Blockchain-enabled AI healthcare system with fireworks algorithm for disease prediction and feature selection in smart medical systems

A blockchain-enabled AI healthcare system is proposed to enhance disease prediction and secure healthcare data management. The framework integrates advanced artificial intelligence techniques with robust data security mechanisms to ensure accurate forecasting and safe handling of sensitive medical information. The system employs a hybrid GRU–LSTM model for real-time chronic disease prediction using electronic health records, IoT sensor outputs, and wearable device data. To further improve prediction performance, the Fireworks Algorithm is utilized for hyperparameter optimization and feature selection. Patient data are encrypted using RSA-2048. The model is evaluated using the ‘Disease Prediction Using Machine Learning’ dataset, which contains demographic and clinical attributes for predicting diseases such as diabetes. Experimental results demonstrate strong predictive performance with 99.87% accuracy, 98.46% precision, 98% recall, and a 97.53% F1-score. In addition, the blockchain layer provides high operational reliability, achieving 99.99% data integrity and system availability, 100% auditability, and 99.8% data-sharing efficiency. The platform supports approximately 1,500 transactions per minute and implements role-based access control through smart contracts with an execution time of 0.2 s. Overall, the integrated AI–blockchain framework offers a scalable, transparent, and privacy-preserving solution suitable for real-world healthcare applications, including hospital decision-support systems and remote patient monitoring.