A novel chaotic-based advanced encryption standard algorithm with blockchain technology for secured IoMT data transfer and disease diagnosis

The increasing adoption of Internet of Medical Things (IoMT) devices generates vast amounts of sensor data in healthcare, posing challenges in data security, privacy and processing efficiency. This work addresses these issues by proposing a secure and efficient storage and classification system for skin disease detection. We present a blockchain-based healthcare storage solution where skin images are collected by IoMT devices, encrypted using the Chaotic-based Advanced Encryption Standard (CAES), and stored on a blockchain for secure, decentralized access. The encrypted images are then retrieved and classified using an Adaptive Multi-Scale Dilation-assisted Residual Attention Network (AMSD-RAN), optimized with a Hybrid Heuristic Algorithm, Renovated Position-based Galactic Swarm with Yellow Saddle Goatfish Optimization (RPGS-YSGO). It provides a scalable and secure solution for real-time healthcare applications. The proposed method shows 95% accuracy which is improved by 6.7% for Harvard Dataverse dataset and by 12% for PH2 dataset and significant reduction in processing time compared to previous approaches. This solution ensures secure storage, accurate skin disease detection and overcomes the limitations of traditional IoT and centralized management systems through blockchain’s decentralized architecture.