Monitoring childhood malnutrition with Microsoft Azure

German famine prevention charity (in English, World Famine Relief) has partnered with Microsoft to use Azure and common smartphone technology in a bid to improve malnutrition detection in children, both for treatment of individuals and for better reporting.

The organization's new Child Growth Monitor (CGM) lets volunteers and NGO workers do a quick scan of a child with a standard smartphone camera. The app use a combination of anthropometric measurements assessed with machine learning and infrared to spot signs of malnutrition, cutting down on the laborious and often inaccurate of manually measuring height and weight.

"Mumbai slums are apocalyptic," stated Joachim Schwarz, head of the project, in a blog. According to Schwarz, the app has been deployed and tested in India with the French organization Action Contre de la Faim. While worldwide food shortage monitoring has helped to spot and prevent many famines since the 1990s except in isolated and war-torn areas, childhood malnutrition has proven to be a much more intractable issue. In the initial test, five ACF teams measured a few children twice, manually and with the app to hone the machine learning algorithm's accuracy.

Jochen Moninger, Welthungerhilfe's innovation lead spoke with MSDW about the organization's work and the promise of the new app. With over 250 million euros of annual revenue, Welthungerhilfe is one of Germany's leading charities. The organization earlier launched an app called AgriShare that links small farmers in Asia and Africa with equipment rental and transport options to help produce and sell their crops which they previously lacked.