‘Child Growth Monitor’ (CGM) - an application using artificial intelligence (AI) to fight hidden hunger

Lack of capacity and skills, errors in data collection due to lack of understanding of the standard cut offs on the measuring scales often results in misinterpretation of a child’s nutritional status in India. Untrained local childcare or Anganwadi workers often leads to incorrect measurement of children’ height and weight which leads to inaccurate detection of malnutrition. In order to strengthen the process of precise and accurate anthropometric measurements of children, Welthungerhilfe has developed an application using artificial intelligence (AI) called the ‘Child Growth Monitor’ (CGM) application.

CGM app helps to detect undernutrition among children under the age group of five years. This innovative mobile based ‘Child Growth Monitor’ (CGM) application records 3D images of children to calculate their body mass with AI, and calculates height, weight and down to the millimetre. The application works on real time and functions offline. Child Growth Monitor app is based on an open-source component which supports data collection and diagnosis and can run in any cloud-environment or data-center. Data collected and diagnosed can then be found on Welthungerhilfe’s Git-Hub repositories. The app loads the captured data into Microsoft Azure. Nutritionists and IT specialists can then evaluate the scans by using Microsoft AI solutions, pinpointing a child’s dietary health. The app decouples data-collection and diagnosis from producing measurement results.

So far, measuring and diagnosing malnutrition accurately could be done by few trained experts with expensive equipment like digital weighing machine by SECA which has TARE (with which we can weigh a child with mother and then it will only display the weight of child after subtracting the mother’s weight) function and Height board validated by UNICEF which gives accurate and precise measurement. However, as a change from the past, the CGM application offers a solution that enables healthcare workers with limited capacity and skills to get accurate measurements and at the same time replaces the expensive equipment with a smartphone. Once operational fully, higher quality of data will be collected at lower costs and in real-time. The CGM app is currently in the learning stage and is being piloted in parts of Maharashtra, Madhya Pradesh and Rajasthan. Till date, 60 Anganwadi workers have been trained on how to use the application and 38,929 children have been scanned. The community members especially the parents are educated on their child’s growth and on the importance of the application.