Data Reliability and Quality in Body Area Networks for Diabetes Monitoring

Abstract

In wireless body area network (WBAN), data captured from different types of wearable, invasive, minimally invasive, or non-invasive sensors have the immense potential to contribute for real-time decisions and effective healthcare services for better diabetes monitoring. Low quality data can be misleading and thus result in inaccurate diagnosis, ineffective health decision-making, and even loss of lives. High data reliability and quality is of paramount importance in WBAN to ensure wide systems’ adoption and technological acceptance. Based on existing literature on WBAN systems, sensor technologies and data quality (DQ) dimensions, a framework is proposed to ensure high data quality and reliability in WBANs for effective and real-time diabetes monitoring. The framework is composed of a set of DQ dimensions to verify that the information gleaned from sensors, processed, and delivered are of high quality so that diabetes patients and healthcare professionals are able to make reliable, high-precision diagnoses, and real-time treatment decisions. Potential research directions are pointed out for further optimization of data quality and reliability in Body Area Network (BAN) on the sensor level, network level, and human-centric level.