1Associate Professor, Department of School of Computational Intelligence, Joy University, Tirunelveli, Tamil Nadu, India, 627116. Email: pkmanojkumar@joyuniversity.edu.in
2Assistant Professor, Department of Computer Science, Sri Ramakrishna College for Arts & Science for Women, Coimbatore, Tamil Nadu, India, 641006. Email: stellacs@srcw.ac.in
3Associate Professor, Department of Mathematics, AMET deemed to be University, Chennai, Tamil Nadu, India, 603112. Email: jenitha.g@ametuniv.ac.in
4Professor & Head, Department of Computer Science and Engineering, A.K.T Memorial College of Engineering and Technology, Kallakurichi, Tamil Nadu, India, 606213. Email: chisundar123@gmail.com
5Associate Professor, Department of Computer Science, SPPU, Pune, Maharashtra, India, 411058. Email: monicaapte@gmail.com
6Assistant Professor, Department of Artificial Intelligence and Machine Learning, Aditya University, Surampalem, Andhra Pradesh, India, 533005. Email: appalakonda.v@adityauniversity.in
7Head & Assistant Professor, Department of Biochemistry, Nadar Saraswathi College of Arts and Science (Autonomous), Theni, Tamilnadu, India, 625531. Email: venibio87@gmail.com
8Associate Professor, Department of Applied Mathematics, Bhilai Institute of Technology Durg, Durg, Chhattisgarh, India, 491001. Email: anilkumardby70@gmail.com
*Corresponding Author: Dr. P. K. Manoj Kumar, Associate Professor, Department of School of Computational Intelligence, Joy University, Tirunelveli, Tamil Nadu, India, 627116. Email: pkmanojkumar@joyuniversity.edu.in
Diabetes mellitus is a chronic metabolic disorder that has emerged as a major global public health challenge, affecting millions worldwide. Early detection plays a crucial role in preventing complications such as cardiovascular diseases, neuropathy, nephropathy, and retinopathy. In recent years, Artificial Intelligence (AI) and Machine Learning (ML) have significantly transformed predictive healthcare, particularly through the utilization of Electronic Health Records (EHRs). This study explores the role of AI-driven models in predicting diabetes using structured and unstructured clinical data extracted from EHR systems. The research focuses on the integration of supervised learning algorithms, deep learning techniques, and data preprocessing methods to enhance predictive accuracy and clinical decision-making. EHR-based prediction systems provide a comprehensive dataset including patient demographics, laboratory results, medical history, and lifestyle factors. AI algorithms such as Random Forest, Support Vector Machines, Logistic Regression, and Neural Networks have demonstrated high accuracy in identifying individuals at risk of developing diabetes. Furthermore, advanced techniques like feature selection, class imbalance handling, and explainable AI contribute to model transparency and reliability. This paper presents a detailed review of existing literature, discusses methodologies, analyzes case studies, and evaluates challenges such as data privacy, bias, and interpretability. The findings highlight that AI-powered EHR systems can significantly improve early diagnosis, reduce healthcare costs, and support personalized treatment strategies. However, ethical considerations and data governance remain critical for successful implementation. Overall, this study emphasizes the transformative potential of AI in predictive healthcare and advocates for interdisciplinary collaboration to enhance diabetes prediction systems.
Keywords: Artificial Intelligence, Machine Learning, Diabetes Prediction, Electronic Health Records, Deep Learning, Healthcare Analytics, Predictive Modeling, Clinical Decision Support
How to cite this article: Kumar PKM, Mary PS, Jenitha G, Jayasundar S, Apte M, Appalakonda V, Krishnaveni M, Dubey AK. AI-Powered Diabetes Prediction using Electronic Health Records. Int J Drug Deliv Technol. 2026;16(12s): 87-93. DOI: 10.25258/ijddt.16.12s.10
Source of support: Nil.
Conflict of interest: None