Efficient data acquisition systems for Internet-of-Medical-Things applications

In recent years, wearable devices for measuring physiological signals and parameters are becoming more complex due to the integration of several sensors and electronic front–ends. Those devices are going to be integrated into Internet-of-Things (IoT) systems where several acquisition nodes must be connected and managed. Therefore, the Internet-of-Medical-Things (IoMT) systems have been proposed to monitor and manage healthcare applications. One of the main challenges of IoMT nodes is to meet the energy consumption and size requirements of wearable devices. Moreover, IoMT systems must collect and store data from up to thousands of nodes, resulting in a significant data rate to be handled. A possible compression of this data rate must however guarantee biosignal integrity to comply with medical standards. Furthermore, IoMT nodes are going to include not only Personal Area Networks, but also Wireless Wide Area Network interfaces, such as Wi-Fi, LoRa, Sigfox, or Narrowband-IoT (NB-IoT), which are characterized by higher power consumptions and lower data rates. Such challenges push the research community to develop novel sampling methods for data acquisition systems.

This talk will review ongoing efforts to develop wearable devices for IoMT applications. Hardware and software techniques used to acquire and compress physiological signals will be reviewed. In particular, the compressive sampling theory and its application for the compression of ECG signals will be presented. An example of an IoMT system based on a wearable device performing compressive sampling and the results obtained from its experimental characterization in terms of energy consumption and signal quality will be discussed.