Archive · 25 Sep 2025
AI Helps Assess Malnutrition in Children, Researchers Find
Scientists have created a new way to check if children are malnourished using artificial intelligence, or AI.
They developed a system called DomainAdapt that analyzes photos of children to estimate their health.
This new method can measure things like height and weight, and also tell if a child is underweight or has other problems.
This is better than older methods because it's faster and can be used in more places, especially where resources are limited.
The AI uses a special collection of pictures called AnthroVision to learn and get better at identifying malnutrition.
This new approach should help healthcare workers and make it easier to provide the right care for kids.
Researchers at IIT-AIIMS Jodhpur developed an AI method to assess childhood malnutrition.
The new method, DomainAdapt, uses a multi-task learning framework.
DomainAdapt predicts key anthropometric measures such as height and weight.
The AI uses a dataset called AnthroVision with 16,938 images.
The method aims to make malnutrition screening faster and more accessible.
- Who
- Researchers at IIT and AIIMS Jodhpur
- What
- Developed an AI method to assess childhood malnutrition.
- When
- Published in the journal MICCAI.
- Where
- Research conducted at IIT-AIIMS Jodhpur, dataset collected in Rajasthan.
- Why
- To improve the accuracy and scalability of malnutrition assessment.
- Researchers
- IIT and AIIMS Jodhpur
- Focus
- Assessing childhood malnutrition
- Method
- AI using DomainAdapt framework
- Data Source
- AnthroVision dataset with 16,938 images
- Measurements
- Height, weight, MUAC, and malnutrition conditions