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{197386,
        author = {Shazli Ansari and Aarzoo Asar and Omar Khan and Er. Waheeda Dhokley},
        title = {A Variational Graph Neural Network for Predictive Modeling of Neurodegenerative Diseases from Electronic Health Records},
        journal = {International Journal of Innovative Research in Technology},
        year = {2026},
        volume = {12},
        number = {11},
        pages = {12716-12721},
        issn = {2349-6002},
        url = {https://ijirt.org/article?manuscript=197386},
        abstract = {Electronic Health Records (EHR) include heteroge- neous clinical information, which includes diagnoses, medications, procedures, and laboratory measurements that are highly inter- dependent and incomplete. The conventional machine learning models normally consider these variables as independent features, which restricts them in capturing complex clinical relations and uncertainty of the real-world healthcare data.
We suggest a Variational Graph Neural Network (VGNN) framework to make predictive models of neurodegenerative dis- eases based on EHR data in this paper. Our model models clinical events as a graph and trains their interactions by attention-based message passing functions. We develop variational regularization, a method that learns a latent embedding distribution subject to KL-divergence to predict with uncertainty, to enhance robustness in the presence of noisy, sparse and incomplete medical records. The suggested framework presents well-organized patient representations to accurately predict diseases and is a base of smart clinical decision support systems.},
        keywords = {Electronic Health Records, Graph Neural Net- works, Variational Inference, Uncertainty Modeling, Disease Risk Prediction},
        month = {April},
        }