Real-Time Anomaly Detection and Data Integrity in Hospitals Using Blockchain and AI

The advancement in technology means that most healthcare facilities are now going digital, causing concern over health information security, privacy, and integrity. This research focuses on applying Blockchain with Artificial Intelligence to promote security in managing health information. The proposed architecture is based on the essence of Blockchain to offer data integrity and the usage of AI models for real-time monitoring and threat identification. A 97% accuracy was obtained using a Random Forest model with a simulated data set. It also demonstrated the Precision of 0 and the Recall value of 0. 97 and 0. 99 for standard access patterns and 0 for any other access patterns not included in Figure 1. 92 and 0. Regarding the hyperparameters, we chose n_estimators= 85 for the anomaly detection model. The AUC value, which stands at 0, indicates that the models perform excellently and can be used in clinical practice. R-squared of 0. 95 shows that the model tested can accurately classify the two classes. Concerning the confusion matrix, studies showed a positive rate of 99% and a false positive rate of 1%. 5% is a strong example of the model for preventing present security breaches. The evaluation of feature importance showed that among all the analyzed features, the object’s access location and user ID were the most valuable signals for detecting outliers. The Blockchain layer ensured that every interaction with patients and data regarding them was indelibly documented, enhancing the data's credibility and creating a complete register of alterations to the information.