Curbing malnutrition with AI

A predictive model helps improve community health in Kenya

In the heart of Samburu County, a county in northern Kenya home to many different communities, towns, and tribes, a group of 20 women till a garden. The bright sun warms their faces as they laugh and tell stories, happy their crops are doing well. Their children run around, excited about making a trip to the local market where the women will sell the food they’ve grown, turning a good profit. But they haven’t always had this garden. In the past, food scarcity was very real. 

A woman wearing glasses and a headscarf holds a child in her lap. The child is looking down.
I’m happy because my kids are healthy, and my family is healthy.
Futama

A community-focused solution

Fatuma is a single mother in her late twenties with three young children. She lives in Samburu County, Kenya, and is part of a mother-to-mother support group that includes pregnant women and those with children 5 years old and younger. As the largest organization in Kenya focused on people and their health, Amref Health Africa is helping Fatuma’s group and other support groups like hers grow, maintain, and sell food. Members of Fatuma’s community keep and sell their crops and have also started raising livestock. Fatuma strongly believes that prevention is better than treatment because she’s seen the results firsthand, and knows that it helps the women in her community live better lives. As a mother, Fatuma has hope for everyone in her community and the other communities Amref has helped. 

Amref has addressed the needs of vulnerable populations in over 35 countries in Africa by improving health and healthcare in a community-focused way. In Kenya alone, Amref is helping over 2 million children avoid malnutrition by establishing a presence within each community. Currently, 18% of children under 5 are stunted, which is an impaired-growth condition resulting from malnutrition. To help solve this, every household in Kenya gets a visit from community health promoters who work with Amref to teach prevention and nutrition to everyone in the family, offering quality care. Amref’s goal is to prevent malnutrition before it occurs, saving the lives of children and ultimately the life of a population, ending country-wide malnutrition by 2027. 

Dr. Shiphrah Kuria, Amref Regional Manager for Reproductive, Maternal, and Child Health, knows that prevention yields better results. Health promoters who understand cultural context visit people in their homes to help them better understand prevention, especially in children 5 years old and younger. The promoters listen to and empower people personally, and refer those who need care to local health facilities. 

Adults and children seated closely in a crowded indoor setting. An adult in blue patterned attire holds an infant wearing a yellow shirt, while another child in blue is partially visible in the foreground.
I love how Amref directly touches communities, for life.
Dr. Shiphrah Kuria
Amref Regional Manager for Reproductive, Maternal, and Child Health
The AI-powered dashboard can help predict and prevent food shortages.

Improving community health with AI 

Amref is working with the University of Southern California (USC), the Kenyan Ministry of Health, and Microsoft AI for Good Lab to develop a model that predicts health risks with AI. Each partner institution contributed their expertise to develop a solution that will empower government and humanitarian groups to tackle community health in a new way.  

“Global health does not fit in a single discipline, so we need to reach out across fields to deal with the biggest issues around the world,” says Dr. Laura Ferguson, Director of Research at the USC Institute on Inequalities in Global Health, and professor of population and public health sciences. 

Built from data gathered throughout the country every month for almost ten years, a tool based on this model will be easy for everyone to use and won’t require any technical training. The model can quickly filter malnutrition data, zoom in on impacted locations, and toggle between immediate and long-term forecasting timeframes. Data is collected regularly so the dashboard can show where malnutrition has affected communities across the country. AI then organizes the information to reflect anticipated seasonal influences and help forecast risks of childhood malnutrition, which is a significant cause of children’s morbidity and mortality.

Amref uses Microsoft 365 for seamless communication along with the predictive model, developed in partnership by USC and the Microsoft AI for Good Lab, giving them insights into current and future risks across Kenya. Food security is critical, and current and future humanitarian crises, worsened by flooding, drought, and other conditions, can be predicted to help prepare for or avoid food shortages altogether. A tool that uses the predictive model will be available to Amref’s partners, allowing them to see real data and better understand how to predict and prevent trends of malnutrition. With AI-powered analysis, organizations will find solutions together faster than ever before. 

The value of prevention

Led by USC experts, the collaborative team is in the process of building a tool that allows Amref, other humanitarian organizations, and policymakers to directly access the predictive model and understand diverse data sources. When combined with other publicly available sources like satellite imagery, one of Microsoft’s key capabilities, data from the Kenyan Ministry of Health can be used to understand the severity of malnutrition in children across Kenya. The model is trained on data that’s being collected in more than 100 countries, and the team hopes it can be adapted to address malnutrition and other health concerns around the world.  

“This work is not only helping mitigate malnutrition risks among children in Kenya, but also preparing us to help tackle this problem globally,” says Dr. Bistra Dilkina, co-director of the USC Center for AI in Society, and associate professor of computer science. 

More accurate information and predictions will enable Amref and others to position resources to prevent malnutrition. 

“The dashboard will change the way partners intervene, enabling them to do evidence-based and timely interventions,” says Dr. Girmaw Abede Tadesse, Principal Research Science Manager, Microsoft AI for Good Lab. 

The mission of the Kenyan Ministry of Health is to have a globally competitive, healthy and productive nation so they can grow economically. The model helps achieve this by organizing resources around planning and budgeting. Malnutrition can be better solved with AI through prediction in the monitoring and evaluation process, which helps conserve resources. The food and nutrition dashboard has the potential to save lives, improve lives, and enhance the quality of life. 

TRANSFORM is Amref’s new 2023-2030 global strategy that champions and supports the people of Africa to have quality and accessible health services. Amref is achieving this through community-led, people-centered primary health systems that address social determinants of health. The new predictive model will help them achieve their goals by showing the risk profiles of malnutrition across different groups, giving them the flexibility to identify hotspots and intervene at the right time. 

Donate Microsoft Rewards points to support Amref.

Imagine a world where wellness thrives, where new hope derives. A vision of health we all adore, transformed!
Dr. Githinji Gitahi
Group Chief Executive Officer, Amref Health Africa