Artificial intelligence against malnutrition

A child on a weighing scale in a district hospital in Cameroon

©Sylvain Liechti


Malnutrition is still widespread in the world. Assessing the height and weight of children is essential to determine their nutritional status. It seems an easy task. In reality, it is challenging.

Body measurements are an essential tool for medical diagnosis and treatment, e.g., to assess the nutritional status of people, to estimate disease risk, and to detect specific medical conditions.

However, these measurements are often difficult to realize in practice, especially with children, and, as a result, are often imprecise and unreliable. Additionally, in many developing countries of today’s world, scales and measuring boards are becoming scarcer than mobile phones.

Several alternatives to traditional body measurement methods have been developed, but to date there are no real appropriate solutions.

45% of the deaths in the children under 5 are attribuable to malnutrition

UNICEF (2018)

The Project

The goal of this project is to develop mobile applications that can estimate body measures from mobile phone pictures in a rapid, easy, automatic, reproducible, and accurate manner.

More precisely, we will develop a mobile application that can estimate the height, weight and nutritional status of children under 5 years of age for developing countries.

In order to do this, we are using state-of-the-art image processing and machine learning techniques.

Key Accomplishments


Data collection to take place in Burkina Faso and Guinea

  • Project Start (April)
  • Data collection in Cambodia during the summer (cf EPFL News)
  • HMP presentation at the Geneva Health Forum (April), the events “Révolution numérique et philanthropie avec Facebook” in Paris (October) and “Technologies for a brighter world” in Bern (October)
Christine Gaulis

Christine Gaulis

Project Manager


EPFL Laboratories




Prof. Jean-Phillippe Thiran


Prof. Robert West