Wearable technology (e.g. smartwatches) that can measure skin conductance signals is a growing market with interesting applications in healthcare and in the workplace (i.e. real-time biofeedback for stress management). In order for this technology to be able to support real-time wearable biofeedback the signals need to be analysed real-time. Skin conductance signals can contain a great deal of noise (e.g. because of motion artifacts). Recently, the first analysis tool for automatic artifact rejection has been made avalaible: http://eda-explorer.media.mit.edu/ (see Figure 1). This tool could be an important building block for the real-time automatic analysis of skin conductance signals. For this thesis, you will do a study into the quality of this tool by using this tool on existing data sets, and on a limited number of your own data in simple scenario’s (e.g. sitting at a desk, walking through a building).
Figure 1. A screenshot of the EDA Explorer website.
This thesis assignment will be done under the supervision of Dr. Matthijs Noordzij.