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{195549,
        author = {SILAMBARASAN G and Dr. I. Laurence Aroquiaraj},
        title = {Explainable Graph Neural Network Framework for Multimodal Early Plant Disease Prediction},
        journal = {International Journal of Innovative Research in Technology},
        year = {2026},
        volume = {12},
        number = {11},
        pages = {656-674},
        issn = {2349-6002},
        url = {https://ijirt.org/article?manuscript=195549},
        abstract = {Early detection of plant diseases is critical for improving crop productivity, reducing agricultural losses, and supporting sustainable farming practices. Traditional plant disease detection methods predominantly rely on visual inspection of leaf symptoms or image-based deep learning models, which often identify diseases only after visible damage has occurred. Such reactive approaches limit the ability of farmers and agricultural systems to perform timely interventions. To address this limitation, this study proposes an explainable graph neural network framework for multimodal early plant disease prediction that integrates heterogeneous agricultural data sources and models the complex interactions between environmental conditions, soil parameters, and plant physiological responses. The proposed framework combines multimodal sensing data, including soil sensor measurements, environmental monitoring variables, plant physiological indicators, and image-derived features, into a unified graph-based representation of plant health dynamics. A graph neural network is employed to learn relational dependencies among plant health indicators, enabling the model to capture complex interactions that influence disease development. In addition, an explainable artificial intelligence module is incorporated to provide interpretable insights into the factors contributing to disease predictions, highlighting key environmental conditions and physiological signals associated with plant stress. Experimental evaluation demonstrates that the proposed framework achieves high predictive performance and improved interpretability compared with conventional machine learning and deep learning approaches. The integration of multimodal data and graph-based relational learning enables early detection of plant stress conditions before visible disease symptoms emerge. (Wang et al., 2025)The explainability component further enhances transparency and supports agronomic decision-making by identifying the underlying causes of plant health deterioration. Overall, the proposed approach provides a scalable and intelligent solution for next-generation precision agriculture systems, enabling proactive plant health monitoring and data-driven crop management strategies.},
        keywords = {Plant disease prediction; Graph neural networks; Multimodal learning; Explainable artificial intelligence; Precision agriculture; Agricultural sensor networks; Plant health monitoring; Deep learning in agriculture.},
        month = {April},
        }