Effect Estimation & Testing

1

Effect Estimation& Testing

Thomas Nichols, Ph.D.Assistant Professor

Department of Biostatistics

http://www.sph.umich.edu/~nichols

fMRI CourseOHBM 2004

2

Outline

• Data Modeling– General Linear Model – GLM Issues

• Statistical Inference– Statistic Images & Hypothesis Testing– Multiple Testing Problem

3

Basic fMRI Example

• Data at one voxel– Rest vs.

passive word listening

• Is there an effect?

4

A Linear Model

IntensityT

ime = 1 2+ + er

ror

x1 x2

• “Linear” in parameters 1 & 2

5

Linear model, in image form…

= + +1 2

Y 11x 22x

6

… in image matrix form…

= +

2

1

Y X

7

… in matrix form.

XY

=

+YY X

N

1

N N

1 1p

p

N: Number of scans, p: Number of regressors

GeneralLinear Model

Really general–Correlation

–ANOVA

–ANCOVA

8

Linear Model Issues

• Signal Predictors– Block– Event-related

• Nuisance Predictors– Drift – Motion parameters

• Autocorrelation

• Random effects

9

Temporal AutocorrelationIn Brief

Advantage Disadvantage Software

Indep. Simple Inflated significance

All

Precoloring Avoids autocorr. est.

Statistically inefficient

SPM99

Whitening Statistically optimal

Requires precise autocorr. est.

FSL, SPM2

10

Random Effects Models

• GLM has only one source of randomness

– Residual error

• But people are another source of error– Everyone activates somewhat differently…

XY

11

Subj. 1

Subj. 2

Subj. 3

Subj. 4

Subj. 5

Subj. 6

0

Fixed vs.RandomEffects

• Fixed Effects– Intra-subject

variation suggests all these subjects different from zero

• Random Effects– Intersubject

variation suggests population not very different from zero

Distribution of each subject’s estimated effect

Distribution of population effect

12

Random Effects for fMRI• Summary Statistic Approach

– Easy• Create contrast images for each subject• Analyze contrast images with one-sample t

– Limited• Only allows one scan per subject• Assumes balanced designs and homogeneous meas. error.

• Full Mixed Effects Analysis– Harder

• Requires iterative fitting• REML to estimate inter- and intra subject variance

– SPM2 & FSL3 implement this differently

– Very flexible

13

Random Effects for fMRIRandom vs. Fixed

• Fixed isn’t “wrong”, just usually isn’t of interest• If it is sufficient to say

“I can see this effect in this cohort”then fixed effects are OK

• If need to say“If I were to sample a new cohort from the population I would get the same result”

then random effects are needed

14

Building Statistic Images

• Contrast– A linear combination

of parameters– Truth: c’ Estimate:

T =

contrast ofestimated

parameters

varianceestimate

T =

ss22c’(X’X)c’(X’X)-1-1cc

c’ = 1 0 0 0 0 0 0 0

...ˆˆˆˆˆ54321

’c

’c

16

Hypothesis Testing

• Assume Null Hypothesis of no signal

• Given that there is nosignal, how likely is our measured T?

• P-value measures this– Probability of obtaining T

as large or larger

level– Acceptable false positive rate

P-val

T

17

Hypothesis Testing in fMRI

• Massively Univariate Modeling– Fit model at each voxel– Create statistic images of effect

• Which of 100,000 voxels are significant? =0.05 5,000 false positives!

t > 0.5 t > 1.5 t > 2.5 t > 3.5 t > 4.5 t > 5.5 t > 6.5

18

MCP Solutions:Measuring False Positives

• Familywise Error Rate (FWER)– Familywise Error

• Existence of one or more false positives

– FWER is probability of familywise error

• False Discovery Rate (FDR)– R voxels declared active, V falsely so

• Observed false discovery rate: V/R

– FDR = E(V/R)

19

FWER MCP Solutions

• Bonferroni

• Maximum Distribution Methods– Random Field Theory– Permutation

20

FWER MCP Solutions

• Bonferroni

• Maximum Distribution Methods– Random Field Theory– Permutation

21

FWER MCP Solutions: Controlling FWER w/ Max

• FWER & distribution of maximum

FWER= P(FWE)= P(One or more voxels u |

Ho)= P(Max voxel u | Ho)

• 100(1-)%ile of max distn controls FWERFWER = P(Max voxel u | Ho)

u

22

FWER MCP Solutions:Random Field Theory

• Euler Characteristic u

– Topological Measure• #blobs - #holes

– At high thresholds,just counts blobs

– FWER = P(Max voxel u | Ho)= P(One or more blobs | Ho) P(u 1 | Ho) E(u | Ho)

Random Field

Suprathreshold Sets

Threshold

23

Random Field Intuition

• Corrected P-value for voxel value t Pc = P(max T > t)

E(t) () ||1/2 t2 exp(-t2/2)

• Statistic value t increases– Pc decreases (of course!)

• Search volume increases– Pc increases (more severe MCP)

• Smoothness increases (||1/2 smaller)– Pc decreases (less severe MCP)

24

Random Field TheoryStrengths & Weaknesses

• Closed form results for E(u)– Z, t, F, Chi-Squared Continuous RFs

• Results depend only on volume & smoothness

• Smoothness assumed known• Sufficient smoothness required

– Results are for continuous random fields

• Multivariate normality• Several layers of approximations

Lattice ImageData

Continuous Random Field

25

FWER MCP Solutions

• Bonferroni

• Maximum Distribution Methods– Random Field Theory– Permutation

26

Nonparametric Permutation Test

• Parametric methods– Assume distribution of

statistic under nullhypothesis

• Nonparametric methods– Use data to find

distribution of statisticunder null hypothesis

– Any statistic!

5%

Parametric Null Distribution

5%

Nonparametric Null Distribution

27

Controlling FWER: Permutation Test

• Parametric methods– Assume distribution of

max statistic under nullhypothesis

• Nonparametric methods– Use data to find

distribution of max statisticunder null hypothesis

– Any max statistic!

5%

Parametric Null Max Distribution

5%

Nonparametric Null Max Distribution

28

Permutation TestStrengths

• Requires only assumption of exchangeability– Under Ho, distribution unperturbed by permutation

• Subjects are exchangeable– Under Ho, each subject’s A/B labels can be flipped

• fMRI scans not exchangeable under Ho– Due to temporal autocorrelation– Need to de-correlate, then permute

(Brammer, Bullmore et al, 1997)

29

Permutation TestLimitations

• Computational Intensity– Analysis repeated for each relabeling– Not so bad on modern hardware

• No analysis discussed below took more than 3 hours

• Implementation Generality– Each experimental design type needs unique

code to generate permutations• Not so bad for population inference with t-tests

30

Measuring False Positives

• Familywise Error Rate (FWER)– Familywise Error

• Existence of one or more false positives

– FWER is probability of familywise error

• False Discovery Rate (FDR)– R voxels declared active, V falsely so

• Observed false discovery rate: V/R

– FDR = E(V/R)

31

False Discovery RateIllustration:

Signal

Signal+Noise

Noise

32

FWE

6.7% 10.4% 14.9% 9.3% 16.2% 13.8% 14.0% 10.5% 12.2% 8.7%

Control of Familywise Error Rate at 10%

11.3% 11.3% 12.5% 10.8% 11.5% 10.0% 10.7% 11.2% 10.2% 9.5%

Control of Per Comparison Rate at 10%

Percentage of Null Pixels that are False Positives

Control of False Discovery Rate at 10%

Occurrence of Familywise Error

Percentage of Activated Pixels that are False Positives

33

Controlling FDR:Benjamini & Hochberg

• Select desired limit q on E(FDR)• Order p-values, p(1) p(2) ... p(V)

• Let r be largest i such that

• Reject all hypotheses corresponding to p(1), ... , p(r).p(i) i/V q p(i)

i/V

i/V qp-

valu

e

0 1

01

34

Example – Working Memory

• fMRI Study of Working Memory – 12 subjects, block design Marshuetz et al (2000)

– Item Recognition• Active:View five letters, 2s pause,

view probe letter, respond

• Baseline: View XXXXX, 2s pause,view Y or N, respond

• Second Level RFX– Difference image, A-B constructed

for each subject

– One sample t test

...

D

yes

...

UBKDA

Active

...

N

no

...

XXXXX

Baseline

Skip

35

Example – Working MemoryRFT Result

• Threshold– S = 110,776– 2 2 2 voxels

5.1 5.8 6.9 mmFWHM

– u = 9.870

• Result– 5 voxels above

the threshold

-log 1

0 p

-va

lue

36

Example – Working MemoryNon-Parametric Result

• Threshold– u = 7.67

• Result– 58 voxels above

the threshold

-log 1

0 p

-va

lue

Permutation Distribution Maximum t

37

Example – Working MemoryFDR Result

• FDR Threshold– u = 3.83

• Result– 3,073 voxels above

threshold

38

Conclusions

• Must account for multiple comparisons• FWER

– Random Field Theory• Simple to apply, but heavy on assumptions

– Nonparametric• Exact, but requires more computation

• FDR– More lenient measure of false positives – more powerful

– Sociological calibration still underway (5%? 1%? 0.1%?)

39

Thanks

• Slide help– Stefan Keibel, Rik Henson, JB Poline, Andrew

Holmes