This Nigerian AI health startup wants to save thousands of babies’ lives with a simple app

A healthy Nigerian baby. Many are not so lucky.
A healthy Nigerian baby. Many are not so lucky.
Image: Reuters/Akintunde Akinleye
We may earn a commission from links on this page.

A Nigerian startup has developed a machine learning system to detect child birth asphyxia earlier and hopes to save thousands of babies’ lives every year when its technology is deployed.

The founders say the AI solution has achieved over 95% prediction accuracy in trials with nearly 1,400 pre-recorded baby cries. The startup is now raising funds to acquire more data to improve accuracy and obtain clinical approval from health institutions.

The young startup is already garnering international attention and is in the second round for the global IBM Watson AI XPRIZE competition, which has a $5 million prize.

Birth asphyxia is the third highest cause of under-five child deaths and is responsible for almost one million neonatal deaths annually, according to WHO. It has also been linked to 1.1 million intrapartum stillbirths, long-term neurological disability and impairment.

Dr. Victoria Feyikemi a physician at Babcock University Teaching Hospital in Ogun state, southwest Nigeria said the condition could be a manifestation of the immaturity of a baby’s respiratory system. “They often present with slow irregular respiration, reduced activity and reduced heart rate,” she said.

Detecting the condition is a challenge in Nigeria and other parts of Africa where the focus usually shifts to preparation for religious and cultural rites right after a child is born. Feyikemi said medical expertise is needed to positively establish diagnosis using blood gas analysis, administer oxygen support and to treat the underlying cause.

Currently deployed as an embedded model on an Android app, Ubenwa, which means baby’s cry, helps parents and caregivers detect asphyxia earlier, without having to wait on doctors.

Charles Onu, Ubenwa’s founder and principal innovator, explained that the startup’s  machine learning system takes an infant’s cry as input, analyses the amplitude and frequency patterns of the cry and provides instant diagnosis of birth asphyxia.

Although the condition is detectable, Onu said few public hospital in the country had the equipment due to its high cost, poor electricity service and an unrealistic routine application for every child.

Ubenwa’s co-founder and engineering lead, Udeogu Innocent, said after being able to achieve a level of success with the model, the startup then deployed its technology to a mobile app for easier mobile diagnosis of birth asphyxia. It builds on techniques that have been developed for speech recognition.

The Ubenwa team is conducting clinical validation exercises in Nigeria at the University of Port Harcourt Teaching Hospital and in Canada at the McGill University Health Centre.

“We want to do the tests in the hospital, interact directly with the babies, and compare how Ubenwa performs given all the new environmental challenges that would come up. The reason we are able to pursue this real-time validation in the clinical setting is as a result of the success of our earlier work,” Onu said.

The company is yet to figure out a definitive monetization business strategy.

“We are still finalizing a hybrid model. But in the meantime, we are planning to make it free for individuals and paid for organizations such as hospitals, clinics, governments, and others),” Onu said.

*Correction: A previous version of this story said Ubenwa is in the final round of the XPRIZE competition. It has been corrected to show it is in the second round.