## Nonparametric Permutation Tests for Functional Neuroimaging: A Primer with Examples. Human Brain Mapping (2001)

Citations: | 146 - 6 self |

### BibTeX

@MISC{Nichols01nonparametricpermutation,

author = {T. E. Nichols and A. P. Holmes},

title = {Nonparametric Permutation Tests for Functional Neuroimaging: A Primer with Examples. Human Brain Mapping},

year = {2001}

}

### Years of Citing Articles

### OpenURL

### Abstract

The statistical analyses of functional mapping experiments usually proceeds at the voxel level, involving the formation and assessment of a statistic image: at each voxel a statistic indicating evidence of the experimental effect of interest, at that voxel, is computed, giving an image of statistics, a statistic

### Citations

557 |
Statistical parametric maps in functional imaging: a general linear approach
- Friston, Holmes, et al.
- 1995
(Show Context)
Citation Context ... Bullmore et al. (2001) use a wavelet transformation to account for more general forms of fMRI correlation. Locascio et al. (1997) describe an application to f mri combining the general linear model (=-=Friston et al., 1995-=-b), arma modelling (Bullmore et al., 1996), and a multiple comparisons permutation procedure (Holmes et al., 1996). (Liu et al., 1998) consider an alternative approach, permuting labels. Bullmore et a... |

314 | The Design of Experiments - Fisher - 1935 |

213 | A unified statistical approach for determining significant voxels in images of cerebral activation
- Worsley, Marrett, et al.
- 1996
(Show Context)
Citation Context ...d for statistical significance, using distributional results for continuous random fields to identify voxels or regions where there is significant evidence against the null hypothesis (Worsley, 1996; =-=Worsley et al., 1995-=-; Friston et al., 1994; Poline et al., 1997; Friston et al., 1996). Holmes et al. (1996) introduced a nonparametric alternative based on permutation test theory. This method is conceptually simple, re... |

201 | Thresholding of statistical maps in functional neuroimaging using the false discovery rate
- Genovese, Lazar, et al.
- 2002
(Show Context)
Citation Context ...e chance of any false positives, another perspective has recently been introduced. The new approach controls the False Discovery Rate (FDR), the fraction of false positives among all detected voxels (=-=Genovese et al., 2001-=-) (see Chapter 15 for a brief description). While this chapter only considers the familywise error rate, we note that a permutation approach to FDR has been proposed (Yekutieli & Benjamini, 1999). We ... |

158 |
Resampling-based Multiple Testing: Examples and Methods for p-Value Adjustment
- Westfall, Young
- 1993
(Show Context)
Citation Context ...996); (Cao, 1999) . . . amongst others). In contrast to these parametric and simulation based methods, a nonparametric resampling based approach provides an intuitive and easily implemented solution (=-=Westfall & Young, 1993-=-). The key realization is that the reasoning presented above for permutation tests at a single voxel rely on relabelling entire images, so the arguments can be extended to image level inference by con... |

150 |
Permutation Tests- A Practical Guide to Resampling Methods for Testing Hypotheses
- Good
- 2000
(Show Context)
Citation Context ...been assigned to scans. Formally, a set of labels on the data (still to be collected) are exchangeable if the distribution of the statistic (still to be evaluated) is the same whatever the labelling (=-=Good, 1994-=-). For our activation example, we would use a statistic expressing the difference between the “active” and “baseline” scans. Thus under the null hypothesis of no difference between the a & b condition... |

146 | Multiple Comparison Procedures - Hochberg, Tamhane - 1987 |

144 | three dimensional statistical analysis for cbf activation studies in human brain - Evans, Yarrett, et al. - 1992 |

122 |
Assessing the significance of focal activations using their spatial extent
- Friston, Worsley, et al.
- 1993
(Show Context)
Citation Context ...ificance, using distributional results for continuous random fields to identify voxels or regions where there is significant evidence against the null hypothesis (Worsley, 1996; Worsley et al., 1995; =-=Friston et al., 1994-=-; Poline et al., 1997; Friston et al., 1996). Holmes et al. (1996) introduced a nonparametric alternative based on permutation test theory. This method is conceptually simple, relies only on minimal a... |

109 | Local maxima and the expected Euler characteristic of excursion sets of χ 2 - Worsley - 1994 |

107 | Human Brain Function - Frackowiak, Friston, et al. - 2003 |

96 | Analysis of fMRI time-series revisited— again - Worsley, Friston - 1995 |

92 |
Randomization Tests
- Edgington
- 1995
(Show Context)
Citation Context ...n a test that is more conservative and less powerful than one using the full permutation distribution. 19Fortunately, as few as 1000 permutations can yield an effective approximate permutation test (=-=Edgington, 1969-=-a). However, for an approximate test with minimal loss of power in comparison to the full test (i.e. with high efficiency), one should consider rather more (Jöckel, 1686), as many as 10,000. Power Fre... |

84 |
Nonparametric analysis of statistic images from functional mapping experiments
- HOLMES, BLAIR, et al.
- 1996
(Show Context)
Citation Context ...io et al. (1997) describe an application to f mri combining the general linear model (Friston et al., 1995b), arma modelling (Bullmore et al., 1996), and a multiple comparisons permutation procedure (=-=Holmes et al., 1996-=-). (Liu et al., 1998) consider an alternative approach, permuting labels. Bullmore et al. (1999) apply nonparametric methods to compare groups of structural MR images. The aim of this chapter is to pr... |

73 | Comparing functional (PET) images: the assessment of significant change - Friston, Frith, et al. - 1991 |

68 |
Modified Randomization Tests for Nonparametric Hypotheses
- Dwass
- 1957
(Show Context)
Citation Context ...llings is also problematic, due to the computations involved. In situations where it is not feasible to compute the statistic images for all the relabellings, a subsample of relabellings can be used (=-=Dwass, 1957-=-) (see also Edgington (1969a) for a less mathematical description). The set of N possible relabellings is reduced to a more manageable N ′ consisting of the true labelling and N ′ − 1 randomly chosen ... |

66 | Colored noise and computational inference in neurophysiological (fMRI) time series analysis: resampling methods in time and wavelet - Bullmore, Long, et al. - 2001 |

65 | Detecting activations in PET and fMRI: Levels of inference and power
- Friston, Holmes, et al.
- 1996
(Show Context)
Citation Context ...continuous random fields to identify voxels or regions where there is significant evidence against the null hypothesis (Worsley, 1996; Worsley et al., 1995; Friston et al., 1994; Poline et al., 1997; =-=Friston et al., 1996-=-). Holmes et al. (1996) introduced a nonparametric alternative based on permutation test theory. This method is conceptually simple, relies only on minimal assumptions, deals with the multiple compari... |

59 | Global, voxel, and cluster tests, by theory and permutation, for a difference between two groups of structural MR images of the brain
- BULLMORE, SUCKLING, et al.
- 1999
(Show Context)
Citation Context ...onsider the exceedance mass, the excess mass of the suprathreshold cluster, defined as the integral of the statistic image above the primary threshold within the suprathreshold cluster (Holmes, 1994; =-=Bullmore et al., 1999-=-). Calculation of the permutation distribution and p-values proceeds exactly as before. 152.5 Considerations Before turning to example applications of the nonparametric permutation tests described ab... |

55 |
Combining spatial extent and peak intensity to test for activations in functional imaging
- Poline, Worsley, et al.
- 1999
(Show Context)
Citation Context ...butional results for continuous random fields to identify voxels or regions where there is significant evidence against the null hypothesis (Worsley, 1996; Worsley et al., 1995; Friston et al., 1994; =-=Poline et al., 1997-=-; Friston et al., 1996). Holmes et al. (1996) introduced a nonparametric alternative based on permutation test theory. This method is conceptually simple, relies only on minimal assumptions, deals wit... |

54 | Nonparametric permutation tests for functional neuroimaging: a primer with examples. Human Brain Mapping 2002;15(1):1–25 - TE, AP |

53 |
Generalisability, random effects and population inference. NeuroImage 7
- Holmes, Friston
- 1998
(Show Context)
Citation Context ...ondition interaction term. Voxel-level Statistic: A random effects analysis is easily effected by collapsing the data within subject and computing the statistic across subjects (Worsley et al., 1991; =-=Holmes & Friston, 1999-=-). In this case the result is a repeated measures t-statistic after proportional scaling global flow normalization: Each scan is proportionally scaled to a common global mean of 50; each subjects data... |

33 | Time series analysis in the time domain and resampling methods for studies of functional magnetic resonance brain imaging - Locascio, Jennings, et al. - 1997 |

27 |
A functional neuroanatomy of hallucinations in schizophrenia. Nature 37'8
- Silbersweig, Stern, et al.
- 1995
(Show Context)
Citation Context ...ging transient, randomly occurring events, specifically events that were shorter than the duration of a pet scan. This work was the foundation for later work imaging hallucinations in schizophrenics (=-=Silbersweig et al., 1995-=-). We consider one subject from the study, who was scanned 12 times. During each scan the subject was presented with brief auditory stimuli. The proportion of each scan over which stimuli were deliver... |

23 | Analysis of individual positron emission tomography activation maps by detection of high signal-to-noise-ratio pixel clusters - Poline, Mazoyer - 1993 |

22 |
Order information in working memory: fMRI evidence for parietal and prefrontal mechanisms
- Marshuetz, Jonides, et al.
- 2000
(Show Context)
Citation Context ...els in 8 distinct regions; in particular there is a pair of bilateral posterior parietal regions, a left thalamus region and an anterior cingulate region; these are typical of working memory studies (=-=Marshuetz et al., 2000-=-). It is informative to compare this result to the traditional t statistic, using both a nonparametric and parametric approach to obtain corrected thresholds. We reran this nonparametric analysis usin... |

22 | Investigation of low frequency drift in fMRI signal
- Smith, Lewis, et al.
- 1999
(Show Context)
Citation Context ...parametric results. 34Before discussing the details of this example, we note that fMRI data presents a special challenge for nonparametric methods. Since fMRI data exhibits temporal autocorrelation (=-=Smith et al., 1999-=-), an assumption of exchangeability of scans within subject is not tenable. However, to analyze a group of subjects for population inference we need only assume exchangeability of subjects. Hence, whi... |

20 | Three-dimensional analysis of clustered voxels in 15O-butanol brain activation images - Roland, Levin, et al. - 1993 |

17 |
The geometry of random images
- Worsley
- 1996
(Show Context)
Citation Context ...is then assessed for statistical significance, using distributional results for continuous random fields to identify voxels or regions where there is significant evidence against the null hypothesis (=-=Worsley, 1996-=-; Worsley et al., 1995; Friston et al., 1994; Poline et al., 1997; Friston et al., 1996). Holmes et al. (1996) introduced a nonparametric alternative based on permutation test theory. This method is c... |

15 |
Statistical issues in functional brain mapping
- Holmes
- 1994
(Show Context)
Citation Context ... could be to consider the exceedance mass, the excess mass of the suprathreshold cluster, defined as the integral of the statistic image above the primary threshold within the suprathreshold cluster (=-=Holmes, 1994-=-; Bullmore et al., 1999). Calculation of the permutation distribution and p-values proceeds exactly as before. 152.5 Considerations Before turning to example applications of the nonparametric permuta... |

9 |
A linear spatial correlation model, with applications to positron emission tomography
- Worsley, Evans, et al.
- 1991
(Show Context)
Citation Context ... a random subject by condition interaction term. Voxel-level Statistic: A random effects analysis is easily effected by collapsing the data within subject and computing the statistic across subjects (=-=Worsley et al., 1991-=-; Holmes & Friston, 1999). In this case the result is a repeated measures t-statistic after proportional scaling global flow normalization: Each scan is proportionally scaled to a common global mean o... |

8 |
The size of the connected components of the excursion sets
- Cao
- 1999
(Show Context)
Citation Context ...occupied many authors (Friston et al. (1991); Worsley et al. (1992; 1994; 1995); Poline et al. (1993; 1997); Roland et al. (1993); Forman et al. (1995); (Friston et al., 1994; Friston et al., 1996); (=-=Cao, 1999-=-) . . . amongst others). In contrast to these parametric and simulation based methods, a nonparametric resampling based approach provides an intuitive and easily implemented solution (Westfall & Young... |

7 | Reliability of PET activation across statistical methods, subject groups, and sample sizes - Grabowski, Frank, et al. - 1996 |

2 | Performance of randomization test for single-subject 15-O-water PET activation studies. J Cereb Blood Flow Metab 17:1033–1039 - Halber, Herholz, et al. - 1997 |

2 |
An estimator and permutation test for single-trial fMRI data
- Liu, Raz, et al.
- 1998
(Show Context)
Citation Context ...be an application to f mri combining the general linear model (Friston et al., 1995b), arma modelling (Bullmore et al., 1996), and a multiple comparisons permutation procedure (Holmes et al., 1996). (=-=Liu et al., 1998-=-) consider an alternative approach, permuting labels. Bullmore et al. (1999) apply nonparametric methods to compare groups of structural MR images. The aim of this chapter is to present the theory of ... |

2 | Imaging transient, randomly occurring neuropsychological events in single subjects with positron emission tomography: an event-related count rate correlational analysis - Silbersweig, Stern, et al. - 1994 |

1 |
Statistical Inference: The Distribution Free
- Edgington
- 1995
(Show Context)
Citation Context ...rces, it is possible that large areas of parametric statistics would have gone undeveloped! Modern texts on the subject include Good’s Permutation Tests (Good, 1994), Edgington’s Randomization Tests (=-=Edgington, 1995-=-), and Manly’s Randomization, Bootstrap and MonteCarlo Methods in Biology (1997). Recent interest in more general resampling methods, such as the bootstrap, has further contributed to the field. For a... |

1 | to the editor, with reply - Holmes, Watson, et al. - 1998 |

1 | Randomization, Bootstrap and Monte-Carlo Methods in Biology - Manly - 1997 |

1 | Signficance tests which may be applied to samples from any population - 1937a |

1 | Signficance tests which may be applied to samples from any population. II. The correlation coefficient test - 1937b |

1 | Signficance tests which may be applied to samples from any population. III. The analysis of variance test - 1937c |

1 | Area V5 of the human brain: 47 from a combined study using positron emission tomography and magnetic resonance imaging - Watson, Myers, et al. - 1993 |