Recent Advances in Securing Healthcare Data with Quaternion-Based Evolutionary Gravitational Neocognitron Neural Networks and Encoder-Elliptic Curve Deep Neural Networks Integrated with Blockchain

Abstract

The rapid expansion of digital healthcare systems, electronic health records (EHRs), and Internet of Medical Things (IoMT) devices has significantly transformed the generation, storage, transmission, and access of sensitive patient data, while also introducing serious concerns related to security, privacy, and data integrity that traditional cryptographic approaches struggle to address. This review explores recent advancements at the intersection of deep learning, evolutionary computation, algebraic cryptography, and blockchain technology, focusing on the integration of Quaternion-Based Evolutionary Gravitational Neocognitron Neural Networks (QEGN-NN) and Encoder-Elliptic Curve Deep Neural Networks (E-ECDNN) for secure healthcare frameworks. Quaternion neural networks effectively represent multidimensional medical data in hypercomplex space, reducing redundancy and enhancing learning efficiency, while their combination with evolutionary and gravitational optimization improves convergence, robustness, and generalization across tasks such as medical imaging and disease prediction. E-ECDNN further enhances security by embedding elliptic curve cryptography into neural processing, enabling efficient and secure data encoding. When integrated with blockchain platforms, these models support decentralized, tamper-resistant data management, ensuring transparency, secure access control, and privacy preservation. Overall, this integrated approach demonstrates strong potential for building scalable, secure, and intelligent next-generation healthcare systems.