Peter A' Bandettini, Ph'D'

1
What further information about human brain
function and physiology can we obtain with fMRI?

• Peter A. Bandettini, Ph.D.
• Section on Functional Imaging Methods
• http//fim.nimh.nih.gov
• Laboratory of Brain and Cognition
• Functional MRI Facility
• http//fmrif.nimh.nih.gov

2
fMRI Setup
3
MRI vs. fMRI
MRI
fMRI
one image
Time

many images (e.g., every 2 sec for 5 mins)
high resolution (1 mm or less)
low resolution (1.5 to 4 mm)
4
BOLD (Blood Oxygen Level Dependent) Contrast
blood flow oxygenated-blood MR signal

5
Basis of BOLD Contrast
Oxygenated and deoxygenated red blood cells have
different magnetic properties
oxygenated
deoxygenated
red blood cells
L. Pauling, C. D. Coryell, Proc.Natl. Acad. Sci.
USA 22, 210-216, 1936. K.R. Thulborn, J. C.
Waterton, et al., Biochim. Biophys. Acta. 714
265-270, 1982. S. Ogawa, T. M. Lee, A. R. Kay, D.
W. Tank, Proc. Natl. Acad. Sci. USA 87,
9868-9872, 1990.
6
BOLD Contrast Imaging
7
1991
8
Activation Statistics
Functional images
ROI Time Course
fMRI Signal ( change)
2s
Condition
Time
Condition 1
Statistical Map superimposed on anatomical MRI
image
Time
Condition 2
...
5 min
Region of interest (ROI)
9
3 2 1 0 -1
Brain Map Column Layer
NIRS NIRS
PET
MEG, EEG
NIRS
Log Size (mm)
fMRI
MRI
-3 -2 -1 0 1 2 3 4 5 6 7
Millisecond Second Minute Hour Day
Log Time (sec)
10
Spatial and Temporal Resolution
Latency Variation
Cheng, et al. (2001) Neuron,32359-374
P. A. Bandettini, (1999) "Functional MRI"
205-220.
Spatial
Temporal
11
Interpretation
Neuronal Activation
Measured Signal
?
?
?
?
Hemodynamics
Noise
12
Medline Search for fMRI or functional MRI
13
Motor (black) Primary Sensory (red) Integrative
Sensory (violet) Basic Cognition
(green) High-Order Cognition (yellow) Emotion
(blue)
J. Illes, M. P. Kirschen, J. D. E. Gabrielli,
Nature Neuroscience, 6 (3)m p.205
14
Methodology
Technology
Interpretation
Applications
15
Methodology
Technology
Connectivity analysis free behavior
designs resting state multi-modal
integration real time fMRI pattern
classification voxel based morphometry
higher field strength parallel excitation/acquisit
ion higher resolution perfusion imaging unique
contrasts Diffusion Tensor Imaging
decision making social cognition genotype/phenotyp
e clinical brain-reading
fluctuations dynamics cross - modal comparisons
Interpretation
Applications
16
Methodology
Technology
Interpretation
Applications
17

• Technology
• Parallel Acquisition
• SENSE Imaging
• High Fields

18
Parallel Acquisition
Technology
J. Bodurka, et al, Magnetic Resonance in Medicine
51 (2004) 165-171.
19
Parallel Acquisition
Technology
J. Bodurka
20
SENSE Imaging
Technology
5 to 30 ms
Pruessmann, et al.
21
SENSE Imaging
Technology
3T single-shot SENSE EPI using 16 channels
1.25x1.25x2mm
22
High Fields
Technology
7T head coil 3T head coil
TSE, 11 echoes, 7 min exam, 20cm FOV, 512x512
(0.4mm x 0.4mm), 3mm thick slices.
7T 3T white matter SNR 65 white matter SNR
26 Gray matter SNR 76 Gray matter SNR 34
Courtesy Larry Wald
23
Different fiber bundles depict different
intensities
24
fiber bundles?
25
FSE images at 0.2x.2x1mm3
26
(No Transcript)
27
(No Transcript)
28
Layered structure i n the visual cortex
29
Susceptibility field (in Gauss) increases w/ Bo
High Fields
Technology
Ping-pong ball in H20 Field maps (DTE 5ms),
black lines spaced by 0.024G (0.8ppm at 3T)
1.5T 3T 7T
Courtesy Larry Wald
30
High Fields
Technology
7T Single Shot whole head EPI
1.5mm inplane
single shot EPI, 7T. 128x128, 20cm FOV (1.5mm
resolution), 2mm slice, TE 20ms
Courtesy Larry Wald
31
7T Blood flow and BOLD based fMRI
High Fields
Technology
Longer T1 means better ASL
6 minute pulse Arterial Spin Labeling blood flow
image 1.56mm x 1.56mm x 4mm (3T typical
resolution 3mmx 3mm x 5mm)
Courtesy Larry Wald
32
Methodology
Technology
Interpretation
Applications
33

• Methodology
• New Contrasts
• Paradigm Designs
• Temporal Resolution
• Spatial Resolution
• Processing Methods

34
New Contrasts
Methodology
fMRI Contrast

• Volume (gadolinium)
• BOLD
• Perfusion (ASL)
• ?CMRO2
• ?Volume (VASO)
• Neuronal Currents
• Diffusion coefficient
• Temperature

35
Methodology
Neuronal Activation Input Strategies
1. Block Design 2. Frequency Encoding 3. Phase
Encoding 4. Event-Related 5. Orthogonal Block
Design 6. Free Behavior Design.
36
Methodology
Neuronal Activation Input Strategies
1. Block Design 2. Frequency Encoding 3. Phase
Encoding 4. Event-Related 5. Orthogonal Block
Design 6. Free Behavior Design.
37
NAIS
Methodology
Free Behavior
Resting State Correlations
Rest seed voxel in motor cortex
Activation correlation with reference function
B. Biswal et al., MRM, 34537 (1995)
38
NAIS
Methodology
Free Behavior
0.3 Hz
FFT
MEG Power Spectrum
39
NAIS
Methodology
Free Behavior
BOLD correlated with 10 Hz power during Rest
Positive
10 Hz power
Negative
Goldman, et al (2002), Neuroreport
40
NAIS
Methodology
Free Behavior
BOLD correlated with SCR during Rest
J. C. Patterson II, L. G. Ungerleider, and P. A
Bandettini, NeuroImage 17 1787-1806, (2002).
41
NAIS
Methodology
Free Behavior
Resting state connectivity
Decreases during cognitive tasks
42
NAIS
Methodology
Free Behavior
Regions showing decreases during cognitive tasks
McKiernan, et al (2003), Journ. of Cog. Neurosci.
15 (3), 394-408
43
The Signal in the Noise
Resting State Connectivity
rest
perception action
strongly reducedvigilance
sleep
tasks
default mode
vegetativestate
generalanaesthesia
(Gusnard and Raichle 2001)
44
The Signal in the Noise
Resting State Connectivity
Spatial similarity of decreased signal change
regions with regions showing resting state
correlations.
45
NAIS
Methodology
Free Behavior
Approaches to assessing spatial
connectivity ICA, PCA, seed voxel
Why not correlate every voxel with every other
voxel? For 64 x 64 resolution, 27 slices, 165
time points -160 voxel ROI 5 min and 63 MB
memory. -Gray matter (10,000 voxels) 5.32 hrs
and 4.3 GB memory -Entire volume (110,000 voxels)
59 hrs and 47.5 GB memory
46
NAIS
Methodology
Free Behavior
Effects of Respiration on Signal
Birn et al. NeuroImage 2005
47
NAIS
Methodology
Free Behavior
Individual Maps
Birn et al. NeuroImage
48
Local Correlations
N. Kriegeskorte, J. Bodurka
49
NAIS
Methodology
Free Behavior
One doesnt need prior knowledge as long as the
task is repeatable
Levin, et al (2001), NeuroImage, 13, 153-160
50
NAIS
Methodology
Free Behavior
Hasson, et al (2004), Science, 303, 1634-1640
51
Temporal Resolution
Methodology
52
Spatial Resolution
Methodology
53
Methodology
Technology
Interpretation
Applications
54

• Interpretation
• Post Undershoot
• Linearity (effect of task duty cycle)

55
Post Undershoot
Interpretation
BOLD post-stimulus undershoot
A BOLD undershoot without a CBF undershoot could
be due to a slow return to baseline of either CBV
or CMRO2
Courtesy Rick Buxton
56
Post Undershoot
Interpretation
BOLD Signal Dynamics
Courtesy Rick Buxton
57
Linearity
Interpretation
Increases linearity
Logothetis et al. (2001) Nature, 412, 150-157.
P. A. Bandettini et al, (2001) Nature
Neuroscience, 4 864-866.
58
Linearity
Interpretation
Sources of this Nonlinearity

• Neuronal
• Hemodynamic
• Oxygen extraction
• Blood volume dynamics

Oxygen Extraction
Flow In
Flow Out
? Volume
59
duty cycle effects
Linearity
Interpretation
R.M. Birn, P. A. Bandettini. NeuroImage, 27,
70-82 (2005)
60
Linearity
Interpretation
duty cycle effects
61
Methodology
Technology
Interpretation
Applications
62
Brain Reading
Rather than mapping what is correlated with a
task, brain reading involves predicting what
the brain is doing based on the pattern of
activation.
63
(No Transcript)
64
Pattern-recognition analysis of fMRI activity
patterns

• Haxby et al. (2001)
• Cox Savoy (2003)
• Carlson et al. (2003)
• Kamitani Tong (2005)
• Haynes Rees (2005)

65
Visual object categories distinguished by widely
distributed inferotemporal activity pattern
Haxby et al. (2001)
66
Orientation informationin early visual
areasKamitani Tong (2005), Haynes Rees (2005)
67
Let's image the fine-scale orientation map
with 3-mm voxels...
Boynton (2005), News Views on Kamitani Tong
(2005) and Haynes Rees (2005)
68
Boynton (2005), News Views on Kamitani Tong
(2005) and Haynes Rees (2005)
69
(No Transcript)
70
fMRI information
neuronal activity pattern
fMRI activity pattern
condition 1
condition 2
71
Interpretation
The Signal
Sources of this Nonlinearity

• Neuronal
• Hemodynamic
• Oxygen extraction
• Blood volume dynamics

Oxygen Extraction
Flow In
Flow Out
? Volume
72
NEUROIMAGE 19 (2) 261-270 Part 1 JUN 2003
73
N. Kriegeskorte
74
Extensive Individual Differences in Brain
Activations During Episodic Retrieval Miller et
al., 2002
Courtesy, Mike Miler, UC Santa Barbara and Jack
Van Horn, fMRI Data Center, Dartmouth University
75
Section on Functional Imaging Methods Rasmus
Birn David Knight Anthony Boemio Nikolaus
Kriegeskorte Kevin Murphy Monica Smith Najah
Waters Marieke Mur Natalia Petridou Jason
Diamond Jon West Functional MRI Facility Kay
Kuhns Sean Marrett Wen-Ming Luh Jerzy
Bodurka Adam Thomas
Karen Bove-Bettis Ellen Condon Sahra Omar Alda
Ottley Paula Rowser Janet Ebron James Hoskie