Detection and Removal of EEG Chewing and Ambulation Artifacts for the Purpose of Increasing Robustness of Seizure Detection

The aim of the internship is for the intern to take up the challenge of detecting and removing noise from brain signals that are recorded using electroencephalogram (EEG). The noise that is of interest in the project is mainly caused by the subject chewing and walking. These noises are found to have caused the inability to have a high accuracy in performing seizure detection using EEG. Machine learning-based approaches are to be taken in the attempt to characterize these noises and subsequently eliminate it from the recorded brain signal. The success of this project is a crucial accomplishment for the partner organization, Avertus Inc, as the organization strives for the ability to provide an accurate seizure detection technology.

Faculty Supervisor:

Julie Audet

Student:

Kevin Ardian

Partner:

University of Toronto

Discipline:

Engineering - biomedical

Sector:

Medical devices

University:

Program: