AI-Powered Diabetes Analysis and Monitoring System

DOI

10.2991/978-94-6463-704-5_14How to use a DOI?

Keywords

Diabetes; Artificial Intelligence; Machine Learning; Ensemble Models; Diagnostic Tool

Abstract

Diabetes is one of the most pressing global health issues, with an estimated 346 million people affected worldwide according to a 2011 WHO survey. Diabetes mellitus, a metabolic disorder caused by improper insulin utilization, increases the risk of heart attacks, kidney damage, and renal failure. Traditional diagnostic methods often require multiple medical examinations, making them costly and time-consuming. However, rapid advancements in Artificial Intelligence (AI) have introduced efficient diagnostic methods that save time and reduce expenses. This study utilizes the K-Nearest Neighbor (KNN) algorithm and an ensemble-based model for diagnosing type-II diabetes. The dataset used is sourced from Hastie et al.’s diabetes dataset. The study evaluates various machine learning techniques for binary classification to determine whether a patient is diabetic. Among the 15 classifiers considered, five primary approaches are highlighted: Artificial Neural Networks (ANN), Support Vector Machines (SVM), KNN, Naive Bayes, and Ensemble models. Tools like MATLAB 2013a and WEKA 3.6.13 were used for implementation. Ensemble methods demonstrated superior performance by combining the predictive capabilities of multiple classifiers, significantly improving accuracy and reducing misclassification risks. For validation, the system used ten-fold cross-validation, achieving robust results with metrics such as accuracy, precision, recall, and F1-score. The random forest model performed best, achieving 81.2% classification accuracy, 79.8% precision, and strong overall reliability. This research also introduces a diagnostic tool featuring a user-friendly Graphical User Interface (GUI), allowing users to input ten attributes—five nominal and five numeric. By facilitating early and accurate diagnoses, this tool aids physicians in timely treatment planning. The study underscores the growing reliance on AI in medical diagnostics, enabling the analysis of complex data and uncovering hidden patterns. Additionally, challenges such as data insufficiency and deployment difficulties are addressed, emphasizing the need for further advancements to refine model performance and provide deeper insights into the factors influencing diabetes prevalence. Evaluation metrics such as the area under the curve (AUC), mean absolute error (MAE), and root mean square error (RMSE) were employed to measure effectiveness, highlighting the potential of AI-driven models in diabetes screening programs.

Copyright

© 2025 The Author(s)

Open Access

Open Access This chapter is licensed under the terms of the Creative Commons Attribution-NonCommercial 4.0 International License (http://creativecommons.org/licenses/by-nc/4.0/), which permits any noncommercial use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license and indicate if changes were made.

Download article (PDF)

TY  - CONF
AU  - Priyata Mishra
AU  - Kunal Agrawal
AU  - Rishit Rathore
AU  - Nidhi H. Soni
PY  - 2025
DA  - 2025/04/30
TI  - AI-Powered Diabetes Analysis and Monitoring System
BT  - Proceedings of the International Conference on Smart Health and Intelligent Technologies (ICSHit-2024)
PB  - Atlantis Press
SP  - 171
EP  - 189
SN  - 1951-6851
UR  - https://doi.org/10.2991/978-94-6463-704-5_14
DO  - 10.2991/978-94-6463-704-5_14
ID  - Mishra2025
ER  -