective data for evaluating performance and refining techniques.
Papers
Object Detection During Newborn Resuscitation Activities
Øyvind Meinich-Bache, Kjersti Engan, Ivar Austvoll, Trygve Eftestøl, Helge Myklebust, Ladislaus Blacy Yarrot, Hussein Kidanto, Hege Ersdal
Activity Recognition From Newborn Resuscitation Videos
Øyvind Meinich-Bache, Simon Lennart Austnes, Kjersti Engan, Ivar Austvoll, Trygve Eftestøl, Helge Myklebust, Simeon Kusulla, Hussein Kidanto, Hege Ersdal