Modeling and Simulation PROJECTS
Information Management PROJECTS
The UNICEF Innovation Fund provides early stage funding and support to frontier technology solutions that benefit children and the world. UNICEF’s Innovation Fund issued Kimetrica a Long Term Agreement (LTA) to back the development and use of open source frontier technology solutions. The LTA will be used by UNICEF offices and other UN Agencies worldwide to solicit and issue contracts based on the needs of requesting offices. Kimetrica will provide UNICEF’s Innovation Fund with services to support the testing, piloting, deployment, customization and further development of frontier technology solutions in UNICEF’s programs. Kimetrica was selected based on its experience and expertise in advanced modeling, data management, information capture, and custom software development. This LTA follows Kimetrica’s previous UNICEF Innovation Fund grant to conduct proof of concept research for MERON.
Modeling and Simulation PROJECTS
Traditional methods for quantifying malnutrition in children involve physical handling of subjects, can be time-consuming and are susceptible to inaccuracy because they require enumerators to interpret the value. Kimetrica has developed an application called Methods for Extremely Rapid Observation of Nutritional Status (MERON) that allows for a non-invasive, time efficient, and tamper-proof approach to assessing the malnutrition status of an individual by using a facial recognition and processing algorithm.
Recently, through a United Nations Children’s Fund (UNICEF) innovation grant, Kimetrica achieved proof of concept with MERON for children and a preliminary classification accuracy level of 60 percent, using 3,500 images of children under-five (6-59 months), collected alongside UNICEF's Standardized Monitoring and Assessment of Relief and Transitions (SMART) survey in Kenya.
MERON's next step for product development is a significant increase in its accuracy for malnutrition detection in children under-five from 60 percent to over 90 percent, which will be achieved through collecting additional image data. Doing so requires the collection of 5,000-15,000 more usable images in tandem with SMART surveys or other nutritional assessments for calibration.
Once MERON achieves high-quality classification ability, it will offer the following benefits:
1. An increase in the accuracy of collecting data on malnutrition.
2. A cost reduction related to the training of enumerators.
3. Use of inconspicuous measurement tools.
4. A less invasive method to measure malnutrition. (In some cultures parents are sensitive to physical handling of their children.)
These benefits could, in turn, result in a number of important outcomes for the diagnosis and treatment of malnutrition in children under five. These include:
1. More appropriate distribution of funding and scarce resources based on accurate measurements.
2. Savings in resources (resources used for training enumerators to take accurate weight for height measurements; transportation of bulky equipment and opportunity cost for communities participating in surveys).
3. Easier data collection in hard to access, high risk or conflict areas, and areas where physical handling of children is culturally not acceptable.
MERON was presented at the Artificial Intelligence for Good Global Summit held in Geneva in May 2018 (Watch the interview) and has been featured in the Smithsonian, New Scientist, Daily Mail and Deutsche Welle.