Mining big data to identify risk factors for sudden unexpected death in infancy (SUDI)

Emeritus Prof Ed Mitchell, University of Auckland

Mining big data to identify new risk factors for sudden unexpected death in infancy (SUDI)

Every year in New Zealand, around one in every 200 pregnancies ends in stillbirth[1] and 40 to 60 babies die suddenly in their sleep.[2] Both are frightening prospects for parents and some babies are more vulnerable than others.

Microsoft scientists are mining big data for insights

John Kahan, Microsoft’s GM Customer Data and Analytics, tragically lost a baby to SUDI in 2003. His experience has inspired a recent collaboration between Microsoft, Seattle Children’s Hospital and some of the top SUDI and stillbirth researchers worldwide, including Emeritus Professor Ed Mitchell of the University of Auckland. The researchers are using AI (artificial intelligence) and machine learning to reveal new insights into the causes of SUDI.

2019 was a productive year for the team

Professor Mitchell brings a wealth of experience with SUDI and late stillbirth epidemiology to this project. During 2019, he’s been working with Microsoft’s data scientists to mine big data from the Centers for Disease Control and prevention birth cohort linked birth/infant death data sets. The team has analysed 20,685,463 births and 19,127 SUDIs computationally to reveal patterns, trends and associations. So far the collaboration has been highly productive, showing for the first time that maternal smoking prior to pregnancy increases the risk of SUDI. Findings have been published in Pediatrics®, with more papers under review and many others in development.

Cure Kids grant means work will continue and the potential for further discoveries to be made

A Cure Kids grant is enabling Professor Mitchell to continue his work with Microsoft’s leading data scientists in 2020. We anticipate more findings that will deepen our understanding of SUDI and late-stillbirth risk.




Donate now to help fund important research into SUDI