Electromyogram-based method to secure wireless body sensor networks for rehabilitation systems

• Published in: Conference proceedings (IEEE Engineering in Medicine and Biology Society. Conf.) Vol. 2017; pp. 1246 - 1249
• Main Authors: Guanghe Zhang, Samuel, Oluwarotimi Williams, Fanghua Liu, Shixiong Chen, Hui Zhou, Haoshi Zhang, Guanglin Li
• Format: Conference Proceeding Journal Article
• Language: English
• Published: United States IEEE 01.07.2017
• Subjects: Electrocardiography; Electromyography; Entropy; Feature extraction; Muscles; NIST; Security
• DOI: 10.1109/EMBC.2017.8037057
• ISSN: 1557-170X

Wireless body sensor networks (WBSNs) provide a platform to track and monitor human health status as well as feedback to the user by capturing and processing certain physiological signals. Since WBSNs need to provide efficient health information privacy, their security has been identified as one of the major challenges, especially for rehabilitation systems. Conventionally, the random numbers (RNs) based on the inter-pulse intervals (IPIs) from electrocardiogram (ECG) recordings have been widely used to secure the data in WBSNs. However, this method is limited in real-time applications such as human posture control. In this study, we proposed a novel electromyogram (EMG) based RN generation method to secure the data acquired from WBSNs systems for rehabilitation. The newly proposed security scheme was tested on EMG signals acquired from 15 healthy subjects by using EMG features. These features were coded into 128-bit RNs with entropy values ranging from 0.96 to 1.00, and hamming distances (HDs) that ranged from 41 to 83. These preliminary results showed that randomness and distinctiveness of those RNs are good enough for authentication and encryption. Findings from the current study suggest that the EMG-based RN generation method would be potential in securing the health information in WBSNs.