basics of fmri analysis: preprocessing, first level ... · 2 overview • neuroanatomy 101 and fmri...
TRANSCRIPT
![Page 1: Basics of fMRI Analysis: Preprocessing, First Level ... · 2 Overview • Neuroanatomy 101 and fMRI Contrast Mechanism • Preprocessing • Hemodynamic Response • “Univariate”](https://reader030.vdocuments.us/reader030/viewer/2022040416/5d49421888c9935f218bd682/html5/thumbnails/1.jpg)
Basics of fMRI Analysis:
Preprocessing, First Level Analysis, and
Group Analysis
![Page 2: Basics of fMRI Analysis: Preprocessing, First Level ... · 2 Overview • Neuroanatomy 101 and fMRI Contrast Mechanism • Preprocessing • Hemodynamic Response • “Univariate”](https://reader030.vdocuments.us/reader030/viewer/2022040416/5d49421888c9935f218bd682/html5/thumbnails/2.jpg)
2
Overview
• Neuroanatomy 101 and fMRI Contrast Mechanism
• Preprocessing
• Hemodynamic Response
• “Univariate” GLM Analysis
• Hypothesis Testing
• Group Analysis (Random, Mixed, Fixed)
![Page 3: Basics of fMRI Analysis: Preprocessing, First Level ... · 2 Overview • Neuroanatomy 101 and fMRI Contrast Mechanism • Preprocessing • Hemodynamic Response • “Univariate”](https://reader030.vdocuments.us/reader030/viewer/2022040416/5d49421888c9935f218bd682/html5/thumbnails/3.jpg)
3
Neuroantomy
• Gray matter
• White matter
• Cerebrospinal Fluid
![Page 4: Basics of fMRI Analysis: Preprocessing, First Level ... · 2 Overview • Neuroanatomy 101 and fMRI Contrast Mechanism • Preprocessing • Hemodynamic Response • “Univariate”](https://reader030.vdocuments.us/reader030/viewer/2022040416/5d49421888c9935f218bd682/html5/thumbnails/4.jpg)
4
Functional Anatomy/Brain Mapping
![Page 5: Basics of fMRI Analysis: Preprocessing, First Level ... · 2 Overview • Neuroanatomy 101 and fMRI Contrast Mechanism • Preprocessing • Hemodynamic Response • “Univariate”](https://reader030.vdocuments.us/reader030/viewer/2022040416/5d49421888c9935f218bd682/html5/thumbnails/5.jpg)
5
Visual Activation Paradigm
Flickering Checkerboard
Visual, Auditory, Motor, Tactile, Pain, Perceptual,
Recognition, Memory, Emotion, Reward/Punishment,
Olfactory, Taste, Gastral, Gambling, Economic, Acupuncture,
Meditation, The Pepsi Challenge, …
• Scientific
• Clinical
• Pharmaceutical
![Page 6: Basics of fMRI Analysis: Preprocessing, First Level ... · 2 Overview • Neuroanatomy 101 and fMRI Contrast Mechanism • Preprocessing • Hemodynamic Response • “Univariate”](https://reader030.vdocuments.us/reader030/viewer/2022040416/5d49421888c9935f218bd682/html5/thumbnails/6.jpg)
6
Magnetic Resonance Imaging
T1-weighted
Contrast
BOLD-weighted
Contrast
![Page 7: Basics of fMRI Analysis: Preprocessing, First Level ... · 2 Overview • Neuroanatomy 101 and fMRI Contrast Mechanism • Preprocessing • Hemodynamic Response • “Univariate”](https://reader030.vdocuments.us/reader030/viewer/2022040416/5d49421888c9935f218bd682/html5/thumbnails/7.jpg)
7
Blood Oxygen Level Dependence (BOLD)
Neurons
Lungs
Oxygen CO2
Oxygenated
Hemoglobin
(DiaMagnetic)
Deoxygenated
Hemoglobin
(ParaMagnetic)
Contrast Agent
![Page 8: Basics of fMRI Analysis: Preprocessing, First Level ... · 2 Overview • Neuroanatomy 101 and fMRI Contrast Mechanism • Preprocessing • Hemodynamic Response • “Univariate”](https://reader030.vdocuments.us/reader030/viewer/2022040416/5d49421888c9935f218bd682/html5/thumbnails/8.jpg)
8
Functional MRI (fMRI)
Localized
Neural
Firing
Localized
Increased
Blood Flow
Stimulus
Localized
BOLD
Changes
Sample BOLD response in 4D
Space (3D) – voxels (64x64x35, 3x3x5mm^3, ~50,000)
Time (1D) – time points (100, 2 sec) – Movie
Time 1 Time 2 Time 3 …
![Page 9: Basics of fMRI Analysis: Preprocessing, First Level ... · 2 Overview • Neuroanatomy 101 and fMRI Contrast Mechanism • Preprocessing • Hemodynamic Response • “Univariate”](https://reader030.vdocuments.us/reader030/viewer/2022040416/5d49421888c9935f218bd682/html5/thumbnails/9.jpg)
9
4D Volume
64x64x35 85x1
![Page 10: Basics of fMRI Analysis: Preprocessing, First Level ... · 2 Overview • Neuroanatomy 101 and fMRI Contrast Mechanism • Preprocessing • Hemodynamic Response • “Univariate”](https://reader030.vdocuments.us/reader030/viewer/2022040416/5d49421888c9935f218bd682/html5/thumbnails/10.jpg)
10
fMRI Analysis Overview
Higher Level GLM
First Level
GLM Analysis
First Level
GLM Analysis
Subject 3
First Level
GLM Analysis
Subject 4
First Level
GLM Analysis
Subject 1
Subject 2
CX
CX
CX
CX
Preprocessing
MC, STC, B0
Smoothing
Normalization
Preprocessing
MC, STC, B0
Smoothing
Normalization
Preprocessing
MC, STC, B0
Smoothing
Normalization
Preprocessing
MC, STC, B0
Smoothing
Normalization
Raw Data
Raw Data
Raw Data
Raw Data
CX
![Page 11: Basics of fMRI Analysis: Preprocessing, First Level ... · 2 Overview • Neuroanatomy 101 and fMRI Contrast Mechanism • Preprocessing • Hemodynamic Response • “Univariate”](https://reader030.vdocuments.us/reader030/viewer/2022040416/5d49421888c9935f218bd682/html5/thumbnails/11.jpg)
11
Preprocessing
• Assures that assumptions of the analysis are met
• Time course comes from a single location
• Uniformly spaced in time
• Spatial “smoothness”
• Analysis – separating signal from noise
![Page 12: Basics of fMRI Analysis: Preprocessing, First Level ... · 2 Overview • Neuroanatomy 101 and fMRI Contrast Mechanism • Preprocessing • Hemodynamic Response • “Univariate”](https://reader030.vdocuments.us/reader030/viewer/2022040416/5d49421888c9935f218bd682/html5/thumbnails/12.jpg)
12
Preprocessing
• Start with a 4D data set
1. Motion Correction
2. Slice-Timing Correction
3. B0 Distortion Correction
4. Spatial Normalization
5. Spatial Smoothing
• End with a 4D data set
• Can be done in other orders
• Not everything is always done
![Page 13: Basics of fMRI Analysis: Preprocessing, First Level ... · 2 Overview • Neuroanatomy 101 and fMRI Contrast Mechanism • Preprocessing • Hemodynamic Response • “Univariate”](https://reader030.vdocuments.us/reader030/viewer/2022040416/5d49421888c9935f218bd682/html5/thumbnails/13.jpg)
13
Motion
• Analysis assumes that time course represents a
value from a single location
• Subjects move
• Shifts can cause noise, uncertainty
• Edge of the brain and tissue boundaries
![Page 14: Basics of fMRI Analysis: Preprocessing, First Level ... · 2 Overview • Neuroanatomy 101 and fMRI Contrast Mechanism • Preprocessing • Hemodynamic Response • “Univariate”](https://reader030.vdocuments.us/reader030/viewer/2022040416/5d49421888c9935f218bd682/html5/thumbnails/14.jpg)
14
Motion and Motion Correction
•Motion correction reduces motion
•Not perfect
Raw Corrected
![Page 15: Basics of fMRI Analysis: Preprocessing, First Level ... · 2 Overview • Neuroanatomy 101 and fMRI Contrast Mechanism • Preprocessing • Hemodynamic Response • “Univariate”](https://reader030.vdocuments.us/reader030/viewer/2022040416/5d49421888c9935f218bd682/html5/thumbnails/15.jpg)
15
Motion Correction
• Motion correction parameters
• Six for each time point
• Sometimes used as nuisance regressors
• How much motion is too much?
![Page 16: Basics of fMRI Analysis: Preprocessing, First Level ... · 2 Overview • Neuroanatomy 101 and fMRI Contrast Mechanism • Preprocessing • Hemodynamic Response • “Univariate”](https://reader030.vdocuments.us/reader030/viewer/2022040416/5d49421888c9935f218bd682/html5/thumbnails/16.jpg)
16
Slice Timing
Ascending Interleaved
• Volume not acquired all at one time
• Acquired slice-by-slice
• Each slice has a different delay
![Page 17: Basics of fMRI Analysis: Preprocessing, First Level ... · 2 Overview • Neuroanatomy 101 and fMRI Contrast Mechanism • Preprocessing • Hemodynamic Response • “Univariate”](https://reader030.vdocuments.us/reader030/viewer/2022040416/5d49421888c9935f218bd682/html5/thumbnails/17.jpg)
• Volume = 30 slices
• TR = 2 sec
• Time for each slice = 2/30 = 66.7 ms
0s
2s
Effect of Slice Delay on Time Course
![Page 18: Basics of fMRI Analysis: Preprocessing, First Level ... · 2 Overview • Neuroanatomy 101 and fMRI Contrast Mechanism • Preprocessing • Hemodynamic Response • “Univariate”](https://reader030.vdocuments.us/reader030/viewer/2022040416/5d49421888c9935f218bd682/html5/thumbnails/18.jpg)
18
Slice Timing Correction
• Temporal interpolation of adjacent time points
• Usually sinc interpolation
• Each slice gets a different interpolation
• Some slices might not have any interpolation
• Can also be done in the GLM
• You must know the slice order!
X
TR1 TR2 TR3
Sli
ce
![Page 19: Basics of fMRI Analysis: Preprocessing, First Level ... · 2 Overview • Neuroanatomy 101 and fMRI Contrast Mechanism • Preprocessing • Hemodynamic Response • “Univariate”](https://reader030.vdocuments.us/reader030/viewer/2022040416/5d49421888c9935f218bd682/html5/thumbnails/19.jpg)
19
B0 Distortion
• Metric (stretching or compressing)
• Intensity Dropout
• A result of a long readout needed to get an entire slice in a
single shot.
• Caused by B0 Inhomogeneity
Stretch
Dropout
![Page 20: Basics of fMRI Analysis: Preprocessing, First Level ... · 2 Overview • Neuroanatomy 101 and fMRI Contrast Mechanism • Preprocessing • Hemodynamic Response • “Univariate”](https://reader030.vdocuments.us/reader030/viewer/2022040416/5d49421888c9935f218bd682/html5/thumbnails/20.jpg)
B0 Map
Echo 2
TE2
Magnitude
Echo 1
TE1
Phase
![Page 21: Basics of fMRI Analysis: Preprocessing, First Level ... · 2 Overview • Neuroanatomy 101 and fMRI Contrast Mechanism • Preprocessing • Hemodynamic Response • “Univariate”](https://reader030.vdocuments.us/reader030/viewer/2022040416/5d49421888c9935f218bd682/html5/thumbnails/21.jpg)
21
Voxel Shift Map
• Units are voxels (3.5mm)
• Shift is in-plane
• Blue = P�A, Red A�P
• Regions affected near air/tissue boundaries
• sinuses
![Page 22: Basics of fMRI Analysis: Preprocessing, First Level ... · 2 Overview • Neuroanatomy 101 and fMRI Contrast Mechanism • Preprocessing • Hemodynamic Response • “Univariate”](https://reader030.vdocuments.us/reader030/viewer/2022040416/5d49421888c9935f218bd682/html5/thumbnails/22.jpg)
22
B0 Distortion Correction
• Can only fix metric distortion
• Dropout is lost forever
![Page 23: Basics of fMRI Analysis: Preprocessing, First Level ... · 2 Overview • Neuroanatomy 101 and fMRI Contrast Mechanism • Preprocessing • Hemodynamic Response • “Univariate”](https://reader030.vdocuments.us/reader030/viewer/2022040416/5d49421888c9935f218bd682/html5/thumbnails/23.jpg)
23
B0 Distortion Correction
• Can only fix metric distortion
• Dropout is lost forever
• Interpolation
• Need:
• “Echo spacing” – readout time
• Phase encode direction
• More important for surface than for volume
• Important when combining from different scanners
![Page 24: Basics of fMRI Analysis: Preprocessing, First Level ... · 2 Overview • Neuroanatomy 101 and fMRI Contrast Mechanism • Preprocessing • Hemodynamic Response • “Univariate”](https://reader030.vdocuments.us/reader030/viewer/2022040416/5d49421888c9935f218bd682/html5/thumbnails/24.jpg)
24
Spatial Normalization
• Transform volume into another volume
• Re-slicing, re-gridding
• New volume is an “atlas” space
• Align brains of different subjects so that a given voxel
represents the “same” location.
• Similar to motion correction
• Preparation for comparing across subjects
• Volume-based
• Surface-based
• Combined Volume-surface-based (CVS)
![Page 25: Basics of fMRI Analysis: Preprocessing, First Level ... · 2 Overview • Neuroanatomy 101 and fMRI Contrast Mechanism • Preprocessing • Hemodynamic Response • “Univariate”](https://reader030.vdocuments.us/reader030/viewer/2022040416/5d49421888c9935f218bd682/html5/thumbnails/25.jpg)
25
Spatial Normalization: Volume
Subject 1
Subject 2
Subject 1
Subject 2
MNI305
MNI152
Native Space MNI305 Space
Affine (12 DOF) Registration
![Page 26: Basics of fMRI Analysis: Preprocessing, First Level ... · 2 Overview • Neuroanatomy 101 and fMRI Contrast Mechanism • Preprocessing • Hemodynamic Response • “Univariate”](https://reader030.vdocuments.us/reader030/viewer/2022040416/5d49421888c9935f218bd682/html5/thumbnails/26.jpg)
26
Subject 1 Subject 2 (Before)
Subject 2 (After) • Shift, Rotate, Stretch
• High dimensional (~500k)
• Preserve metric properties
• Take variance into account
• Common space for group
analysis (like Talairach)
Spatial Normalization: Surface
![Page 27: Basics of fMRI Analysis: Preprocessing, First Level ... · 2 Overview • Neuroanatomy 101 and fMRI Contrast Mechanism • Preprocessing • Hemodynamic Response • “Univariate”](https://reader030.vdocuments.us/reader030/viewer/2022040416/5d49421888c9935f218bd682/html5/thumbnails/27.jpg)
27
Spatial Smoothing
• Replace voxel value with a weighted average of nearby voxels
(spatial convolution)
• Weighting is usually Gaussian
• 3D (volume)
• 2D (surface)
• Do after all interpolation, before computing a standard deviation
• Similarity to interpolation
• Improve SNR
• Improve Intersubject registration
• Can have a dramatic effect on your results
![Page 28: Basics of fMRI Analysis: Preprocessing, First Level ... · 2 Overview • Neuroanatomy 101 and fMRI Contrast Mechanism • Preprocessing • Hemodynamic Response • “Univariate”](https://reader030.vdocuments.us/reader030/viewer/2022040416/5d49421888c9935f218bd682/html5/thumbnails/28.jpg)
Spatial Smoothing
Full-Width/Half-max
• Spatially convolve image with Gaussian kernel.
• Kernel sums to 1
• Full-Width/Half-max: FWHM = σ/sqrt(log(256))
σ = standard deviation of the Gaussian
0 FWHM 5 FWHM 10 FWHM
2mm FWHM
10mm FWHM
5mm FWHM
Full Max
Half Max
![Page 29: Basics of fMRI Analysis: Preprocessing, First Level ... · 2 Overview • Neuroanatomy 101 and fMRI Contrast Mechanism • Preprocessing • Hemodynamic Response • “Univariate”](https://reader030.vdocuments.us/reader030/viewer/2022040416/5d49421888c9935f218bd682/html5/thumbnails/29.jpg)
29
Spatial Smoothing
![Page 30: Basics of fMRI Analysis: Preprocessing, First Level ... · 2 Overview • Neuroanatomy 101 and fMRI Contrast Mechanism • Preprocessing • Hemodynamic Response • “Univariate”](https://reader030.vdocuments.us/reader030/viewer/2022040416/5d49421888c9935f218bd682/html5/thumbnails/30.jpg)
30
Effect of Smoothing on Activation
• Working memory paradigm
• FWHM: 0, 2, 4, 6, 8, 10, 12, 14, 16, 18, 20
![Page 31: Basics of fMRI Analysis: Preprocessing, First Level ... · 2 Overview • Neuroanatomy 101 and fMRI Contrast Mechanism • Preprocessing • Hemodynamic Response • “Univariate”](https://reader030.vdocuments.us/reader030/viewer/2022040416/5d49421888c9935f218bd682/html5/thumbnails/31.jpg)
31
Volume- vs Surface-based Smoothing
• 5 mm apart in 3D
• 25 mm apart on surface
• Averaging with other tissue
types (WM, CSF)
• Averaging with other
functional areas
14mm FWHM
![Page 32: Basics of fMRI Analysis: Preprocessing, First Level ... · 2 Overview • Neuroanatomy 101 and fMRI Contrast Mechanism • Preprocessing • Hemodynamic Response • “Univariate”](https://reader030.vdocuments.us/reader030/viewer/2022040416/5d49421888c9935f218bd682/html5/thumbnails/32.jpg)
32
Preprocessing
• Start with a 4D data set
1. Motion Correction - Interpolation
2. Slice-Timing Correction
3. B0 Distortion Correction - Interpolation
4. Spatial Normalization - Interpolation
5. Spatial Smoothing – Interpolation-like
• End with a 4D data set
• Can be done in other orders
• Not all are done
![Page 33: Basics of fMRI Analysis: Preprocessing, First Level ... · 2 Overview • Neuroanatomy 101 and fMRI Contrast Mechanism • Preprocessing • Hemodynamic Response • “Univariate”](https://reader030.vdocuments.us/reader030/viewer/2022040416/5d49421888c9935f218bd682/html5/thumbnails/33.jpg)
fMRI Time-Series Analysis
![Page 34: Basics of fMRI Analysis: Preprocessing, First Level ... · 2 Overview • Neuroanatomy 101 and fMRI Contrast Mechanism • Preprocessing • Hemodynamic Response • “Univariate”](https://reader030.vdocuments.us/reader030/viewer/2022040416/5d49421888c9935f218bd682/html5/thumbnails/34.jpg)
34
fMRI Analysis Overview
Higher Level GLM
First Level
GLM Analysis
First Level
GLM Analysis
Subject 3
First Level
GLM Analysis
Subject 4
First Level
GLM Analysis
Subject 1
Subject 2
CX
CX
CX
CX
Preprocessing
MC, STC, B0
Smoothing
Normalization
Preprocessing
MC, STC, B0
Smoothing
Normalization
Preprocessing
MC, STC, B0
Smoothing
Normalization
Preprocessing
MC, STC, B0
Smoothing
Normalization
Raw Data
Raw Data
Raw Data
Raw Data
CX
![Page 35: Basics of fMRI Analysis: Preprocessing, First Level ... · 2 Overview • Neuroanatomy 101 and fMRI Contrast Mechanism • Preprocessing • Hemodynamic Response • “Univariate”](https://reader030.vdocuments.us/reader030/viewer/2022040416/5d49421888c9935f218bd682/html5/thumbnails/35.jpg)
35
Visual/Auditory/Motor Activation Paradigm
15 sec ‘ON’, 15 sec ‘OFF’
• Flickering Checkerboard
• Auditory Tone
• Finger Tapping
![Page 36: Basics of fMRI Analysis: Preprocessing, First Level ... · 2 Overview • Neuroanatomy 101 and fMRI Contrast Mechanism • Preprocessing • Hemodynamic Response • “Univariate”](https://reader030.vdocuments.us/reader030/viewer/2022040416/5d49421888c9935f218bd682/html5/thumbnails/36.jpg)
36
Block Design: 15s Off, 15s On
Voxel 1
Voxel 2
Stimulus Schedule
Paradigm File
![Page 37: Basics of fMRI Analysis: Preprocessing, First Level ... · 2 Overview • Neuroanatomy 101 and fMRI Contrast Mechanism • Preprocessing • Hemodynamic Response • “Univariate”](https://reader030.vdocuments.us/reader030/viewer/2022040416/5d49421888c9935f218bd682/html5/thumbnails/37.jpg)
37
Contrasts and Inference
p = 10-11, sig=-log10(p) =11 p = .10, sig=-log10(p) =1
OFFin N Var, Mean,,,
ONin N Var, Mean,,,
)2(
)1()1(
2
2
2
22
−
−
−+
−+−
−=
OFFOFFOFF
ONONON
OFFON
OFFOFFONON
OFFON
N
N
NN
NNt
σβ
σβ
σσ
ββOFFON ββ −=Contrast
2
22
)2(
)1()1(st)Var(Contra
−+
−+−=
OFFON
OFFOFFONON
NN
NN σσ
ONβOFFβ
2
ONσ2
OFFσ
Voxel 1 Voxel 2
Note: z, t, F monotonic with p
![Page 38: Basics of fMRI Analysis: Preprocessing, First Level ... · 2 Overview • Neuroanatomy 101 and fMRI Contrast Mechanism • Preprocessing • Hemodynamic Response • “Univariate”](https://reader030.vdocuments.us/reader030/viewer/2022040416/5d49421888c9935f218bd682/html5/thumbnails/38.jpg)
38
Statistical Parametric Map (SPM)
+3%
0%
-3%
Contrast AmplitudeCON, COPE, CES
Contrast Amplitude
Variance
(Error Bars)VARCOPE, CESVAR
Significance
t-Map (p,z,F)
(Thresholded p<.01)
sig=-log10(p)“Massive Univariate Analysis”
-- Analyze each voxel separately
![Page 39: Basics of fMRI Analysis: Preprocessing, First Level ... · 2 Overview • Neuroanatomy 101 and fMRI Contrast Mechanism • Preprocessing • Hemodynamic Response • “Univariate”](https://reader030.vdocuments.us/reader030/viewer/2022040416/5d49421888c9935f218bd682/html5/thumbnails/39.jpg)
39
Hemodynamics
• Delay
• Dispersion
• Grouping by simple
time point inaccurate
![Page 40: Basics of fMRI Analysis: Preprocessing, First Level ... · 2 Overview • Neuroanatomy 101 and fMRI Contrast Mechanism • Preprocessing • Hemodynamic Response • “Univariate”](https://reader030.vdocuments.us/reader030/viewer/2022040416/5d49421888c9935f218bd682/html5/thumbnails/40.jpg)
40
Convolution with HRF
• Shifts, rolls off; more accurate
• Loose ability to simply group time points
• More complicated analysis
• General Linear Model (GLM)
![Page 41: Basics of fMRI Analysis: Preprocessing, First Level ... · 2 Overview • Neuroanatomy 101 and fMRI Contrast Mechanism • Preprocessing • Hemodynamic Response • “Univariate”](https://reader030.vdocuments.us/reader030/viewer/2022040416/5d49421888c9935f218bd682/html5/thumbnails/41.jpg)
41
GLM
Data fom
one voxel
= βTask βbase+
Baseline Offset
(Nuisance)Task
βTask=βon−βoff•Implicit Contrast
•HRF Amplitude
βbase=βoff
![Page 42: Basics of fMRI Analysis: Preprocessing, First Level ... · 2 Overview • Neuroanatomy 101 and fMRI Contrast Mechanism • Preprocessing • Hemodynamic Response • “Univariate”](https://reader030.vdocuments.us/reader030/viewer/2022040416/5d49421888c9935f218bd682/html5/thumbnails/42.jpg)
42
Matrix Model
y = X * β
βTask
βbase
Data from
one voxel
Design Matrix
Regressors
=
Vector of
Regression
Coefficients
(“Betas”)
Design Matrix
Ob
serv
atio
ns
![Page 43: Basics of fMRI Analysis: Preprocessing, First Level ... · 2 Overview • Neuroanatomy 101 and fMRI Contrast Mechanism • Preprocessing • Hemodynamic Response • “Univariate”](https://reader030.vdocuments.us/reader030/viewer/2022040416/5d49421888c9935f218bd682/html5/thumbnails/43.jpg)
43
Two Task Conditions
y = X * β
βOdd
βEven
βbase
Data from
one voxel
Design Matrix
Regressors
=
Design Matrix
Ob
serv
atio
ns
![Page 44: Basics of fMRI Analysis: Preprocessing, First Level ... · 2 Overview • Neuroanatomy 101 and fMRI Contrast Mechanism • Preprocessing • Hemodynamic Response • “Univariate”](https://reader030.vdocuments.us/reader030/viewer/2022040416/5d49421888c9935f218bd682/html5/thumbnails/44.jpg)
44
Working Memory Task (fBIRN)
Images Stick Figs
“Scrambled” Encode Distractor Probe
16s 16s 16s 16s
Stick Figs
0. “Scrambled” – low-level baseline, no response
1. Encode – series of passively viewed stick figures
Distractor – respond if there is a face
2. Emotional
3. Neutral
Probe – series of two stick figures (forced choice)
4. Following Emotional Distractor
5. Following Neutral Distractor
fBIRN: Functional Biomedical Research Network (www.nbirn.net)
![Page 45: Basics of fMRI Analysis: Preprocessing, First Level ... · 2 Overview • Neuroanatomy 101 and fMRI Contrast Mechanism • Preprocessing • Hemodynamic Response • “Univariate”](https://reader030.vdocuments.us/reader030/viewer/2022040416/5d49421888c9935f218bd682/html5/thumbnails/45.jpg)
45
Five Task Conditions
βEncode
βEmotDist
βNeutDist
βEmotProbe
βNeutProbe
=
y = X * β
![Page 46: Basics of fMRI Analysis: Preprocessing, First Level ... · 2 Overview • Neuroanatomy 101 and fMRI Contrast Mechanism • Preprocessing • Hemodynamic Response • “Univariate”](https://reader030.vdocuments.us/reader030/viewer/2022040416/5d49421888c9935f218bd682/html5/thumbnails/46.jpg)
46
GLM Solution
βEncode
βEmotDist
βNeutDist
βEmotProbe
βNeutProbe
=
y = X * β
• Set of simultaneous equations • Each row of X is an equation
• Each column of X is an unknown
• βs are unknown
• 142 Time Points (Equations)
• 5 unknowns
)(ˆ 1yXXX
TT −=β
![Page 47: Basics of fMRI Analysis: Preprocessing, First Level ... · 2 Overview • Neuroanatomy 101 and fMRI Contrast Mechanism • Preprocessing • Hemodynamic Response • “Univariate”](https://reader030.vdocuments.us/reader030/viewer/2022040416/5d49421888c9935f218bd682/html5/thumbnails/47.jpg)
47
Estimates of the HRF Amplitude
βEncode
βEmotDist
βNeutDist
βEmotProbe
βNeutProbe
=
Distractor Neutral following Probe toresponsein amplitude cHemodynamiˆ
Distractor Emotional following Probe toresponsein amplitude cHemodynamiˆ
Distractor Neutral toresponsein amplitude cHemodynamiˆ
Distractor Emotional toresponsein amplitude cHemodynamiˆ
Encode toresponsein amplitude cHemodynamiˆ
)(ˆ ,
NeuttProbe
EmotProbe
NeutDist
EmotDist
Encode
1
=
=
=
=
=
=+= −
β
β
β
β
β
ββ yXXXnXyTT
![Page 48: Basics of fMRI Analysis: Preprocessing, First Level ... · 2 Overview • Neuroanatomy 101 and fMRI Contrast Mechanism • Preprocessing • Hemodynamic Response • “Univariate”](https://reader030.vdocuments.us/reader030/viewer/2022040416/5d49421888c9935f218bd682/html5/thumbnails/48.jpg)
48
Hypotheses and ContrastsWhich voxels respond more/less/differently to the Emotional
Distractor than to the Neutral Distractor?
0,0,0ˆˆ
ˆˆ
ˆ
ˆˆ
NDistEDist
NDistEDist
NDistEDist
NDistEDist
≠<>−=
≠
<
>
ββγ
ββ
ββ
ββ
[ ] MatirxContrast 00110
0c
0c
1c
1c
0c
ˆcˆcˆcˆcˆc
NProbe
EProbe
NDist
EDist
Encode
NProbeNProbeEProbeEProbeNDistNDistEDistEDistEncodeEncode
−+=
=
=
−=
+=
=
++++=
C
βββββγ
Contrast: Assign Weights to each Beta
EDistβ̂NDistβ̂
![Page 49: Basics of fMRI Analysis: Preprocessing, First Level ... · 2 Overview • Neuroanatomy 101 and fMRI Contrast Mechanism • Preprocessing • Hemodynamic Response • “Univariate”](https://reader030.vdocuments.us/reader030/viewer/2022040416/5d49421888c9935f218bd682/html5/thumbnails/49.jpg)
49
Hypotheses
• Which voxels respond more to the Emotional Distractor
than to the Neutral Distractor?
• Which voxels respond to Encode (relative to baseline)?
• Which voxels respond to the Emotional Distractor?
• Which voxels respond to either Distractor?
• Which voxels respond more to the Probe following the
Emotional Distractor than to the Probe following the
Neutral Distractor?
![Page 50: Basics of fMRI Analysis: Preprocessing, First Level ... · 2 Overview • Neuroanatomy 101 and fMRI Contrast Mechanism • Preprocessing • Hemodynamic Response • “Univariate”](https://reader030.vdocuments.us/reader030/viewer/2022040416/5d49421888c9935f218bd682/html5/thumbnails/50.jpg)
50
Which voxels respond more to the Emotional Distractor than to
the Neutral Distractor?
1.
En
cod
e
2.
ED
ist
3. N
Dis
t
4.
EP
rob
e
5. N
Pro
be
• Only interested in Emotional and Neutral Distractors
• No statement about other conditions
Condition: 1 2 3 4 5
Weight 0 +1 -1 0 0
Contrast Matrix
C = [0 +1 -1 0 0]
![Page 51: Basics of fMRI Analysis: Preprocessing, First Level ... · 2 Overview • Neuroanatomy 101 and fMRI Contrast Mechanism • Preprocessing • Hemodynamic Response • “Univariate”](https://reader030.vdocuments.us/reader030/viewer/2022040416/5d49421888c9935f218bd682/html5/thumbnails/51.jpg)
51
Contrasts and the Full Model
( )
( )ate)(multivariTest -F ˆˆˆF
e)(univariatTest - tˆ)(
ˆ
ˆ
ˆt
Cin Rows J
Estimate VarianceContrast ˆ)(1ˆˆ
Contrast ˆˆ
ˆˆ Variance, Residual ˆˆ
ˆ
EstimatesParameter )(ˆ
),0(~ , ,
1
JDOF,
21DOF
212
2
1
2
γγ
σ
β
σ
γ
σσ
βγ
βσ
β
σβ
γ
γ
γγ
−
−
−
−
Σ=
==
=
=Σ=
=
−==
=
+=+=
T
n
TT
n
TT
T
n
TT
n
CXXC
C
CXXCJ
C
XynDOF
nn
yXXX
NnnsynXy
ONβ
OFFβ
2
ONσ2
OFFσ
2
22
)2(
)1()1(
−+
−+−
−=
OFFON
OFFOFFONON
OFFON
NN
NNt
σσ
ββ
![Page 52: Basics of fMRI Analysis: Preprocessing, First Level ... · 2 Overview • Neuroanatomy 101 and fMRI Contrast Mechanism • Preprocessing • Hemodynamic Response • “Univariate”](https://reader030.vdocuments.us/reader030/viewer/2022040416/5d49421888c9935f218bd682/html5/thumbnails/52.jpg)
52
First Level GLM Outputs
+3%
0%
-3%
Contrast AmplitudeCON, COPE, CES
Contrast Amplitude
Variance
(Error Bars)VARCOPE, CESVAR
Significance
t-Map (p,z,F)
(Thresholded p<.01)
sig=-log10(p)
![Page 53: Basics of fMRI Analysis: Preprocessing, First Level ... · 2 Overview • Neuroanatomy 101 and fMRI Contrast Mechanism • Preprocessing • Hemodynamic Response • “Univariate”](https://reader030.vdocuments.us/reader030/viewer/2022040416/5d49421888c9935f218bd682/html5/thumbnails/53.jpg)
53
Time Series Analysis Summary
• Correlational
• Design Matrix (HRF shape)
• Estimate HRF amplitude (Parameters)
• Contrasts to test hypotheses
• Results at each voxel:
• Contrast Value
• Contrast Value Variance
• p-value (Volume of Activation)
• Pass Contrast Value and Variance up to higher
level analyses
![Page 54: Basics of fMRI Analysis: Preprocessing, First Level ... · 2 Overview • Neuroanatomy 101 and fMRI Contrast Mechanism • Preprocessing • Hemodynamic Response • “Univariate”](https://reader030.vdocuments.us/reader030/viewer/2022040416/5d49421888c9935f218bd682/html5/thumbnails/54.jpg)
fMRI Group Analysis
![Page 55: Basics of fMRI Analysis: Preprocessing, First Level ... · 2 Overview • Neuroanatomy 101 and fMRI Contrast Mechanism • Preprocessing • Hemodynamic Response • “Univariate”](https://reader030.vdocuments.us/reader030/viewer/2022040416/5d49421888c9935f218bd682/html5/thumbnails/55.jpg)
55
fMRI Analysis Overview
Higher Level GLM
First Level
GLM Analysis
First Level
GLM Analysis
Subject 3
First Level
GLM Analysis
Subject 4
First Level
GLM Analysis
Subject 1
Subject 2
CX
CX
CX
CX
Preprocessing
MC, STC, B0
Smoothing
Normalization
Preprocessing
MC, STC, B0
Smoothing
Normalization
Preprocessing
MC, STC, B0
Smoothing
Normalization
Preprocessing
MC, STC, B0
Smoothing
Normalization
Raw Data
Raw Data
Raw Data
Raw Data
CX
![Page 56: Basics of fMRI Analysis: Preprocessing, First Level ... · 2 Overview • Neuroanatomy 101 and fMRI Contrast Mechanism • Preprocessing • Hemodynamic Response • “Univariate”](https://reader030.vdocuments.us/reader030/viewer/2022040416/5d49421888c9935f218bd682/html5/thumbnails/56.jpg)
56
Overview
• Goal of Group Analysis
• Types of Group Analysis
– Random Effects, Mixed Effects, Fixed Effects
• Multi-Level General Linear Model (GLM)
![Page 57: Basics of fMRI Analysis: Preprocessing, First Level ... · 2 Overview • Neuroanatomy 101 and fMRI Contrast Mechanism • Preprocessing • Hemodynamic Response • “Univariate”](https://reader030.vdocuments.us/reader030/viewer/2022040416/5d49421888c9935f218bd682/html5/thumbnails/57.jpg)
57
Spatial Normalization, Atlas Space
Subject 1
Subject 2
Subject 1
Subject 2
MNI305
Native Space MNI305 Space
Affine (12 DOF) Registration
![Page 58: Basics of fMRI Analysis: Preprocessing, First Level ... · 2 Overview • Neuroanatomy 101 and fMRI Contrast Mechanism • Preprocessing • Hemodynamic Response • “Univariate”](https://reader030.vdocuments.us/reader030/viewer/2022040416/5d49421888c9935f218bd682/html5/thumbnails/58.jpg)
58
Inter-Subject Averaging
Subje
ct 1
Subje
ct 2
NativeSpherical Spherical
Surface-to-
Surface
Surface-to-
Surface
GLM
Demographics
cf. Talairach
![Page 59: Basics of fMRI Analysis: Preprocessing, First Level ... · 2 Overview • Neuroanatomy 101 and fMRI Contrast Mechanism • Preprocessing • Hemodynamic Response • “Univariate”](https://reader030.vdocuments.us/reader030/viewer/2022040416/5d49421888c9935f218bd682/html5/thumbnails/59.jpg)
59
Is Pattern Repeatable Across Subject?
Subject 1 Subject 2 Subject 3 Subject 4 Subject 5
![Page 60: Basics of fMRI Analysis: Preprocessing, First Level ... · 2 Overview • Neuroanatomy 101 and fMRI Contrast Mechanism • Preprocessing • Hemodynamic Response • “Univariate”](https://reader030.vdocuments.us/reader030/viewer/2022040416/5d49421888c9935f218bd682/html5/thumbnails/60.jpg)
60
Group Analysis
Does not have to
be all positive!
Contrast Amplitudes
Contrast Amplitudes Variances
(Error Bars)
![Page 61: Basics of fMRI Analysis: Preprocessing, First Level ... · 2 Overview • Neuroanatomy 101 and fMRI Contrast Mechanism • Preprocessing • Hemodynamic Response • “Univariate”](https://reader030.vdocuments.us/reader030/viewer/2022040416/5d49421888c9935f218bd682/html5/thumbnails/61.jpg)
61
GβGσ ( )
( )1
1
,
2
2
22
−=
−
−=
==
∑
G
G
iG
G
G
GG
NDOF
N
Nt
G
G
ββσ
σσ
σ
ββ
β
“Random Effects (RFx)” Analysis
RFx
![Page 62: Basics of fMRI Analysis: Preprocessing, First Level ... · 2 Overview • Neuroanatomy 101 and fMRI Contrast Mechanism • Preprocessing • Hemodynamic Response • “Univariate”](https://reader030.vdocuments.us/reader030/viewer/2022040416/5d49421888c9935f218bd682/html5/thumbnails/62.jpg)
62
Gβ
Gσ
“Random Effects (RFx)” Analysis
RFx
• Model Subjects as a Random Effect
• Variance comes from a single source:
variance across subjects – Mean at the population mean
– Variance of the population variance
• Does not take first-level noise into account
(assumes 0)
• “Ordinary” Least Squares (OLS)
• Usually less activation than individuals
![Page 63: Basics of fMRI Analysis: Preprocessing, First Level ... · 2 Overview • Neuroanatomy 101 and fMRI Contrast Mechanism • Preprocessing • Hemodynamic Response • “Univariate”](https://reader030.vdocuments.us/reader030/viewer/2022040416/5d49421888c9935f218bd682/html5/thumbnails/63.jpg)
63
“Mixed Effects (MFx)” Analysis
MFx
RFx
• Down-weight each subject based on variance.
• Weighted Least Squares vs (“Ordinary” LS)
![Page 64: Basics of fMRI Analysis: Preprocessing, First Level ... · 2 Overview • Neuroanatomy 101 and fMRI Contrast Mechanism • Preprocessing • Hemodynamic Response • “Univariate”](https://reader030.vdocuments.us/reader030/viewer/2022040416/5d49421888c9935f218bd682/html5/thumbnails/64.jpg)
64
“Mixed Effects (MFx)” Analysis
MFx
• Down-weight each subject based on variance.
• Weighted Least Squares vs (“Ordinary” LS)
• Protects against unequal variances across group or
groups (“heteroskedasticity”)
• May increase or decrease significance with respect to
simple Random Effects
• More complicated to compute
• “Pseudo-MFx” – simply weight by first-level
variance (easier to compute)
![Page 65: Basics of fMRI Analysis: Preprocessing, First Level ... · 2 Overview • Neuroanatomy 101 and fMRI Contrast Mechanism • Preprocessing • Hemodynamic Response • “Univariate”](https://reader030.vdocuments.us/reader030/viewer/2022040416/5d49421888c9935f218bd682/html5/thumbnails/65.jpg)
65
“Fixed Effects (FFx)” Analysis
FFx
RFx
( )
∑
∑
=
=
=
i
G
i
G
DOFDOF
N
t
G
G
2
2
2
2
σσ
σ
β
β
β
2
iσ
![Page 66: Basics of fMRI Analysis: Preprocessing, First Level ... · 2 Overview • Neuroanatomy 101 and fMRI Contrast Mechanism • Preprocessing • Hemodynamic Response • “Univariate”](https://reader030.vdocuments.us/reader030/viewer/2022040416/5d49421888c9935f218bd682/html5/thumbnails/66.jpg)
66
“Fixed Effects (FFx)” Analysis
FFx
( )
∑
∑
=
=
=
i
G
i
G
DOFDOF
N
t
G
G
2
2
2
2
σσ
σ
β
β
β
• As if all subjects treated as a single subject (fixed effect)
• Small error bars (with respect to RFx)
• Large DOF
• Same mean as RFx
• Huge areas of activation
• Not generalizable beyond sample.
![Page 67: Basics of fMRI Analysis: Preprocessing, First Level ... · 2 Overview • Neuroanatomy 101 and fMRI Contrast Mechanism • Preprocessing • Hemodynamic Response • “Univariate”](https://reader030.vdocuments.us/reader030/viewer/2022040416/5d49421888c9935f218bd682/html5/thumbnails/67.jpg)
67
fMRI Analysis Overview
Higher Level GLM
First Level
GLM Analysis
First Level
GLM Analysis
Subject 3
First Level
GLM Analysis
Subject 4
First Level
GLM Analysis
Subject 1
Subject 2
C1X1
C1X1
C1X1
C1X1
Preprocessing
MC, STC, B0
Smoothing
Normalization
Preprocessing
MC, STC, B0
Smoothing
Normalization
Preprocessing
MC, STC, B0
Smoothing
Normalization
Preprocessing
MC, STC, B0
Smoothing
Normalization
Raw Data
Raw Data
Raw Data
Raw Data
CGXG
![Page 68: Basics of fMRI Analysis: Preprocessing, First Level ... · 2 Overview • Neuroanatomy 101 and fMRI Contrast Mechanism • Preprocessing • Hemodynamic Response • “Univariate”](https://reader030.vdocuments.us/reader030/viewer/2022040416/5d49421888c9935f218bd682/html5/thumbnails/68.jpg)
68
GβGσ
Higher Level GLM Analysis
=
1
1
1
1
1
βG
y = X * β
Data from
one voxel
Design Matrix
(Regressors)
Vector of
Regression
Coefficients
(“Betas”)
Ob
serv
atio
ns
(Lo
w-L
evel
Co
ntr
asts
)
Contrast Matrix:
C = [1]
Contrast = C*β = βG
One-Sample Group Mean (OSGM)
![Page 69: Basics of fMRI Analysis: Preprocessing, First Level ... · 2 Overview • Neuroanatomy 101 and fMRI Contrast Mechanism • Preprocessing • Hemodynamic Response • “Univariate”](https://reader030.vdocuments.us/reader030/viewer/2022040416/5d49421888c9935f218bd682/html5/thumbnails/69.jpg)
69
Summary
• Preprocessing – MC, STC, B0, Normalize, Smooth
• First Level GLM Analysis – Design matrix, HRF,
Nuisance
• Contrasts, Hypothesis Testing – contrast matrix
• Group Analysis
– Random, Mixed, Fixed
– Multi-level GLM (Design and Contrast Matrices)