Key points are not available for this paper at this time.
Abstract The integration of Electronic Health Records (EHRs) in healthcare has significantly advanced the field but introduced challenges related to data security and precise diagnoses due to the large data volume. To address these issues, we propose the SA-GBO-ODBN model, combining Blockchain and deep learning (DL) for secure medical data management and diagnostics. This model includes Hyperledger Fabric for tamper-proof storage, optimal key generation, data encryption and decryption, and disease detection functionalities. The Key features of the proposed framework include emergency contact notifications, user data access management, and administrative data modifications. The framework employs SHA-256 and elliptical curve cryptography (ECC) for enhanced data security. ECC uses the Self-Adaptive Gradient-Based Optimizer (SA-GBO) to generate optimal encryption and decryption keys. Hyperledger blockchain technology enables secure medical data sharing, patient visit data storage, and EHR link recording in external databases via multiple channels. After decryption, the Optimized Deep Belief Network–based approach diagnoses epilepsy using real-time EEG datasets. The qualitative and quantitative performance analysis shows the proposed framework’s superiority over existing techniques, with accuracy, False Positive Rate (FPR), and FNR of 98.93%, 0.0199, and 0.0034 for the Bonn EEG dataset, and 99.40%, 0.0196, and 0.0034 for the CHB-MIT dataset, respectively.
Sharma et al. (Thu,) studied this question.