AI-Blockchain-Enabled Framework for Cardio-Diabetic Risk Prediction

Abstract

Cardiovascular diseases and diabetes are increasingly prevalent health concerns in modern society, creating a significant burden on healthcare systems worldwide. This study proposes an AI- and blockchain-enabled framework for cardio-diabetic risk prediction and secure healthcare data management. Patient information is initially collected from medical sources and edge-enabled devices. The collected data undergoes preprocessing and analysis to improve data quality and prediction reliability. A hybrid ensemble machine learning model integrating Support Vector Machine (SVM), Random Forest (RF), and Naive Bayes (NB) is employed to enhance the accuracy and robustness of disease risk prediction. To ensure the security, integrity, and privacy of patient records, encryption techniques and blockchain-based storage are incorporated into the framework. In addition, a rule-based clustering algorithm is introduced to organize healthcare data efficiently and facilitate faster analysis. The blockchain layer provides a transparent, tamper-resistant, and access-controlled environment for managing sensitive medical records, while the machine learning module enables accurate prediction of cardio-diabetic disease risk. Experimental evaluation using real-world healthcare datasets demonstrates that the proposed framework outperforms conventional standalone machine learning models and provides an effective approach for secure healthcare data management and disease prediction, achieving an accuracy of over 88.5%.