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{198149,
        author = {Shaik Lathifbasha and Morla Avinash and Anusha},
        title = {Multi-Disease Risk Prediction Using Clinical–Lifestyle Feature Fusion and Deep Tabular Neural Networks},
        journal = {International Journal of Innovative Research in Technology},
        year = {2026},
        volume = {12},
        number = {11},
        pages = {11652-11660},
        issn = {2349-6002},
        url = {https://ijirt.org/article?manuscript=198149},
        abstract = {The rising trends of non-communicable diseases such as cardiovascular disease and diabetes are some of the reasons why reliable early risk prediction systems are required. Physiological conditions and lifestyle behaviours influence these diseases, but most of the existing machine learning approaches focus on disease prediction in isolation and do not sufficiently factor in the interactions between the clinical and lifestyle factors. The current paper introduces a risk prediction framework of multiple diseases combining both clinical and lifestyle features by using a systematic feature fusion strategy combined with deep tabular neural networks. The assessment of clinical indica- tors is done through threshold based abnormality scoring, and lifestyle factors are normalised and synthesised to demonstrate behavioural risk patterns. To describe nonlinear interactions between clinical and lifestyle factors, a sigmoid-gated fusion mechanism is presented, which generates a better feature representation to make predictions.
The feature set is combined and then a deep learning model is trained using TabNet which is able to select the features selectively and learn using structured healthcare data. The system classifies the forecasts in diabetes and heart diseases into three levels of risks namely low, moderate and high. It does this by the possibility-based classification. This helps people to make decisions on how to take care of them so that they can avoid falling sick. The determination of the model performance is done by the use of the F1-score, the confusion matrix and the ROC- AUC.
Experiments indicate that TabNet-based explainable artificial intelligence can be used to accurately predict the probability of various diseases and yet is simple to interpret. The importance of global features and patient-level explanations are considered to identify the most common clinical and lifestyle factors influencing the disease prediction. The suggested framework improves predictive stability, transparency, and scalability. This is why it is suitable in clinical decision support and proactive health monitoring systems.},
        keywords = {Multi-Disease Prediction, Clinical–Lifestyle Feature Fusion, TabNet, Deep Tabular Learning, Healthcare Ana- lytics, Risk Stratification, and Explainable Artificial Intelligence (XAI)},
        month = {April},
        }