Abstract
This paper presents a novel feature extraction approach for motor imagery classification overcoming the weakness of conventional common spatial pattern (CSP) methods, especially for small sample settings. We consider local CSPs generated from individual channels and their neighbors (termed 'local regions') rather than a global CSP generated from all channels. The novelty is to select a few good local regions using interquartile range (IQR) or an 'above the mean' rule based on variance ratio dispersion score (VRDS) and inter-class feature distance (ICFD); instead of computationally expensive cross-validation method. Furthermore, we develop frequency optimization using filter banks by extending the VRDS and ICFD to frequency-optimized local CSPs. The proposed methods are tested on three publicly available brain-computer interface (BCI) datasets: BCI competition III dataset IVa, BCI competition IV dataset I, and BCI competition IV dataset IIb. The proposed method exhibits substantially improved classification accuracy compared to recent related motor imagery (MI) classification methods.
| Original language | English |
|---|---|
| Article number | 8736402 |
| Pages (from-to) | 1378-1388 |
| Number of pages | 11 |
| Journal | IEEE Transactions on Neural Systems and Rehabilitation Engineering |
| Volume | 27 |
| Issue number | 7 |
| DOIs | |
| Publication status | Published - 2019 Jul 1 |
Fingerprint
Keywords
- Brain-computer interfaces (BCIs)
- Common spatial pattern (CSP)
- Electroencephalography (EEG)
- Local feature
- Motor imagery (MI)
ASJC Scopus subject areas
- Neuroscience(all)
- Biomedical Engineering
- Computer Science Applications
Cite this
Frequency-Optimized Local Region Common Spatial Pattern Approach for Motor Imagery Classification. / Park, Yongkoo; Chung, Wonzoo.
In: IEEE Transactions on Neural Systems and Rehabilitation Engineering, Vol. 27, No. 7, 8736402, 01.07.2019, p. 1378-1388.Research output: Contribution to journal › Article
}
TY - JOUR
T1 - Frequency-Optimized Local Region Common Spatial Pattern Approach for Motor Imagery Classification
AU - Park, Yongkoo
AU - Chung, Wonzoo
PY - 2019/7/1
Y1 - 2019/7/1
N2 - This paper presents a novel feature extraction approach for motor imagery classification overcoming the weakness of conventional common spatial pattern (CSP) methods, especially for small sample settings. We consider local CSPs generated from individual channels and their neighbors (termed 'local regions') rather than a global CSP generated from all channels. The novelty is to select a few good local regions using interquartile range (IQR) or an 'above the mean' rule based on variance ratio dispersion score (VRDS) and inter-class feature distance (ICFD); instead of computationally expensive cross-validation method. Furthermore, we develop frequency optimization using filter banks by extending the VRDS and ICFD to frequency-optimized local CSPs. The proposed methods are tested on three publicly available brain-computer interface (BCI) datasets: BCI competition III dataset IVa, BCI competition IV dataset I, and BCI competition IV dataset IIb. The proposed method exhibits substantially improved classification accuracy compared to recent related motor imagery (MI) classification methods.
AB - This paper presents a novel feature extraction approach for motor imagery classification overcoming the weakness of conventional common spatial pattern (CSP) methods, especially for small sample settings. We consider local CSPs generated from individual channels and their neighbors (termed 'local regions') rather than a global CSP generated from all channels. The novelty is to select a few good local regions using interquartile range (IQR) or an 'above the mean' rule based on variance ratio dispersion score (VRDS) and inter-class feature distance (ICFD); instead of computationally expensive cross-validation method. Furthermore, we develop frequency optimization using filter banks by extending the VRDS and ICFD to frequency-optimized local CSPs. The proposed methods are tested on three publicly available brain-computer interface (BCI) datasets: BCI competition III dataset IVa, BCI competition IV dataset I, and BCI competition IV dataset IIb. The proposed method exhibits substantially improved classification accuracy compared to recent related motor imagery (MI) classification methods.
KW - Brain-computer interfaces (BCIs)
KW - Common spatial pattern (CSP)
KW - Electroencephalography (EEG)
KW - Local feature
KW - Motor imagery (MI)
UR - http://www.scopus.com/inward/record.url?scp=85068793555&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85068793555&partnerID=8YFLogxK
U2 - 10.1109/TNSRE.2019.2922713
DO - 10.1109/TNSRE.2019.2922713
M3 - Article
C2 - 31199263
AN - SCOPUS:85068793555
VL - 27
SP - 1378
EP - 1388
JO - IEEE Transactions on Neural Systems and Rehabilitation Engineering
JF - IEEE Transactions on Neural Systems and Rehabilitation Engineering
SN - 1534-4320
IS - 7
M1 - 8736402
ER -