Explainable and Fair AI-Based Models for Early Diabetes Risk Prediction: A Review

DOI

10.2991/978-94-6239-697-5_36How to use a DOI?

Keywords

Explainable Artificial Intelligence (XAI); Fair Artificial Intelligence; SHAP; LIME; Early Diabetes Risk Prediction; Type 2 Diabetes Mellitus (T2DM); Clinical Decision Support Systems

Abstract

It has been clearly observed that Type 2 Diabetes Mellitus, which is also known as T2DM, is rising tremendously on a global scale, which not only needs accuracy to detect at an early stage, but also requires sound ethical practices. As per traditional statistical methods, it seems difficult to handle the complicated and non-linear patterns that are in the large multimodal clinical datasets. In contrast to this, artificial intelligence models are better at achieving predictive performance, but it lacks in terms of fairness and explainability. It is crystal clear that in the medical field, it becomes necessary to gain the trust of either the patient or the doctors. So, in this review, I have reviewed some research papers from 2015 to 2025 that include separately explainable AI (XAI) and fairness-based AI models for early Diabetes risk prediction. Yet both explainable AI and fairness AI-based models have not used till now, and this review paper consists of the combination of both AI models to elevate the interpretability and equality in terms of genders and ethnicity. It has already seen that some basic machine learning algorithms, such as Random Forest (RF) and Cat-boost, utilising Electronic Health Record (EHR) data, are achieving the predictive accuracy of 0.99. Along with that, if 38% of SHAP and 26% of LIME usage can increase clinical trust. In this paper, the work of the mitigation framework in the Smart User Interface is highlighted, which successfully diminishes the Equal Opportunity Difference for different age and BMI up to 18 percentage points. Moreover, this review concludes by recommending XAI-fairness frameworks to make sure the AI systems predict accurately and fairly.

Copyright

© 2026 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  - Kavita Kavita
AU  - Gurbinder Singh Brar
AU  - Amandeep Singh
AU  - Kamalpreet Kaur
PY  - 2026
DA  - 2026/06/04
TI  - Explainable and Fair AI-Based Models for Early Diabetes Risk Prediction: A Review
BT  - Proceedings of the Conference on Bridging Engineering Disciplines with AI and Machine Learning (BEDAIML 2026)
PB  - Atlantis Press
SP  - 444
EP  - 450
SN  - 1951-6851
UR  - https://doi.org/10.2991/978-94-6239-697-5_36
DO  - 10.2991/978-94-6239-697-5_36
ID  - Kavita2026
ER  -