An enhanced hybrid framework for IoT healthcare security using blockchain-driven multimedia data analysis and cybersecurity techniques
Jyoti Parashar, Bùi Thanh Hùng, Kamal Upreti, Pravin Ramdas Kshirsagar, Shubham Mahajan, Seifedine Kadry · Array · 2026
In the era of digital healthcare, safeguarding sensitive patient information while ensuring real-time access and decision-making is paramount. This study presents a novel Hybrid Blockchain-IoT Framework for secure healthcare data management, integrating Elman neural network–based Blowfish encryption with blockchain and deep anomaly detection. The framework leverages IoT sensor data and utilizes a Proof-of-Authority (PoA) consensus mechanism to ensure tamper-proof transaction recording across decentralized nodes. A Long Short-Term Memory (LSTM) autoencoder combined with a Support Vector Machine (SVM) classifier enables accurate anomaly detection, while cryptographic functions ensure privacy and data integrity. The proposed system is evaluated using a healthcare dataset comprising over 1000 patient records across three network configurations (195, 585, and 1171 nodes). Results demonstrate a Wormhole Attack Probability (%) as low as 1.1%, Product Drop Ratio (%) between 1.2 and 2.7%, and Authentication Delay under 111 ms—outperforming existing systems. Although the anomaly detection accuracy (98.98%) and F1-Score (0.90) are slightly below leading deep learning models, our framework uniquely combines encrypted transmission, distributed validation, and intelligent threat detection in a practical healthcare setting. The architecture ensures security, scalability, and efficiency, positioning it as a robust solution for next-generation smart healthcare ecosystems.