huett el

Association for Consumer Research Scott Huettel, Duke University

Neuroimaging Methods

Scott HuettelBrain Imaging and Analysis CenterDepartment of PsychiatryDuke University

All uncredited figures are from Huettel, Song, & McCarthy (2004). Functional Magnetic Resonance Imaging.

This presentation, save for credited figures from other sources, is copyrighted by Scott Huettel (2006).


Methods for Creating Images of (Human) Brain Function

1. Electroencephalography (EEG)2. Magnetoencephalography (MEG)3. Transcranial Magnetic Stimulation (TMS)4. Positron Emission Tomography (PET)5. Structural Magnetic Resonance Imaging (MRI)6. Functional Magnetic Resonance Imaging (fMRI)

7. Examples: Neuroimaging of Choice


The Cardinal Principles

Functional neuroimaging comprises methods for mapping information processing within the brain.

All functional neuroimaging is limited by two factors: the physical properties of the recording system and the

physiological constraints of the brain.

Images of brain activity only have meaning when acquired using the correct experimental design and interpreted using

the correct analyses.


1. Electroencephalography (EEG)


From Cognition to Neuron


Electrophysiological Recording

Brain

Amplifier Bank

Electrode Array (e.g., n = 64)


EEG recordings by Hans Berger (c. 1925-1935)


Event-Related Potentials (ERPs)

TIME (ms)

+ V

OLT

AG

E -

TIME (in 20ms Intervals)

from Khoe et al. (2004)

Using selective averaging across trials, ERPs have exquisite temporal resolution (but coarse spatial resolution)


2. Magnetoencephalography (MEG)


Magnetoencephalography (MEG)

from Woldorff et al. (1999)Courtesy 4D Neuroimaging


3. Transcranial Magnetic Stimulation (TMS)


Transcranial Magnetic Stimulation (TMS)

TMS allows transient* and safe* disruption of local neuronal activity, in effect creating reversible lesions.


4. Positron Emission Tomography (PET)


PET Scanning: Principles


http://www.idac.tohoku.ac.jp/dep/nmr/pet1.jpg

Positron Emission Tomography

Image from Utah Center for Advanced Imaging Research

Cyclotron Radio-isotope (FDG)

Image Scanner

http://www.med-ed.virginia.edu/courses/rad/PETCT/Emission.html


PET: Strengths and Limitations

• Strengths– Uses a simple physiological mechanism– Provides absolute, quantitative data– Allows imaging of anything that can be tagged

• Limitations– Poor temporal resolution (many minutes)– Poor spatial resolution (several centimeters)– Requires injection of radioactive material


5. Structural MRI


MRI Scanning Hardware

“Imaging”(Weak Gradient

Magnetic Fields)“Magnetic”(Strong Static Magnetic Field)

“Resonance”(Radiofrequency Energy)


Structural MRI


6. Functional MRI (fMRI)


Fact #1: Energy is supplied to the brain via the vascular system

Glucose image from NYU Library of 3-D Molecular Structures

Glucose

Hemoglobin

Hemoglobin image from Pittsburgh Supercomputing CenterFrom Duvernoy et al., 1982

(Oxygen)


From Mandeville et al., 1999

Fact #2: More hemoglobin is supplied than needed, causing a decrease in deoxygenated hemoglobin.


Fact #3: Deoxygenated hemoglobin reduces some forms (T2*) of MR signal.

Baseline

Task

Blood-Oxygenation-Level Dependent Contrast

(BOLD Contrast)


From Cognition to Neuron to fMRI


fMRI: Strengths and Limitations

• Strengths– Non-invasive, replicable– Potentially good spatial localization– Common, well-validated technique

• Limitations– Mediocre temporal resolution (seconds)– Complex, highly variable data analyses– Expensive and time-consuming


Neuroimaging of Decision Preferences

1. Uncertainty: Risk vs. Ambiguity2. Probability: High vs. Low3. Choice: Safe vs. Risky

In all of these cases, there is some derived parameter that is related to the

neuroimaging activation.


Parietal Cortex

Prefrontal cortex

Huettel et al. (2006) Neuron

Dissociable Systems for Risk and Ambiguity

Risky - Certain

Risky - Risky

Ambiguity Preference

Ambiguity preference (1-α) Risk preference (β)

Risk Preference

Ambiguity Preference Risk Preference

Ambiguity preference (1-α) Risk preference (β)


Probability

Preuschoff, Boessarts, & Quartz (2006) Neuron

Huettel et al. (2005) J Neuroscience

Probability of Error


Safe vs. Risky Choice

Insula activation predicts safe choice.

Nucleus accumbens activation predicts risky choice.

Kuhnen & Knutson (2005) Neuron


Summary

• Neuroimaging techniques create maps of brain function.

• The most common approaches measure neuronal activity (EEG, ERP, MEG) or brain hemodynamics (PET, FMRI).

• The neuroimaging approaches relevant for consumer research involve relating neuroimaging data to economic parameters.


Acknowledgments

neuroeconomics.duke.edu

Recommended Readings:• Huettel, Song, & McCarthy (2004). Functional Magnetic Resonance Imaging.• Buxton (2002). Introduction to fMRI.• Luck (2005). An Introduction to the ERP Technique.• Purves et al. (2004). Neuroscience, 3rd Edition.

FMRI education colleagues:• Allen Song (Duke University), Gregory McCarthy (Yale University)

Laboratory members:• Bethany Weber, Dharol Tankersley, John Clithero, Luke Vicens, Lily Kinross-

Wright, Parker Goyer, Jason Chen

huett el

Documents

duke universityassociation

duke universityhttp

duke universitymethods

duke university1

duke university2

duke university3

duke university4

duke universitypet scanning