Copyright © 2026 Authors retain the copyright of this article. This article is an open access article distributed under the Creative Commons Attribution License which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

@article{196218,
        author = {Mangali Divya and K.Ajay Chowdary and Karingula Navya and Vinayak G Biradar},
        title = {A Multimodal Deep Learning Framework For Predicting The Malnutrition Risks},
        journal = {International Journal of Innovative Research in Technology},
        year = {2026},
        volume = {12},
        number = {11},
        pages = {2945-2952},
        issn = {2349-6002},
        url = {https://ijirt.org/article?manuscript=196218},
        abstract = {Malnutrition remains a major public health concern, particularly for children under five, for whom effective treatment depends on early detection. The complex relationships between dietary, environmental, and socioeconomic factors are often difficult to capture by conventional assessment methods, which rely on anthropometric measurements and manual evaluation. We offer a novel multimodal deep learning framework that combines Knowledge Graphs (KG), Graph Neural Networks (GNN), and Long Short-Term Memory (LSTM) networks to overcome these limitations and provide understandable insights while accurately predicting the risk of malnutrition. The LSTM component simulates children’s temporal growth patterns by capturing trends in nutritional and anthropometric data. The KG component, which encodes domain knowledge from medical literature and dietary recommendations, enables reasoning over risk factors. By simulating the relationships between children, households, and communities, the GNN captures social and environmental dependencies. To create a comprehensive risk assessment, an attention-based mechanism integrates these modalities. The suggested method outperforms baseline models (LSTM-only, GNN-only, and KG-only) in accuracy, precision, recall, and F1-score, according to experiments conducted on public and simulated datasets. It also produces interpretable predictions that can help medical professionals make proactive decisions. This study provides a scalable and explicable solution for public health applications by demonstrating the potential of integrating temporal, relational, and knowledge-based modelling for efficient malnutrition risk prediction.},
        keywords = {Malnutrition prediction, Deep Learning, LSTM, Knowledge Graphs, Graph Neural Networks, Explainable AI, Public Health.},
        month = {April},
        }