According to the World Health Organisation, nearly 1 million newborns die every year from birth asphyxia, the majority in sub-Saharan Africa. Many of these deaths are preventable if warning signs are recognised early enough. Birth asphyxia occurs when a baby’s brain and vital organs do not receive enough oxygen during or shortly after delivery.
Charles Onu, a Nigerian-born AI researcher, witnessed the devastating impact of these preventable neonatal deaths in low-resource settings and decided to act. This conviction led to the founding of Ubenwa Health. This innovative project uses artificial intelligence to analyse an infant’s cry as a low-cost, accessible tool for the early detection of birth asphyxia.
The name Ubenwa, which means “the cry of a child” in Igbo, reflects the heart of the technology. Using machine learning algorithms, Ubenwa analyses the acoustic patterns of a newborn’s cry to detect abnormalities linked to birth asphyxia. The system is designed as a mobile app, requiring only a smartphone and a microphone. Health workers or caregivers can record a short sample of an infant’s cry, and the AI model interprets it within seconds, offering an immediate, non-invasive, and affordable screening result.
Clinical validation studies have shown promising outcomes, with Ubenwa’s accuracy approaching that of conventional, hospital-based diagnostic methods but at a fraction of the cost and infrastructure requirements.
By lowering the barrier to detection, Ubenwa provides a scalable and accessible solution to a major cause of infant mortality. Furthermore, Ubenwa’s approach is non-invasive and cost-effective, requiring no blood tests, probes, or specialised training. This makes it especially relevant for communities where healthcare budgets and personnel are already stretched thin.
Ubenwa Health is an inspiring example of how African innovation is addressing global health challenges. By turning a baby’s first cry into a powerful diagnostic signal, this startup is redefining what’s possible in neonatal care.
Learn more about Ubenwa : https://www.ubenwa.ai/
