Comparison of the Use of Blink Rate and Blink Rate Variability for Mental State Recognition

Recent research has unearthed that blink rate variability (BRV) can be employed as a psychophysiological measure. However, its efficiency for mental state recognition (MSR) has not been investigated yet. Because BRV can indicate dynamics inherent in eye blinks, we conjectured that BRV might exhibit...

Full description

Saved in:
Bibliographic Details
Published in:IEEE transactions on neural systems and rehabilitation engineering 2019-05, Vol.27 (5), p.867-875
Main Authors: Ren, Peng, Ma, Xiaole, Lai, Wenjia, Zhang, Min, Liu, Shengnan, Wang, Ying, Li, Min, Ma, Dan, Dong, Yeyun, He, Yongsheng, Xu, Xiaolei
Format: Article
Language:English
Subjects:
Algorithms
Area Under Curve
Biomedical monitoring
Biomedical signal processing
blink rate (BR)
blink rate variability (BRV)
Blinking - physiology
Cognition - physiology
Cognitive ability
cognitive neuroscience
dynamics
Emotions - physiology
Feature extraction
Female
Frequency-domain analysis
Hospitals
Humans
Male
Memory - physiology
mental state recognition (MSR)
neural engineering
Psychology
psychophysiological measure
psychophysiological response
Recognition
Recognition, Psychology
Signal Processing, Computer-Assisted
Task analysis
Time series analysis
Variability
Wavelet Analysis
Young Adult
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Recent research has unearthed that blink rate variability (BRV) can be employed as a psychophysiological measure. However, its efficiency for mental state recognition (MSR) has not been investigated yet. Because BRV can indicate dynamics inherent in eye blinks, we conjectured that BRV might exhibit stronger abilities for the MSR if compared with blink rate (BR), known as the leading indicator derived from eye blinks for MSR. Therefore, in this paper, we attempted to differentiate between high and low cognitive loads of an individual through the analyses of BR and BRV, respectively, which could be viewed as a preliminary study for comparing their MSR abilities. First, an n -back experiment was performed to collect data. Then, in order to characterize the phenomenon of BRV, the features were extracted from its time and frequency domains, respectively. Finally, the area under the curve (AUC) values of BRV and BR for MSR were estimated by the ten commonly used classifiers, respectively. The results indicated that BRV achieves significantly higher AUC values than BR, which suggests its strong potentiality for MSR. In sum, the BRV may prove to be a promising method for the MSR, which should be considered in the future.
ISSN:1534-4320
1558-0210
DOI:10.1109/TNSRE.2019.2906371