What Neuroscientists Can and Cannot Learn from fMRI Last Update: January 18, 2012 Last Course:...

What Neuroscientists Can and Cannot Learn from fMRI

http://www.fmri4newbies.com/

Last Update: January 18, 2012Last Course: Psychology 9223, W2010, University of Western Ontario

Jody CulhamBrain and Mind Institute

Department of PsychologyUniversity of Western Ontario

Section 1The BOLD Signal

Deoxygenated Blood Signal Loss

Oxygenated blood?• Diamagnetic• Doesn’t distort surrounding

magnetic field• No signal loss…

Deoxygenated blood?

• Paramagnetic

• Distorts surrounding magnetic field

• Signal loss !!!

Images from Huettel, Song & McCarthy, 2004, Functional Magnetic Resonance Imagingbased on two papers from Ogawa et al., 1990, both in Magnetic Resonance in Medicine

rat breathing pure oxygen

rat breathing normal air (less oxygen than pure oxygen)

Hemoglobin (Hb)

Figure Source, Huettel, Song & McCarthy, 2004, Functional Magnetic Resonance Imaging

BOLD Time CourseBlood Oxygenation Level-Dependent Signal

Positive BOLD response

InitialDip

OvershootPost-stimulusUndershoot

0

1

2

3

BO

LD R

espo

nse

(% s

igna

l cha

nge)

Time

Stimulus

Perhaps it should be BDLD?Blood DE-oxygenation level-dependent signal?

• Technically, “BOLD” is a misnomer• The fMRI signal is dependent on deoxygenation

rather than oxygenation per se• The more deoxy-Hb there is the lower the signal

fMRISignal

Amount of deoxy-Hb

“BDLD” idea from Bruce Pike, MNI

Deoxy-Hb

0

1

2

3

Deo

xy-H

b C

once

ntra

tion

Time

Stimulus

Initial Dip (Hypo-oxic Phase)• Transient increase in oxygen consumption, before

change in blood flow – Menon et al., 1995; Hu, et al., 1997

• Smaller amplitude than main BOLD signal– 10% of peak amplitude (e.g., 0.1% signal change)

• Potentially more spatially specific– Oxygen utilization may be more closely associated with

neuronal activity than positive response

Slide modified from Duke course

Rise (Hyperoxic Phase)

• Results from vasodilation of arterioles, resulting in a large increase in cerebral blood flow

• Inflection point can be used to index onset of processing

Slide modified from Duke course

Peak – Overshoot

• Over-compensatory response– More pronounced in BOLD signal measures than flow

measures

• Overshoot found in blocked designs with extended intervals– Signal saturates after ~10s of stimulation

Slide modified from Duke course

Sustained Response

• Blocked design analyses rest upon presence of sustained response– Comparison of sustained activity vs. baseline– Statistically simple, powerful

• Problems– Difficulty in identifying magnitude of activation– Little ability to describe form of hemodynamic response

Slide modified from Duke course

Undershoot

• Cerebral blood flow more locked to stimuli than cerebral blood volume– Increased blood volume with baseline flow leads to

decrease in MR signal

• More frequently observed for longer-duration stimuli (>10s)– Short duration stimuli may not evidence– May remain for 10s of seconds

Slide modified from Duke course

Hu et al., 1997, MRM

Evolution of BOLD Response

Trial to Trial Variability

Huettel, Song & McCarthy, 2004,Functional Magnetic Resonance Imaging

Variability of HRF Between SubjectsAguirre, Zarahn & D’Esposito, 1998• HRF shows considerable variability between subjects

• Within subjects, responses are more consistent, although there is still some variability between sessions

different subjects

same subject, same session same subject, different session

Variability of HRF Between AreasPossible caveat: HRF may also vary between areas, not just subjects

• Buckner et al., 1996: • noted a delay of .5-1 sec between visual and prefrontal regions• vasculature difference?• processing latency?

• Bug or feature? • Menon & Kim – mental chronometry

Buckner et al., 1996

Variability Between Subjects/Areas

• greater variability between subjects than between regions

• deviations from canonical HRF cause false negatives (Type II errors)

• Consider including a run to establish subject-specific HRFs from robust area like M1

Handwerker et al., 2004, Neuroimage

Sampling Rate

Huettel, Song & McCarthy, 2004, Functional Magnetic Resonance Imaging

Linearity of BOLD response

Source: Dale & Buckner, 1997

Linearity:“Do things add up?”

red = 2 - 1

green = 3 - 2

Sync each trial response to start of trial

Not quite linear but good enough!

Section 2From Neurons to BOLD

From Neurons to BOLD

• Any similarity in the shapes of the curves for action potentials and the BOLD response is purely coincidental (but still kinda cool)

-70

-55

0

40

Refractory period

Dep

olar

izat

ion R

epolarization

Vo

ltage

(m

V)

Time (ms)

0

1

UndershootBO

LD

Sig

na

l Ch

ang

e (

%)

Time (s)

Positive BOLD Response

Stimulus to BOLD

Source: Arthurs & Boniface, 2002, Trends in Neurosciences

Neural Networks

Post-Synaptic Potentials

• The inputs to a neuron (post-synaptic potentials) increase (excitatory PSPs) or decrease (inhibitory PSPs) the membrane voltage

• If the summed PSPs at the axon hillock push the voltage above the threshold, the neuron will fire an action potential

What does electrophysiology measure?

Source: http://www.cin.uni-tuebingen.de/research/methods-in-neuroscience/networks.php

Raw microelectrode signal

Filter out low frequencies Action Potentials (APs)

Filter out high frequencies Local Field Potentials (LFPs)

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

4 s stimulus

12 s stimulus

24 s stimulus

Even Simple Circuits Aren’t Simple

Will BOLD activation from the blue voxel reflect:

• output of the black neuron (action potentials)?

• excitatory input (green synapses)?

• inhibitory input (red synapses)?

• inputs from the same layer (which constitute ~80% of synapses)?

• feedforward projections (from lower-tier areas)?

• feedback projections (from higher-tier areas)?

Lower tier area (e.g., thalamus)

Middle tier area (e.g., V1, primary

visual cortex)

Higher tier area (e.g., V2, secondary

visual cortex)

…

gray matter(dendrites, cell bodies

& synapses)

white matter(axons)

Comparing Electrophysiolgy and BOLD

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

fMRI Measures the Population Activity

Ideas from: Scannell & Young, 1999, Proc Biol Sci fMRI for Dummies

Effects of Practice

Verb generation Verb generation after 15 min practice

Raichle & Posner, Images of Mind cover image

fMRI for Dummies

Bug or feature?• fMRI adaptation enables us to study the tuning of

neurons

Stimulus to BOLD

Source: Arthurs & Boniface, 2002, Trends in Neurosciences

Vasculature: Brain vs. Vein

Source: Menon & Kim, TICS

Contents of a Voxel

Source: Logothetis, 2008, Nature

Capillary beds within the cortex

Source: Duvernoy, Delon & Vannson, 1981, Brain

Research Bulletin

“Brain vs. Vein”• 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

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

Vessel Valves

Source: Harrison et al. (2002). Cerebral Cortex.

Dilation of Arterioles

• biggest changes in arteriole dilation occurred near stimulation; however, effects could also be observed several mm upstream

Source: Adapted from Takano et al., 2006, Nat Neurosci, by Huettel, et al., 2nd ed.

Tim

e

stim

max dilation ~3-6 s after stim

vasodilation could be induced by either electrical stimulation or release of Ca2+

Upstream Effects

• biggest changes in arteriole dilation occurred near stimulation; however, effects could also be observed several mm upstream

Source: Adapted from Iadecola et al., 1997, J Neurophysiol, by Huettel, et al., 2nd ed.

arteriole

veins

Don’t Trust Sinus Activity

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

Energy Budget

Data Source: Atwell & Laughlin, 2001, J. Cereb. Blood Flow Metab.Figure Source, Huettel, Song & McCarthy, Functional Magnetic Resonance Imaging

The Forgotten Brain Cells

Common (i.e., Wrong) Wisdom

“Glial cells are probably not essential for processing information”(Kandel, Schwartz & Jessell, Principles of Neural Science 3rd Ed.)

• Astrocytes are adjacent to both synapses and blood vessels– well poised to adjust vascular response to neural activity

• Astrocytes outnumber neurons by at least 10:1 and comprise ~50% of the total CNS volume

Astrocytes perform a number of critically important functions:

1. Neurotransmitter uptake and recycling2. Neurometabolic regulation3. Cerebrovascular regulation4. Release of signaling molecules

(“gliotransmitters”)

Tripartite Synapse

Source: Figley & Stroman, 2011, EJN

Glycolysis

Source: Raichle, 2001, Nature

Vasoactive Substances

• substances that cause the vessels to dilate• potassium ions (K+)

– move from intra- to extra-cellular space during synaptic activity

• adenosine– increases with high metabolic activity

• nitric oxide– released by local and distant activation

• gap junctions• calcium (Ca2+)

– triggered by neuronal activation

• dopamine

Information Source: Huettel, Song & McCarthy, 2nd ed.

What about inhibitory synapses?

• GABA = inhibitory neurotransmitter hyperpolarization (IPSP)

• less metabolically demanding than excitatory (glutamatergic) activity

• GABA can be taken up presynaptically rather than recycled through astrocytes

• Therefore, neurotransmission at inhibitory synapses likely influences the BOLD signal less than at excitatory synapses

Not Just Neurons

Leopold, 2009, based on data of Sirotin & Das, 2009, NatureSirotin & Das, 2009• awake macaque monkey sees tiny light in dark room

– red: keep tight fixation; green: relax– timing of red-green is periodic

• measure blood flow in area of peripheral visual cortex – away from foveal representation of fixation point– on some trials visual stimuli were presented to activate the measured area

Non-Neuronal Effects

Leopold, 2009, based on data of Sirotin & Das, 2009, Nature

Sirotin & Das, 2009• two components to blood flow in visual cortex (V1)

1. related to neuronal responses to visual stimuli

2. related anticipation of neural events

Properties of Predictive Response

• response follows expected trial timing– when trial timing is changed, monkey performs correctly but

this response persists for a few trials

• occurs even without stimulation• correlated with pupil diameter

– is it just general arousal?

• visual cortex response does not occur with predictive sequence of auditory events – suggests it’s more regionally specific than general arousal

• occurs in arterial signal

Stimulus to BOLD

Source: Arthurs & Boniface, 2002, Trends in Neurosciences

Gradient Echo vs. Spin Echo

Gradient Echo• high SNR• strong contribution of vessels

Spin Echo• lower SNR• weaker contribution of vessels

Source: Logothetis, 2008, Nature

The Concise SummaryWe sort of understand this

(e.g., psychophysics, neurophysiology)

We sort of understand this (MR Physics)We’re *&^%$#@ clueless here!

Is the fMRI Sky Falling?

Don’t Panic

• 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)– maybe anticipatory changes in blood flow are interesting too

Section 3Spatial Limits of fMRI

fMRI in the Big Picture

What Limits Spatial Resolution

• noise– smaller voxels have lower SNR

• head motion– the smaller your voxels, the more contamination head motion

induces

• temporal resolution– the smaller your voxels, the longer it takes to acquire the same

volume• 4 mm x 4 mm at 16 slices/sec• OR 1 mm x 1 mm at 1 slice/sec

• vasculature– depends on pulse sequences

• e.g., spin echo sequences reduce contributions from large vessels

– some preprocessing techniques may reduce contribution of large vessels (Menon, 2002, MRM)

Ocular Dominance Columns

• Columns on the order of ~0.5 mm have been observed with fMRI

Submillimeter Resolution

Goenze, Zappe & Logothetis, 2007, Magnetic Resonance Imaging• anaesthetized monkey; 4.7 T; contrast agent (MION)• ~0.3 x 0.3 x 2 mm

Gradient EchoFunctional

(superficial activationincludes vessels)

Spin EchoFunctional(activation

localized to Layer IV)

Spin EchoAnatomical

Gradient EchoAnatomical

vein

Stria of Gennari(Layer IV)

EXCEPT when the activated region does not fill the voxel (partial voluming)

Voxel Size

3 x 3 x 6= 54 mm3

e.g., SNR = 100

3 x 3 x 3= 27 mm3

e.g., SNR = 71

2.1 x 2.1 x 6= 27 mm3

e.g., SNR = 71

isotropic

non-isotropic

non-isotropic

In general, larger voxels buy you more SNR.

Partial Voluming

• The fMRI signal occurs in gray matter (where the synapses and dendrites are)

• If your voxel includes white matter (where the axons are), fluid, or space outside the brain, you effectively water down your signal

fMRI for Dummies

Partial Voluming

This voxel contains mostly gray matter

This voxel contains mostly white matter

This voxel contains both gray and white matter. Even if neurons within the voxel are strongly activated, the signal may be washed out by the absence of activation in white matter.

Partial voluming becomes more of a problem with larger voxel sizes

Worst case scenario: A 22 cm x 22 cm x 22 cm voxel would contain the whole brain

Partial volume effects: The combination, within a single voxel, of signal contributions from two or more distinct tissue types or functional regions (Huettel, Song & McCarthy, 2004)

The Initial Dip

• The initial dip seems to have better spatial specificity• However, it’s often called the “elusive initial dip” for a reason