During resuscitation of non-breathing babies, effective ventilation of the lungs is the most critical action. Real-time feedback and guidance systems may contribute to more optimal treatment. In this study we utilized deep neural networks for detecting ventilations from audio signals extracted from video recordings of 25 newborn resuscitations at Haydom. We managed to detect >95% of ongoing ventilations using the audio signals. This ventilation detection functionality will be integrated as user feedback and guidance system and tested in future research.