Practical aspects of…
Statistical Analysis of M/EEG Sensor- and
Source-Level Data
Jason Taylor
MRC Cognition and Brain Sciences Unit (CBU)
Cambridge Centre for Ageing and Neuroscience (CamCAN)
Cambridge, UK
jason.taylor@mrc-cbu.cam.ac.uk | http://imaging.mrc-cbu.cam.ac.uk/meg (wiki)
14 May 2012 | London | Thanks to Rik Henson and Dan Wakeman @ CBU
Overview
The SPM approach to M/EEG Statistics:
A mass-univariate statistical approach to inference regarding
effects in space/time/frequency (using replications across
trials or subjects).
•
In Sensor Space:
2D Time-Frequency
3D Topography-by-Time
•
In Source Space:
3D Time-Frequency contrast images
4-ish-D space by (factorised) time
Alternative approaches: SnPM, PPM, etc.
Example Datasets
CTF Multimodal Face-Evoked Dataset | SPM8 Manual (Rik Henson)
CTF MEG/ActiveTwo EEG:
275 MEG
128 EEG
3T fMRI (with nulls)
1mm3 sMRI
2 sessions
Stimuli:
~160 face trials/session
~160 scrambled trials/session
Group:
N=12 subjects
(Henson et al, 2009 a,b,c)
Chapter 37 SPM8 Manual
Neuromag Faces Dataset | Dan Wakeman & Rik Henson @ CBU
Neuromag MEG & EEG Data:
102 Magnetometers
204 Planar Gradiometers
70 EEG
HEOG, VEOG, ECG
1mm3 sMRI
400600ms
8001000ms
1700ms
6 sessions
Stimuli:
faces & scrambled-face images
(480 trials total)
Group:
N=18
(Henson et al, 2011)
Biomag 2010 Award-Winning Dataset!
Neuromag Lexical Decision Dataset | Taylor & Henson @ CBU
Neuromag MEG Data:
102 Magnetometers
204 Planar Gradiometers
HEOG, VEOG, ECG
1mm3 sMRI
4 sessions
Stimuli:
words & pseudowords
presented visually
(480 trials total)
Group:
N=18
Taylor & Henson (submitted)
Neuromag Lexical Decision
The Multiple Comparison Problem
Where is the effect?: The curse of too much data
The Multiple Comparison Problem
•
The more comparisons we conduct, the more Type I errors (false
positives) we will make when the Null Hypothesis is true.
* Must consider Familywise (vs. per-comparison) Error Rate
•
Comparisons are often made implicitly, e.g., by viewing (“eyeballing”) data before selecting a time-window or set of channels for
statistical analysis. RFT correction depends on the search volume.
-> When is there
an effect in time
e.g., GFP (1D)?
time
The Multiple Comparison Problem
•
The more comparisons we conduct, the more Type I errors (false
positives) we will make when the Null Hypothesis is true.
* Must consider Familywise (vs. per-comparison) Error Rate
Comparisons are often made implicitly, e.g., by viewing (“eyeballing”) data before selecting a time-window or set of channels for
statistical analysis. RFT correction depends on the search volume.
-> When/at what frequency
is there an effect
an effect in time/frequency
(2D)?
frequency
•
time
The Multiple Comparison Problem
•
The more comparisons we conduct, the more Type I errors (false
positives) we will make when the Null Hypothesis is true.
* Must consider Familywise (vs. per-comparison) Error Rate
•
Comparisons are often made implicitly, e.g., by viewing (“eyeballing”) data before selecting a time-window or set of channels for
statistical analysis. RFT correction depends on the search volume.
x
-> When/where
is there an effect
in sensor-topography
space/time (3D)?
y
The Multiple Comparison Problem
•
The more comparisons we conduct, the more Type I errors (false
positives) we will make when the Null Hypothesis is true.
* Must consider Familywise (vs. per-comparison) Error Rate
•
Comparisons are often made implicitly, e.g., by viewing (“eyeballing”) data before selecting a time-window or set of channels for
statistical analysis. RFT correction depends on the search volume.
-> When/where
is there an effect
in source
space/time
x
(4-ish-D)?
z
y
The Multiple Comparison Problem
•
The more comparisons we conduct, the more Type I errors (false
positives) we will make when the Null Hypothesis is true.
* Must consider Familywise (vs. per-comparison) Error Rate
•
Comparisons are often made implicitly, e.g., by viewing (“eyeballing”) data before selecting a time-window or set of channels for
statistical analysis. RFT correction depends on the search volume.
Random Field Theory (RFT) is a method for correcting for multiple
statistical comparisons with N-dimensional spaces (for parametric
statistics, e.g., Z-, T-, F- statistics).
* Takes smoothness of images into account
Worsley Et Al (1996). Human Brain Mapping, 4:58-73
GLM: Condition Effects after removing variance due
to confounds
Each trial-type (6)
Confounds (4)
Each trial
W/in Subject
(1st-level) model
-> one image per trial
-> one covariate value
per trial
Also: Group Analysis
(2nd-level)
-> one image per subject
per condition
-> one covariate value per
subject per condition
beta_00* image volumes reflect
(adjusted) condition effects
Henson et al., 2008, NImage
Sensor-space analyses
Where is an effect in time-frequency space?
1 subject (1st-level analysis)
1 MEG channel
Faces > Scrambled
Faces
Morlet wavelet projection
Scrambled
1 t-x-f image per trial
Kilner et al., 2005, Nsci Letters
CTF Multimodal Faces
Where is an effect in time-frequency space?
18 subjects
1 channel
(2nd-level
analysis)
Faces > Scrambled
EEG
MEG (Mag)
6
6
Freq (Hz)
Freq (Hz)
90
90
100-220ms
8-18Hz
-500
+1000
Time (ms)
-500
+1000
Time (ms)
Neuromag Faces
Where is an effect in sensor-time space?
1 subject
ALL EEG
channels
3D
Topographyx-Time
Image
volume
(EEG)
time
Each trial,
Topographic
projection (2D) for
each time point (1D)
= topo-time image
volume (3D)
Faces vs
Scrambled
y
x
CTF Multimodal Faces
Where is an effect in sensor-time space?
Analysis over subjects (2nd Level)
NOTE for MEG: Complicated by variability in head-position
SOLUTION(?): Virtual transformation to same position (sessions, subjects)
- using Signal Space Separation (SSS; Taulu et al, 2005; Neuromag systems)
Without transformation to Device Space
With transformation to Device Space
Stats over 18 subjects, RMS of planar gradiometers
Improved: (i.e., more blobs, larger T values) w/ head-position transform
Other evidence: ICA ‘splitting’ with movement
Taylor & Henson (2008) Biomag
Neuromag Lexical Decision
Where is an effect in sensor-time space?
Analysis over subjects (2nd Level):
Magnetometers: Pseudowords vs Words PPM (P>.95 effect>0)
Effect Size
18
fT
-18
Taylor & Henson (submitted)
Neuromag Lexical Decision
Source-space analyses
Where is an effect in source space (3D)?
STEPS:
Estimate evoked/induced
energy (RMS) at each dipole for
a certain time-frequency window
of interest.
- e.g., 100-220ms, 8-18 Hz
- For each condition
(Faces, Scrambled)
- For each sensor type
OR fused modalities
Write data to 3D image
- in MNI space
- smooth along 2D surface
Smooth by 3D Gaussian
Submit to GLM
Henson et al., 2007, NImage
Where is an effect in source space (3D)?
RESULTS: Faces > Scrambled
fMRI
EEG
MEG
sensor fusion
E+MEG
Henson et al, 2011
Neuromag Faces
Where is an effect in source space (3D)?
RESULTS: Faces > Scrambled
fMRI
with group optimised
fMRI priors
EEG+MEG+fMRI
with fMRI priors
EEG+MEG+fMRI
mean
stats
Henson et al, 2011
Neuromag Faces
Where and When do effects emerge/disappear
in source space (4-ish-D: time factorised)?
Condition x Time-window
Interactions
Factorising time allows
you to infer (rather than
simply describe) when
effects emerge or disappear.
We've used a data-driven
method (hierarchical
cluster analysis) in sensor
space to define timewindows of interest for
source-space analyses.
* high-dimensional distance
between topography vectors
* cut the cluster tree where the
solution is stable over longest
range of distances
Taylor & Henson, submitted
Neuromag Lexical Decision
Where and When do effects emerge/disappear
in source space (4-ish-D)?
Condition x Time-window
Interactions
Factorising time allows you
to infer (rather than simply
describe) when effects
emerge or disappear.
source
timecourses
We've used a data-driven
method (hierarchical cluster
analysis) in sensor space to
define time-windows of
interest for source-space
analyses.
* estimate sources in each
sub-time-window
* submit to GLM with conditions
& time-windows as factors
(here: Cond effects per t-win)
Taylor & Henson, submitted
Neuromag Lexical Decision
Where and When do effects emerge/disappear
in source space (4-ish-D)?
Condition x Time-window
Interactions
Factorising time allows you
to infer (rather than simply
describe) when effects
emerge or disappear.
We've used a data-driven
method (hierarchical cluster
analysis) in sensor space to
define time-windows of
interest for source-space
analyses.
* estimate sources in each subtime-window
* submit to GLM with conditions
& time-windows as factors:
Cond x TW interactions
Taylor & Henson, submitted
Neuromag Lexical Decision
Alternative Approaches
Alternative Approaches
Classical SPM approach: Caveats
Inverse operator induces long-range error correlations (e.g., similar
gain vectors from non-adjacent dipoles with similar orientation),
making RFT correction conservative
Distributions over subjects/voxels may not be Gaussian (e.g., if
sparse, as in MSP)
(Taylor & Henson, submitted; Biomag 2010)
Alternative Approaches
Non-Parametric Approach
(SnPM)
p<.05 FWE
Robust to non-Gaussian
distributions
Less conservative than
RFT when dfs<20
Caveats:
No idea of effect size (e.g.,
for power, future expts)
Exchangeability difficult for
more complex designs
(Taylor & Henson, Biomag 2010)
SnPM Toolbox by Holmes & Nichols:
http://go.warwick.ac.uk/tenichols/software/snpm/
CTF Multimodal Faces
Alternative Approaches
Posterior Probability
Maps (PPMs)
p>.95 (γ>1SD)
Bayesian Inference
No need for RFT (no MCP)
Threshold on posterior
probabilty of an effect
greater than some size
Can show effect size after
thresholding
Caveats:
Assume Gaussian
distribution (e.g., of mean
over voxels)
CTF Multimodal Faces
Future Directions
• Extend RFT to 2D cortical surfaces (+1D time)?
(e.g., Pantazis et al., 2005, NImage)
• Novel approaches to controlling for multiple
comparisons
(e.g., 'unique extrema' in sensors: Barnes et al., 2011, NImage)
• Go multivariate?
To identify linear combinations of spatial (sensor or
source) effects in time and space
(e.g., Carbonnell, 2004, NImage; but see Barnes et al., 2011)
To detect spatiotemporal patterns in 3D images
(e.g., Duzel, 2004, NImage; Kherif, 2003, NImage)
-- The end --
• Thanks for listening
• Acknowledgements:
• Rik Henson (MRC CBU)
• Dan Wakeman (MRC CBU / now at MGH)
• Vladimir, Karl, and the FIL Methods Group
• More info:
• http://imaging.mrc-cbu.cam.ac.uk/meg (wiki)
jason.taylor@mrc-cbu.cam.ac.uk