Wireless Body Sensor Networks

 

 

The availability of inexpensive biomedical and inertial sensors, that can be worn on or even implanted in the body, combined with advanced signal processing and network communication techniques is driving a revolution in physiological monitoring and intervention. Wireless Body Sensor Networks (WBSNs), which are a very special kind of Wireless Sensor Networks (WSNs), are an enabling technology for unobtrusive patient monitoring. Signals monitored by sensors span from physiological vital signs, such blood pressure, oxygen saturation, electrocardiogram , electroencephalogram or glucose levels, to bio-kinetic and ambient parameters, such as accelerations, temperature, humidity.

The research activity is focused on developing and optimizing of the various components of the distributed measurement system, both at sensor and communication level, and on a careful risk assessment in order to guarantee the safety of monitored patients, as expressly mentioned by the MDD. Significant open issues identified for the current research proposal are the design of advanced signal processing algorithms for operating with signals usually characterized by a very low signal-to noise and disturbance ratio, the analysis of the robustness of current available technologies for supporting the communication of critical signals, the design of suitable data compression algorithms for optimizing the resources consumed for transmitted sensor data without altering the informative content and finally the design of suitable synchronization algorithms for guaranteeing temporal-correlation of data acquired from multiple sensors.

 

Issues

 

Photoplethysmography (PPG) is a widely used technology, routinely employed for heart rate measurement in low-cost medical devices. Monitoring is notoriously more difficult during physical exercise, since motion artifacts may considerably degrade PPG signals. The approach we propose estimates human heart rate and reliably tracks its changes by a robust algorithm, whose main steps include denoising by joint principal component analysis, Fourier-based heart rate measurement and, finally, smoothing and tracking by a Kalman filter. To illustrate its good overall performance, experimental results are presented using publicly available real-life PPG traces.

Matlab code:

DOWNLOAD, please refers to the paper: "A. Galli. G. Frigo, C. Narduzzi, G. Giorgi, Robust Estimation and Tracking of Heart Rate by PPG Signal Analysis, Proc. IEEE International Instrumentation and Measurement Conference, I2MTC 2017, May 2017, Torino, Italy."

Papers:

  • A. Galli. G. Frigo, C. Narduzzi, G. Giorgi, Robust Estimation and Tracking of Heart Rate by PPG Signal Analysis, Proc. IEEE International Instrumentation and Measurement Conference, I2MTC 2017, May 2017, Torino, Italy.