Introduction:fMRI for Newbies

MRI studies brain anatomy.Functional MRI (fMRI) studies brain function.

MRI vs. fMRI

Brain Imaging: Anatomy

Photography

CAT

PET

MRI

Source: modified from Posner & Raichle, Images of Mind

MRI vs. fMRI

neural activity blood oxygen fMRI signal

MRI fMRI

one image

many images (e.g., every 2 sec for 5 mins)

high resolution(1 mm)

low resolution(~3 mm but can be better)

fMRI Blood Oxygenation Level Dependent (BOLD) signal

indirect measure of neural activity

…

E = mc2

???

The First “Brain Imaging Experiment”

“[In Mosso’s experiments] the subject to be observed lay on a delicately balanced table which could tip downward either at the head or at the foot if the weight of either end were increased. The moment emotional or intellectual activity began in the subject, down went the balance at the head-end, in consequence of the redistribution of blood in his system.”

-- William James, Principles of Psychology (1890)

Angelo MossoItalian physiologist

(1846-1910)

… and probably the cheapest one too!

100

200

300

400

500

600

700

800

0

Num

ber

of p

aper

s (P

ubM

ed)

1990 1995 2000

746 papers (2001)

Year of Publication Slide modified from Mel Goodale

The Rise of fMRI

fMRI Activation

Time

BrainActivity

Source: Kwong et al., 1992

Flickering CheckerboardOFF (60 s) - ON (60 s) -OFF (60 s) - ON (60 s) - OFF (60 s)

PET and fMRI Activation

Source: Posner & Raichle, Images of Mind

fMRI Setup

Category-Specific Visual Areas

• Lateral Occipital (LO)– object-selective– objects > (faces & scenes)– objects > scrambled images

faces

places

objects

Malach, 2002, TICS

• Fusiform Face Area (FFA) or pFs– face-selective– faces > (objects & scenes)– faces > scrambled images– ~ posterior fusiform sulcus (pFs)

• Parahippocampal Place Area (PPA)– place-selective– places > (objects and faces)– places > scrambled images

A Simple Experiment: LO Localizer

IntactObjects

ScrambledObjects

BlankScreen

TIME

One volume (12 slices) every 2 seconds for 272 seconds (4 minutes, 32 seconds)

Condition changes every 16 seconds (8 volumes)

Lateral Occipital Complex• responds when subject views objects

fMRI Experiment Stages: Prep

1) Prepare subject• Consent form• Safety screening• Instructions and practice trials if appropriate

2) Shimming • putting body in magnetic field makes it non-uniform• adjust 3 orthogonal weak magnets to make magnetic field as homogenous as

possible

3) SagittalsTake images along the midline to use to plan slices

Note: That’s one g, two t’s

In this example, these are the functional slices we want: 12 slices x 6 mm

fMRI Experiment Stages: Anatomicals4) Take anatomical (T1) images

• high-resolution images (e.g., 0.75 x 0.75 x 3.0 mm)• 3D data: 3 spatial dimensions, sampled at one point in time• 64 anatomical slices takes ~4 minutes

64 slices x 3 mm

Slice Thicknesse.g., 6 mm

Number of Slicese.g., 10

SAGITTAL SLICE IN-PLANE SLICE

Field of View (FOV)e.g., 19.2 cm

VOXEL(Volumetric Pixel)

3 mm

3 mm6 mm

Slice Terminology

Matrix Sizee.g., 64 x 64

In-plane resolutione.g., 192 mm / 64

= 3 mm

fMRI Experiment Stages: Functionals5) Take functional (T2*) images

• images are indirectly related to neural activity• usually low resolution images (3 x 3 x 6 mm)• all slices at one time = a volume (sometimes also called an image)• sample many volumes (time points) (e.g., 1 volume every 2 seconds for 136

volumes = 272 sec = 4:32)• 4D data: 3 spatial, 1 temporal

…

Anatomic Slices Corresponding to Functional Slices

Time Courses

TIME

MR

SIG

NA

L(A

RB

ITR

AR

Y U

NIT

S)

MR

SIG

NA

L(%

Cha

nge)

Arbitrary signal varies from voxel to voxel, day to day, subject to subject

To make the y-axis more meaningful, we usually convert the signal into units of % change:

100*(x - baseline)/baseline

Changes are typically in the order of 0.5-4 %.

Statistical Mapsuperimposed on

anatomical MRI image

~2s

Functional images

Time

Condition 1

Condition 2 ...

~ 5 min

Time

fMRISignal

(% change)

ROI Time Course

Condition

Activation Statistics

Region of interest (ROI)

Statistical Maps & Time Courses

Use stat maps to pick regions

Then extract the time course

2D 3D

Design Jargon: Runs

run (or scan): one continuous period of fMRI scanning (~5-7 min) session: all of the scans collected from one subject in one day

experiment: a set of conditions you want to compare to each othercondition: one set of stimuli or one task

4 stimulus conditions+ 1 baseline condition (fixation)

A session consists of one or more experiments.Each experiment consists of several (e.g., 1-8) runsMore runs/expt are needed when signal:noise is low or the effect is weak.Thus each session consists of numerous (e.g., 5-20) runs (e.g., 0.5 – 3 hours)

Note: Terminology can vary from one fMRI site to another (e.g., some places use “scan” to refer to what we’ve called a volume).

Design Jargon: Paradigm

paradigm (or protocol): the set of conditions and their order used in a particular run

Time

volume #1(time = 0)

volume #105(time = 105 vol x 2 sec/vol = 210 sec = 3:30)

runepoch: one instance of a condition

first “objects right” epochsecond “objects right” epoch

epoch 8 vol x 2 sec/vol = 16 sec

fMRI Equipment

Magnet (4T)

Head Coil

Source: Joe Gati, photos

RF Coil

4T magnet

gradient coil(inside)

Surface Coil

Gradient Coil

What Does fMRI Measure?

• Big magnetic field– protons (hydrogen molecules) in body become aligned to field

• RF (radio frequency) coil– radio frequency pulse– knocks protons over– as protons realign with field, they emit energy that coil receives

(like an antenna)

• Gradient coils– make it possible to encode spatial information

• MR signal differs depending on– concentration of hydrogen in an area (anatomical MRI)– amount of oxy- vs. deoxyhemoglobin in an area (functional MRI)

BOLD signal

Source: fMRIB Brief Introduction to fMRI

neural activity blood flow oxyhemoglobin T2* MR signal

Blood Oxygen Level Dependent signal

JODY – THE FOLLOWING SLIDES ARE FROM THE NORWAY PAGE

Statistical Maps & Time Courses

Use stat maps to pick regions

Then extract the time course

Percent Signal Change

500

505

200

205

1%

1%

Slide from Duke course

Stats on Anatomical

2D 3D

Design Jargon: Runs

run (or scan): one continuous period of fMRI scanning (~5-7 min) session: all of the scans collected from one subject in one day

experiment: a set of conditions you want to compare to each othercondition: one set of stimuli or one task

2 stimulus conditions+ 1 baseline condition (fixation)

A session consists of one or more experiments.Each experiment consists of several (e.g., 1-8) runsMore runs/expt are needed when signal:noise is low or the effect is weak.Thus each session consists of numerous (e.g., 5-20) runs (e.g., 0.5 – 3 hours)

Note: Terminology can vary from one fMRI site to another (e.g., some places use “scan” to refer to what we’ve called a volume).

Design Jargon: Paradigm

paradigm (or protocol): the set of conditions and their order used in a particular run

Time

volume #1(time = 0)

volume #136(time = 136 vol x 2 sec/vol = 272 sec = 4:32)

run

first “intact objects” epoch

second “intact objects” epoch

epoch: one instance of a condition

epoch 8 vol x 2 sec/vol = 16 sec

first “scrambled objects” epoch

Recipe for MRI

1) Put subject in big magnetic field (leave him there)

2) Transmit radio waves into subject [about 3 ms]

3) Turn off radio wave transmitter

4) Receive radio waves re-transmitted by subject– Manipulate re-transmission with magnetic fields during this readout

interval [10-100 ms: MRI is not a snapshot]

5) Store measured radio wave data vs. time– Now go back to 2) to get some more data

6) Process raw data to reconstruct images

7) Allow subject to leave scanner (this is optional)

Source: Robert Cox’s web slides

Necessary Equipment

Magnet Gradient Coil RF Coil

Source for Photos: Joe Gati

RF Coil

4T magnet

gradient coil(inside)

Susceptibility Artifacts

-artifacts occur near junctions between air and tissue• sinuses, ear canals

sinuses

earcanals

T1-weighted image T2*-weighted image

The Benefit of Susceptibility

Modified from: Robert Cox’s web slides

Susceptibility variations can also be seen around blood vessels where deoxyhemoglobin affects T2* in nearby tissue

BOLD Correlations

Local Field Potentials (LFP)• reflect post-synaptic potentials• similar to what EEG (ERPs) and MEG

measure

Multi-Unit Activity (MUA)• reflects action potentials• similar to what most electrophysiology

measures

Logothetis et al. (2001)• combined BOLD fMRI and

electrophysiological recordings • found that BOLD activity is more closely

related to LFPs than MUA

Source: Logothetis et al., 2001, Nature

Comparing Electrophysiolgy and BOLD

Data Source: Disbrow et al., 2000, PNASFigure Source, Huettel, Song & McCarthy, Functional Magnetic Resonance Imaging

fMRI Measures the Population Activity• population activity depends on

– how active the neurons are– how many neurons are active

• manipulations that change the activity of many neurons a little have a show bigger activation differences than manipulations that change the activation of a few neurons a lot– attention

activity

– learning activity

• fMRI may notmatch single neuronphysiology results

Verb generation Verb generation after 15 min practice

Raichle & Posner, Images of Mind cover imageIdeas from: Scannell & Young, 1999, Proc Biol Sci

Why are vessels a problem?• large vessels produce BOLD activation further from the true site of activation than small vessels (especially problematic for high-resolution fMRI)• large vessels line the sulci and make it hard to tell which bank of a sulcus the activity arises from • the % signal change in large vessels can be considerably higher than in small vessels (e.g., 10% vs. 2%)• activation in large vessels occurs later than in small ones• vessel artifacts are worse with gradient echo sequences (compared to asymmetric spin echo for example) and low field strengths

Source: Ono et al., 1990, Atlas of the Cerebral Sulci

Don’t Trust Sinus Activity

• You will sometimes see bogus “activity” in the sagittal sinus

More Caveats

• “brain vs. vein” debate – source of signal affects spatial resolution

• scientists haven’t agreed on a single theory to explain the relationship between oxygen, glucose metabolism and blood flow

• no one really understands how neurons trigger increased blood flow– neural synchrony may be a factor

Bottom Line

• Despite all the caveats, questions and concerns, BOLD imaging is well-correlated with results from other methods

• BOLD imaging can resolve activation at a fairly small scale (e.g., retinotopic mapping)

• PSPs and action potentials are correlated so either way, it’s getting at something meaningful

• even if BOLD activation doesn’t correlate completely with electrophysiology, that doesn’t mean it’s wrong– may be picking up other processing info (e.g., PSPs, synchrony)