MLNN-Based Anomaly Detection and Blockchain-Enabled Security for Wearable Healthcare Systems
Year: 2026
Type: article
Abstract: Wearable healthcare systems have become integral components of modern healthcare delivery by enabling continuous monitoring of patient physiological conditions, remote patient management, and real-time clinical decision-making. Despite their benefits, wearable devices remain vulnerable to cyber threats that compromise patient privacy, system integrity, and healthcare service availability. Traditional security approaches struggle to identify sophisticated attacks due to their reliance on static security mechanisms and signature-based detection methods, making them ineffective against emerging and previously unknown threats. This study presents an integrated security model combining Multi-Layer Neural Network (MLNN)-based anomaly detection with blockchain-enabled security to enhance cyber threat identification and mitigation within wearable healthcare systems. The model integrates intelligent anomaly detection, decentralized transaction validation, immutable audit trails, and data integrity protection within a unified cybersecurity architecture. The MLNN component analyzes network traffic characteristics, authentication records, user behavior patterns, device communications, and session activities to detect abnormal system behavior in real time. Blockchain technology complements the detection mechanism by providing secure transaction validation, tamper-resistant record management, transparency, accountability, and decentralized storage of healthcare transactions. A Design Science Research (DSR) methodology was employed to guide the design, development, implementation, and evaluation of the model. A prototype system was developed within a wearable healthcare environment and evaluated using healthcare monitoring data generated by wearable devices. Experimental evaluation demonstrated excellent detection performance, achieving an accuracy of 97.1%, precision of 96.8%, recall of 97.4%, F1-score of 97.1%, and ROC-AUC value of 0.993. Blockchain evaluation further demonstrated efficient transaction processing, a consensus success rate of 99.4%, and a data integrity verification rate of 99.8%. The findings demonstrate that integrating machine learning with blockchain technology significantly enhanced cybersecurity resilience, strengthened data protection, and improved trust within wearable healthcare environments. The model provides an effective and scalable approach for securing healthcare data against contemporary and emerging cyber threats.
Institution: Kabarak University
Language: English
Subfield: Information Systems
Field: Computer Science
Domain: Physical Sciences
Open Access status: diamond
OpenAlex ID: W7200154698