TY - JOUR
T1 - Assessing and compensating for zero-lag correlation effects in time-lagged granger causality analysis of fMRI
AU - Deshpande, Gopikrishna
AU - Sathian, K.
AU - Hu, Xiaoping
N1 - Funding Information:
Manuscript received June 18, 2009; revised November 24, 2009; accepted November 15, 2009. Date of publication November 15, 2009; date of current version May 14, 2010. This work was supported by the Georgia Research Alliance and the National Institutes of Health. The work of G. Deshpande was supported by Biomedical Imaging Technology Center. The work of X. Hu was supported by the National Institutes of Health under Grant R01EB002009. The work of K. Sathian was supported by the National Institutes of Health under Grant R01EY12440 and Grant K24EY17332. Asterisk indicates corresponding author.
PY - 2010/6
Y1 - 2010/6
N2 - Effective connectivity in brain networks can be studied using Granger causality analysis, which is based on temporal precedence, while functional connectivity is usually derived using zero-lag correlation. Due to the smoothing of the neuronal activity by the hemodynamic response inherent in the functional magnetic resonance imaging (fMRI) acquisition process, Granger causality, as normally computed from fMRI data, may be contaminated by zero-lag correlation. Simulations performed in this paper showed that the zero-lag correlation does leak into estimates of time-lagged causality. To eliminate this leak, we introduce a method in which the zero-lag influences are explicitly modeled in the vector autoregressive model but omitted while calculating Granger causality. The effectiveness of this method is demonstrated using fMRI data obtained from healthy humans performing a verbal working memory task.
AB - Effective connectivity in brain networks can be studied using Granger causality analysis, which is based on temporal precedence, while functional connectivity is usually derived using zero-lag correlation. Due to the smoothing of the neuronal activity by the hemodynamic response inherent in the functional magnetic resonance imaging (fMRI) acquisition process, Granger causality, as normally computed from fMRI data, may be contaminated by zero-lag correlation. Simulations performed in this paper showed that the zero-lag correlation does leak into estimates of time-lagged causality. To eliminate this leak, we introduce a method in which the zero-lag influences are explicitly modeled in the vector autoregressive model but omitted while calculating Granger causality. The effectiveness of this method is demonstrated using fMRI data obtained from healthy humans performing a verbal working memory task.
UR - http://www.scopus.com/inward/record.url?scp=77952564109&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=77952564109&partnerID=8YFLogxK
U2 - 10.1109/TBME.2009.2037808
DO - 10.1109/TBME.2009.2037808
M3 - Article
C2 - 20659822
AN - SCOPUS:77952564109
SN - 0018-9294
VL - 57
SP - 1446
EP - 1456
JO - IEEE Transactions on Biomedical Engineering
JF - IEEE Transactions on Biomedical Engineering
IS - 6
M1 - 5464489
ER -