J 2017

Testing of features for fatigue detection in EOG

NĚMCOVÁ, Andrea, Oto JANOUŠEK, Martin VITEK and Ivo PROVAZNÍK

Basic information

Original name

Testing of features for fatigue detection in EOG

Authors

NĚMCOVÁ, Andrea (203 Czech Republic), Oto JANOUŠEK (203 Czech Republic), Martin VITEK (203 Czech Republic) and Ivo PROVAZNÍK (203 Czech Republic, guarantor, belonging to the institution)

Edition

BIO-MEDICAL MATERIALS AND ENGINEERING, AMSTERDAM, IOS PRESS, 2017, 0959-2989

Other information

Language

English

Type of outcome

Článek v odborném periodiku

Field of Study

20601 Medical engineering

Country of publisher

Netherlands

Confidentiality degree

není předmětem státního či obchodního tajemství

Impact factor

Impact factor: 0.872

RIV identification code

RIV/00216224:14110/17:00097867

Organization unit

Faculty of Medicine

DOI

UT WoS

000408296300005

Keywords in English

Biopac; blink; electrooculography; REM; scenes; SEM

Tags

Tags

International impact, Reviewed
Změněno: 20/3/2018 18:40, Soňa Böhmová

Abstract

V originále

The article deals with the testing of features for fatigue detection in electrooculography (EOG) records. An optimal methodology for EOG signal acquisition is described; the Biopac data acquisition system was used. EOG signals were being recorded while 10 volunteers were watching prepared scenes. Three scenes were created for this purpose a rotating ball, a video of driving a car, and a cross. Recorded EOG signals were processed and 20 features were extracted. The features involved blinks, slow eye movement (SEM), rapid eye movement (REM), eye instability, magnitude, and periodicity. These features were statistically tested and discussed in terms of fatigue detection ability. Some of the features were compared with published results. Finally, the best features - fatigue indicators - were selected.