Fast Random Permutation Tests Enable Objective Evaluation of Methods for Single-Subject fMRI Analysis

International Journal of Biomedical Imaging - Tập 2011 - Trang 1-15 - 2011
Anders Eklund1,2, Mats Andersson1,2, Hans Knutsson1,2
1Center for Medical Image Science and Visualization (CMIV), Linköping University, Linköping, Sweden
2Division of Medical Informatics, Department of Biomedical Engineering, Linköping University, Linköping, Sweden

Tóm tắt

Parametric statistical methods, such asZ-,t-, andF-values, are traditionally employed in functional magnetic resonance imaging (fMRI) for identifying areas in the brain that are active with a certain degree of statistical significance. These parametric methods, however, have two major drawbacks. First, it is assumed that the observed data are Gaussian distributed and independent; assumptions that generally are not valid for fMRI data. Second, the statistical test distribution can be derived theoretically only for very simple linear detection statistics. With nonparametric statistical methods, the two limitations described above can be overcome. The major drawback of non-parametric methods is the computational burden with processing times ranging from hours to days, which so far have made them impractical for routine use in single-subject fMRI analysis. In this work, it is shown how the computational power of cost-efficient graphics processing units (GPUs) can be used to speed up random permutation tests. A test with 10000 permutations takes less than a minute, making statistical analysis of advanced detection methods in fMRI practically feasible. To exemplify the permutation-based approach, brain activity maps generated by the general linear model (GLM) and canonical correlation analysis (CCA) are compared at the same significance level.

Từ khóa


Tài liệu tham khảo

10.1002/mrm.1910340618

10.1016/S1053-8119(03)00149-6

10.1016/j.neuroimage.2005.07.005

10.1155/IJBI/2006/12014

10.1006/nimg.2000.0609

10.1006/nimg.2001.0933

10.1002/1522-2594(200102)45:2<323::AID-MRM1041>3.0.CO;2-#

10.1016/S1053-8119(03)00077-6

2003, Magnetic Resonance in Medicine, 49, 1152, 10.1002/mrm.10469

10.1016/j.neuroimage.2005.06.070

10.1073/pnas.0600244103

10.1016/j.neuroimage.2008.06.037

10.1109/JSTSP.2008.2007788

1987

1957, The American Statistician, 11, 13

1996, Journal of Cerebral Blood Flow & Metabolism, 16, 7, 10.1097/00004647-199601000-00002

10.1002/mrm.1910350219

10.1002/(SICI)1097-0193(1997)5:3<168::AID-HBM3>3.0.CO;2-1

10.1016/S0730-725X(97)00135-5

10.1109/42.918475

10.1002/1097-0193(200102)12:2<61::AID-HBM1004>3.0.CO;2-W

10.1002/hbm.1058

10.1191/0962280203sm341ra

10.1016/j.neuroimage.2004.04.035

10.1002/hbm.20045

10.1016/j.neuroimage.2004.11.046

10.1016/j.neuroimage.2006.03.058

10.1016/j.neuroimage.2006.10.025

10.1016/j.cmpb.2011.07.007

10.1007/s11554-010-0162-9

2011, Computer Methods and Programs in Biomedicine, 102, 238, 10.1016/j.cmpb.2010.05.003

2010, BMC Bioinformatics, 11, 329, 10.1186/1471-2105-11-329

10.1016/j.cmpb.2010.04.010

1993, Human Brain Mapping, 1, 153

1994, Human Brain Mapping, 2, 189, 10.1002/hbm.460020402

10.1006/nimg.1999.0508

2004

1957, The Annals of Mathematical Statistics, 28, 181, 10.1214/aoms/1177707045

10.1006/nimg.1999.0435

10.1016/j.neuroimage.2004.01.033

10.1002/mrm.10671

10.1016/j.neuroimage.2004.07.048

10.1109/JSTSP.2008.2007819

1978, Biometrika, 65, 297, 10.1093/biomet/65.2.297

1936, Biometrika, 28, 322

1963, Annals of Mathematical Statistics, 34, 1270, 10.1214/aoms/1177703863

1994, Linear Algebra and Its Applications, 210, 29

2007

2010

10.1002/jmri.21494

10.1016/j.neuroimage.2006.11.037

1999, IEEE Transactions on Medical Imaging, 18, 32, 10.1109/42.750253

Thông tin
Thông tin xuất bản

International Journal of Biomedical Imaging

Tập 2011

1-15

Thông tin tác giả