A Newborn Cry-Based Diagnosis System

Chakib Tadj of École de Technologie Supérieure in Canada will design a non-invasive diagnostic tool using computer acoustical analysis of newborn cries to detect serious medical conditions such as heart defects and infections. Cry production in newborns can be influenced by neurological and physiological states. In this project's Phase I, Tadj and his team analyzed cries from 120 healthy and sick newborns to identify acoustic features that can be used for diagnostic purposes. In Phase II, they will build on these results by incorporating an additional 1000 infants, and expanding to other serious pathological conditions afflicting newborns. They will also develop a software-based diagnostic tool that can interpret recorded cries to help neonatologists detect specific pathologies affecting newborns.