Login Search
FMDB Transactions on Sustainable Health Science Letters Submit Paper
    Crossref logo

    Archiving

    Portico

    Fernando Martins De Bulhão is a member of

    Crossref logo

Predictive Analysis for Stroke in Genetically Predisposed Diabetic Patients Using Glycemic Variability Patterns

Authors:
S. Jaya Varshini, Balika J. Chelliah, Shriya Santosh, Akshay Sai Abbireddy, G. Mary Amirtha Sagayee

Addresses:
Department of Computer Science and Engineering, SRM Institute of Science and Technology, Ramapuram, Chennai, Tamil Nadu, India. Department of Computing and Information Sciences, University of Technology and Applied Sciences, Nizwa, Muscat, Sultanate of Oman.

Abstract:

The issue of health risks associated with diabetes mellitus, particularly the risk of developing a stroke in case of genetic predisposition, is critical and requires special attention. Nevertheless, current approaches for identifying health risks do not account for the intricate relationships among multiple factors that can lead to certain health conditions, and the predictive ability of modern technologies is limited. That is why the main aim of this research is to develop an advanced predictive machine learning model for identifying stroke risk among diabetes patients. To meet the stated objective, the following healthcare data attributes were used: age, hypertension, heart disease, BMI, smoking status, and mean glucose. Moreover, missing values, categorical variables, and numerical values were processed and normalized. As a result, four algorithms were analyzed: Extra Trees, a Multi-Layer Perceptron (MLP) neural network, XGBoost, and a stacking model consisting of a Random Forest and an Extra Trees ensemble, as well as an MLP classifier. According to the results, the stacking approach achieved the highest prediction accuracy, exceeding 0.90. Meanwhile, the highest prediction accuracy rates (0.91 and 0.86) were achieved with the XGBoost and MLP neural network algorithms, while Extra Trees achieved lower accuracy (0.82). The model has the best accuracy, precision, and recall, proving that the health risk prediction solution is effective. 

Keywords: Stroke Prediction; Diabetes Mellitus; Glycemic Variability; Machine Learning; Extra Trees; MLP Neural Network; Stacking Ensemble; Predictive Analytics; Healthcare Data Analysis.

Received on: 20/05/2025, Revised on: 23/07/2025, Accepted on: 10/10/2025, Published on: 07/05/2026

DOI: 10.69888/FTSHSL.2026.000638

FMDB Transactions on Sustainable Health Science Letters, 2026 Vol. 4 No. 2, Pages: 129-140

  • Views : 245
  • Downloads : 14
Download PDF