Accelerated Sleep: Automatic State of Vigilance Detection for Implantable Neural Devices

Open Access
Author:
Nohar, Ravendra Mavlankar
Area of Honors:
Engineering Science
Degree:
Bachelor of Science
Document Type:
Thesis
Thesis Supervisors:
  • Bruce Gluckman, Thesis Supervisor
  • Patrick James Drew, Faculty Reader
  • Judith A Todd, Faculty Reader
Keywords:
  • epilepsy
  • seizure
  • acceleration
  • EEG
  • sleep
  • linear discriminant analysis
  • state of vigilance
Abstract:
We demonstrate that head acceleration improves state of vigilance discrimination when combined with electroencephalogram (EEG) measures. We wanted to replace what is normally done in polysomnogram (PSG) studies with just EEG and head acceleration. Therefore we designed a device that was used in the context of ongoing PSG studies that was compatible and synchronizable with the rest of the PSG measurements. Linear discriminant analysis of EEG features with and without accelerometer features was used to classify sleep states. The addition of head acceleration significantly improves the discriminative capability in most cases. This has potential application in automated sleep scoring which could prove useful in an implantable neural device for seizure control in epileptic patients.