EEG anywhere and anytime: signal processing algorithm design for high-density mobile electroencephalography

From 01-01-2016 to 31-12-2018

Description

Electroencephalography (EEG) is a non-invasive technique to measure brain activity using scalp electrodes. Currently, there is a growing need for mobile EEG systems that can be worn
continuously during every-day life, outside a lab or hospital environment. This is believed to be a ‘game-changing’ feature for several potential EEG applications. However, moving from highlycontrolled lab experiments to daily-life environments poses many new signal processing (SP) challenges. In this project, we tackle two of these challenges:

1)EEG signals are often contaminated by muscle and motion artifacts, which can be avoided in lab experiments by sitting still. However, when using EEG in real life, such artifacts become the rule rather than the exception. Therefore, we will design novel multi-channel SP algorithms to remove muscle and motion artifacts.

2)To further miniaturize chronic EEG systems, modular architectures have been proposed, in which the head is covered with a multitude of EEG nodes. To reduce the energy consumption in such systems, we will develop novel distributed SP algorithms, which do not require a power-hungry data centralization, but where the processing is distributed over the EEG nodes instead.

Furthermore, we also target a specific application in the context of hearing prostheses, where the signals of a high-density EEG system are used to detect which sound source a user is attending to, to inform a noise reduction algorithm to not remove that source.

Team

• Alexander Bertrand, Promoter

Financing

Funding: FWO - Research Foundation - Flanders

Program/Grant Type: FWO other - Other FWO Grants