Limits...
Channel selection and feature projection for cognitive load estimation using ambulatory EEG.

Lan T, Erdogmus D, Adami A, Mathan S, Pavel M - Comput Intell Neurosci (2007)

Bottom Line: The ambulatory cognitive state estimator is utilized in the context of a real-time augmented cognition (AugCog) system that aims to enhance the cognitive performance of a human user through computer-mediated assistance based on assessments of cognitive states using physiological signals including, but not limited to, EEG.This paper focuses particularly on the offline channel selection and feature projection phases of the design and aims to present mutual-information-based techniques that use a simple sample estimator for this quantity.Analyses conducted on data collected from 3 subjects performing 2 tasks (n-back/Larson) at 2 difficulty levels (low/high) demonstrate that the proposed mutual-information-based dimensionality reduction scheme can achieve up to 94% cognitive load estimation accuracy.

View Article: PubMed Central - PubMed

Affiliation: Department of Biomedical Engineering, Oregon Health and ScienceUniversity, Portland, OR 97239, USA.

ABSTRACT
We present an ambulatory cognitive state classification system to assess the subject's mental load based on EEG measurements. The ambulatory cognitive state estimator is utilized in the context of a real-time augmented cognition (AugCog) system that aims to enhance the cognitive performance of a human user through computer-mediated assistance based on assessments of cognitive states using physiological signals including, but not limited to, EEG. This paper focuses particularly on the offline channel selection and feature projection phases of the design and aims to present mutual-information-based techniques that use a simple sample estimator for this quantity. Analyses conducted on data collected from 3 subjects performing 2 tasks (n-back/Larson) at 2 difficulty levels (low/high) demonstrate that the proposed mutual-information-based dimensionality reduction scheme can achieve up to 94% cognitive load estimation accuracy.

No MeSH data available.


PSD-based feature extraction (left) and dimensionality reduction, classification, and postprocessing flow diagrams (right).
© Copyright Policy - open-access
Related In: Results  -  Collection


getmorefigures.php?uid=PMC2267884&req=5

fig1: PSD-based feature extraction (left) and dimensionality reduction, classification, and postprocessing flow diagrams (right).

Mentions: All channels reference the right mastoid. EEG isrecorded at 256 Hz sampling frequency while the subject is performing tasks withvarious cognitive loads. EEG signals are preprocessed to remove eye blinksusing an adaptive linear filter based on the Widrow-Hoff training rule[18]. Information fromthe VEOGLB ocular reference channel was used as the noise reference source forthe adaptive ocular filter. DC drifts were removed using high-pass filters (0.5Hzcut-off). A bandpass filter (between 2 Hz and 50 Hz) was also employed, as thisinterval is generally associated with cognitive activity. The PSD of the EEGsignals, estimated using the Welch method [29] with 1-second windows, is integrated over 5 frequencybands: 4–8 Hz (theta), 8–12 Hz (alpha), 12–16 Hz (low beta), 16–30 Hz (highbeta), 30–44 Hz (gamma). The energy levels in these bands sampled every 0.2seconds (i.e., sliding windows with 80% overlap) are used as the basic inputfeatures for cognitive classification. The particular selection of thefrequency bands is based on well-established interpretations of EEG signals inprior experimental and clinical contexts [24]. The overall schematic diagram of the signal processingsystem is shown in Figure 1.


Channel selection and feature projection for cognitive load estimation using ambulatory EEG.

Lan T, Erdogmus D, Adami A, Mathan S, Pavel M - Comput Intell Neurosci (2007)

PSD-based feature extraction (left) and dimensionality reduction, classification, and postprocessing flow diagrams (right).
© Copyright Policy - open-access
Related In: Results  -  Collection

Show All Figures
getmorefigures.php?uid=PMC2267884&req=5

fig1: PSD-based feature extraction (left) and dimensionality reduction, classification, and postprocessing flow diagrams (right).
Mentions: All channels reference the right mastoid. EEG isrecorded at 256 Hz sampling frequency while the subject is performing tasks withvarious cognitive loads. EEG signals are preprocessed to remove eye blinksusing an adaptive linear filter based on the Widrow-Hoff training rule[18]. Information fromthe VEOGLB ocular reference channel was used as the noise reference source forthe adaptive ocular filter. DC drifts were removed using high-pass filters (0.5Hzcut-off). A bandpass filter (between 2 Hz and 50 Hz) was also employed, as thisinterval is generally associated with cognitive activity. The PSD of the EEGsignals, estimated using the Welch method [29] with 1-second windows, is integrated over 5 frequencybands: 4–8 Hz (theta), 8–12 Hz (alpha), 12–16 Hz (low beta), 16–30 Hz (highbeta), 30–44 Hz (gamma). The energy levels in these bands sampled every 0.2seconds (i.e., sliding windows with 80% overlap) are used as the basic inputfeatures for cognitive classification. The particular selection of thefrequency bands is based on well-established interpretations of EEG signals inprior experimental and clinical contexts [24]. The overall schematic diagram of the signal processingsystem is shown in Figure 1.

Bottom Line: The ambulatory cognitive state estimator is utilized in the context of a real-time augmented cognition (AugCog) system that aims to enhance the cognitive performance of a human user through computer-mediated assistance based on assessments of cognitive states using physiological signals including, but not limited to, EEG.This paper focuses particularly on the offline channel selection and feature projection phases of the design and aims to present mutual-information-based techniques that use a simple sample estimator for this quantity.Analyses conducted on data collected from 3 subjects performing 2 tasks (n-back/Larson) at 2 difficulty levels (low/high) demonstrate that the proposed mutual-information-based dimensionality reduction scheme can achieve up to 94% cognitive load estimation accuracy.

View Article: PubMed Central - PubMed

Affiliation: Department of Biomedical Engineering, Oregon Health and ScienceUniversity, Portland, OR 97239, USA.

ABSTRACT
We present an ambulatory cognitive state classification system to assess the subject's mental load based on EEG measurements. The ambulatory cognitive state estimator is utilized in the context of a real-time augmented cognition (AugCog) system that aims to enhance the cognitive performance of a human user through computer-mediated assistance based on assessments of cognitive states using physiological signals including, but not limited to, EEG. This paper focuses particularly on the offline channel selection and feature projection phases of the design and aims to present mutual-information-based techniques that use a simple sample estimator for this quantity. Analyses conducted on data collected from 3 subjects performing 2 tasks (n-back/Larson) at 2 difficulty levels (low/high) demonstrate that the proposed mutual-information-based dimensionality reduction scheme can achieve up to 94% cognitive load estimation accuracy.

No MeSH data available.