TY - JOUR
T1 - Performance of an MEG Adaptive-Beamformer Source Reconstruction Technique in the Presence of Additive Low-Rank Interference
AU - Sekihara, Kensuke
AU - Nagarajan, Srikantan S.
AU - Poeppel, David
AU - Marantz, Alec
N1 - Funding Information:
Manuscript received January 20, 2003; revised May 11, 2003. The work of K. Sekihara was supported in part by Grants-in-Aid from the Kayamori Foundation of Informational Science Advancement, in part by Grants-in-Aid from Magnetic Health Science Foundation, and in part by the Ministry of Education, Science, Culture and Sports in Japan under Grant-in-Aid C13680948. The work of S. S. Nagarajan was supported in part by the Whitaker Foundation and in part by the National Institute of Health (NIH) under Grant P41RR12553-03 and Grant R01-DC004855-01A1. The work of D. Poeppel was supported in part by the National Institute of Health (NIH) under Grant NIH-R01-DC5660. Asterisk indicates corresponding author. *K. Sekihara is with the Department of Electronic Systems and Engineering, Tokyo Metropolitan Institute of Technology, Asahigaoka 6-6, Hino, Tokyo 191-0065, Japan (e-mail: [email protected]).
PY - 2004/1
Y1 - 2004/1
N2 - The influence of external interference on neuromagnetic source reconstruction by adaptive beamformer techniques was investigated. In our analysis, we assume that the interference has the following two properties: First, it is additive and uncorrelated with brain activity. Second, its temporal behavior can be characterized by a few distinct activities, and as a result, the spatio-temporal matrix of the interference has a few distinctly large singular values. Namely, the interference can be modeled as a low-rank signal. Under these assumptions, our analysis shows that the adaptive beamformer techniques are insensitive to interference when its spatial singular vectors are so different from a lead field vector of a brain source that the generalized cosine between these two vectors is much smaller than unity. Four types of numerical examples verifying this conclusion are presented.
AB - The influence of external interference on neuromagnetic source reconstruction by adaptive beamformer techniques was investigated. In our analysis, we assume that the interference has the following two properties: First, it is additive and uncorrelated with brain activity. Second, its temporal behavior can be characterized by a few distinct activities, and as a result, the spatio-temporal matrix of the interference has a few distinctly large singular values. Namely, the interference can be modeled as a low-rank signal. Under these assumptions, our analysis shows that the adaptive beamformer techniques are insensitive to interference when its spatial singular vectors are so different from a lead field vector of a brain source that the generalized cosine between these two vectors is much smaller than unity. Four types of numerical examples verifying this conclusion are presented.
KW - Adaptive beamformer
KW - Biomagnetism
KW - Functional neuroimaging
KW - MEG inverse problems
KW - Magnetoencephalography
KW - Neuromagnetic signal processing
UR - http://www.scopus.com/inward/record.url?scp=0348147673&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=0348147673&partnerID=8YFLogxK
U2 - 10.1109/TBME.2003.820329
DO - 10.1109/TBME.2003.820329
M3 - Article
C2 - 14723498
AN - SCOPUS:0348147673
SN - 0018-9294
VL - 51
SP - 90
EP - 99
JO - IEEE Transactions on Biomedical Engineering
JF - IEEE Transactions on Biomedical Engineering
IS - 1
ER -