Adaptive Filter to Remove Motion Artifacts from GSR Sensor Embedded on Handle Cane.

Abstract

Wearable devices capable of recording physiological variables such as pulse and electrodermal activity have become popular in the form of smart watches or bracelets. The measurement is affected by the relative movement of the device with respect to the skin, generating artifacts. An alternative to watches or bracelets that can improve adherence for population groups such as elderly people may be the integration of sensors into everyday objects, such as walking sticks. In this case, the challenge of motion artifacts is particularly difficult. This paper presents a strategy based on a Notch filter that adapts its center frequency from the reading of a force sensor placed under the GSR sensor. A noticeable reduction of the interference from motion artifacts is observed.