Signal processing for estimation of vital signs from wearable sensors
Title:Signal processing for estimation of vital signs from wearable sensors
Continuous monitoring of physiological changes of patients inside the intensive care units (ICU), following ICU discharge or at their home environments is a crucial step to transform the current healthcare systems. One important vital sign that has recently attracted many research studies is the respiratory rate (RR). Validation of this vital sign obtained from wearable sensors in the clinical environment during continuous monitoring over few hours/days is an extremely important and challenging task.
On the other hand, estimation of RR from wearable sensors is significantly useful in various home-based monitoring applications such as monitoring patients with chronic obstructive pulmonary disease (COPD) at their home environments to manage their symptoms.
Theoretical and Practical part:
In this tutorial, we look into methods for estimation of RR from patients in the hospital ward following ICU discharge using three different simultaneous recordings including photoplethysmography (PPG), electroencephalography (ECG) and accelerometer signals with a focus on accelerometry based estimation of RR.
In addition, we look into signal processing methods such as adaptive filters, synchrosqueezed wavelet transform and Kalman filters applied to process bio-signals from wearable sensors.