NutriScreener: AI to screen child malnutrition from images
Malnutrition screening is often slow and hard to scale. NutriScreener is a new AI tool that flags undernutrition from children’s images and estimates key body measurements to support early intervention.
- Blends visual cues with a curated knowledge base matched to a child’s age and region to improve fairness and accuracy.
- Uses a multi-pose model that reads body cues across different angles and real-world conditions.
- Clinician feedback: rated 4.3/5 for accuracy and 4.6/5 for efficiency in a clinical study.
- Results: recall 0.79, AUC 0.82; significantly lower measurement errors than prior methods. Matching the knowledge base to demographics boosted recall by up to 25% and cut error by up to 3.5 cm.
Trained/tested on 2,141 children (AnthroVision) and evaluated across diverse populations (including ARAN and CampusPose), NutriScreener aims to scale early screening in low-resource settings so more kids get help sooner.
Responsible use matters: deployments should ensure consent, secure data handling, and local validation.
Paper: https://arxiv.org/abs/2511.16566v1
Paper: https://arxiv.org/abs/2511.16566v1
Register: https://www.AiFeta.com
AI Healthcare GlobalHealth Malnutrition ChildHealth ComputerVision MachineLearning Pediatrics PublicHealth LowResource ResponsibleAI
