NutriScreener: Retrieval-Augmented Multi-Pose Graph Attention Network for Malnourishment Screening

2025-11-20

AI Summary

Manual malnutrition screening in children is time-consuming and difficult to scale in low-resource settings. Method: We propose a Retrieval-Augmented Multi-Pose Graph Attention Network (RA-MGAT) that integrates CLIP-based visual embeddings, context-aware modeling, and class-enhanced knowledge retrieval to enable end-to-end image-level nutritional status classification and anthropometric value prediction. Contribution/Results: RA-MGAT significantly improves recognition of underrepresented classes (e.g., severe wasting) and cross-domain generalization. Clinical evaluation by physicians yields accuracy and efficiency scores of 4.3/5 and 4.6/5, respectively; recall is 0.79 and AUC is 0.82. Anthropometric predictions achieve substantially lower RMSE versus baselines, with robust performance across diverse datasets. This work establishes a novel paradigm for automated, scalable nutritional screening in resource-constrained environments.

Technology Category

• Machine Learning: Graph-based Machine Learning
• Reasoning under Uncertainty: Graphical Models
• Computer Vision: Biometrics, Face, Gesture & Pose

Application Category

• Search and Retrieval-Augmented AI: Web evaluation methodologies and metrics
• Responsible Web: Machine-in-the-loop, human agency and autonomy
• Web Mining and Content Analysis: Large pretrained models with web data

Abstract

Child malnutrition remains a global crisis, yet existing screening methods are laborious and poorly scalable, hindering early intervention. In this work, we present NutriScreener, a retrieval-augmented, multi-pose graph attention network that combines CLIP-based visual embeddings, class-boosted knowledge retrieval, and context awareness to enable robust malnutrition detection and anthropometric prediction from children's images, simultaneously addressing generalizability and class imbalance. In a clinical study, doctors rated it 4.3/5 for accuracy and 4.6/5 for efficiency, confirming its deployment readiness in low-resource settings. Trained and tested on 2,141 children from AnthroVision and additionally evaluated on diverse cross-continent populations, including ARAN and an in-house collected CampusPose dataset, it achieves 0.79 recall, 0.82 AUC, and significantly lower anthropometric RMSEs, demonstrating reliable measurement in unconstrained pediatric settings. Cross-dataset results show up to 25% recall gain and up to 3.5 cm RMSE reduction using demographically matched knowledge bases. NutriScreener offers a scalable and accurate solution for early malnutrition detection in low-resource environments.

Problem

• Automating malnutrition screening from children's images to replace laborious manual methods
• Addressing class imbalance and generalizability in pediatric malnutrition detection
• Providing scalable anthropometric prediction for low-resource healthcare settings

Innovation

• Retrieval-augmented graph network for malnutrition screening
• CLIP embeddings with class-boosted knowledge retrieval
• Multi-pose attention with context awareness for anthropometry