Shackman Psyc210 Module09 IntermediatePhenotypesImaging2 030515

Stable individual differences in temperament and personality reflect variation in the structure and function of the brain

PSYC 612

Tools for measuring the brain:

What You Can and Cant Do With

EEG and MRI

AJ Shackman 13 October 2014

Just As People Differ, Brains Differ

Individual maps from 17 subjects These subjects are all supposed to be "the same Activation blobs are common, but strength (relative to noise) varies a lot

Just As People Differ, Brains Differ

Neurons video (4 min) https://www.youtube.com/watch?v=GIGqp6_PG6k

There are many ways to measure the brain

Resolving Time & Space

Temporal ResoluBon

SpaB

al ResoluB

on

Resolving Time & Space

Temporal ResoluBon

SpaB

al ResoluB

on

Resolving Time & Space

Temporal ResoluBon

SpaB

al ResoluB

on

Note: Techniques also dier in the amount of the brain they can survey at one Bme

Whadya Mean by Spatial Resolution?

Temporal resoluBon is analogous

EEG/ERP

Hans Berger 1924

EEG (0:44 to 4:21) https://www.youtube.com/watch?v=YUAPUoV56gM

See also Jackson & Bolger Psychophysiol 2014

How do you compute ERPs?

You compute the average potenAal (voltage) evoked by a parAcular event

Strengths & Weaknesses of EEG/ERP Strengths Cheap

EEG can be used in more places than fMRI/PET, as these techniques require bulky and immobile equipment (fMRI requires the use of a mulN-ton magnet in a shielded room) Millisecond temporal resoluNon RelaNvely tolerant of subject movement, unlike most other neuroimaging techniques (kids!)

Silent, which allows for beUer study of the responses to auditory sNmuli. Does not aggravate claustrophobia, unlike fMRI

Does not involve exposure to radioligands, unlike positron emission tomography Weaknesses Poor spaNal resoluNon, parNcularly for deep structures CorrelaNonal (not mechanisNc), like all brain imaging techniques

Strengths & Weaknesses of EEG/ERP Strengths Cheap

EEG can be used in more places than fMRI/PET, as these techniques require bulky and immobile equipment (fMRI requires the use of a mulN-ton magnet in a shielded room) Millisecond temporal resoluNon RelaNvely tolerant of subject movement, unlike most other neuroimaging techniques (kids!)

Silent, which allows for beUer study of the responses to auditory sNmuli. Does not aggravate claustrophobia, unlike fMRI

Does not involve exposure to radioligands, unlike positron emission tomography Weaknesses Poor spaNal resoluNon, parNcularly for deep structures CorrelaNonal (not mechanisNc), like all brain imaging techniques

Strengths & Weaknesses of EEG/ERP Strengths Cheap

EEG can be used in more places than fMRI/PET, as these techniques require bulky and immobile equipment (fMRI requires the use of a mulN-ton magnet in a shielded room) Millisecond temporal resoluNon RelaNvely tolerant of subject movement, unlike most other neuroimaging techniques (kids!)

Silent, which allows for beUer study of the responses to auditory sNmuli. Does not aggravate claustrophobia, unlike fMRI

Does not involve exposure to radioligands, unlike positron emission tomography Weaknesses Poor spaNal resoluNon, parNcularly for deep structures CorrelaNonal (not mechanisNc), like all brain imaging techniques

Strengths & Weaknesses of EEG/ERP Strengths Cheap

EEG can be used in more places than fMRI/PET, as these techniques require bulky and immobile equipment (fMRI requires the use of a mulN-ton magnet in a shielded room) Millisecond temporal resoluNon RelaNvely tolerant of subject movement, unlike most other neuroimaging techniques (kids!)

Silent, which allows for beUer study of the responses to auditory sNmuli. Does not aggravate claustrophobia, unlike fMRI

Does not involve exposure to radioligands, unlike positron emission tomography Weaknesses Poor spaNal resoluNon, parNcularly for deep structures CorrelaNonal (not mechanisNc), like all brain imaging techniques

Strengths & Weaknesses of EEG/ERP Strengths Cheap

EEG can be used in more places than fMRI/PET, as these techniques require bulky and immobile equipment (fMRI requires the use of a mulN-ton magnet in a shielded room) Millisecond temporal resoluNon RelaNvely tolerant of subject movement, unlike most other neuroimaging techniques (kids!)

Silent, which allows for beUer study of the responses to auditory sNmuli. Does not aggravate claustrophobia, unlike fMRI

Does not involve exposure to radioligands, unlike positron emission tomography Weaknesses Poor spaNal resoluNon, parNcularly for deep structures CorrelaNonal (not mechanisNc), like all brain imaging techniques

Strengths & Weaknesses of EEG/ERP Strengths Cheap

EEG can be used in more places than fMRI/PET, as these techniques require bulky and immobile equipment (fMRI requires the use of a mulN-ton magnet in a shielded room) Millisecond temporal resoluNon RelaNvely tolerant of subject movement, unlike most other neuroimaging techniques (kids!)

Silent, which allows for beUer study of the responses to auditory sNmuli. Does not aggravate claustrophobia, unlike fMRI

Does not involve exposure to radioligands, unlike positron emission tomography Weaknesses Poor spaNal resoluNon, parNcularly for deep structures CorrelaNonal (not mechanisNc), like all brain imaging techniques

Strengths & Weaknesses of EEG/ERP Strengths Cheap

EEG can be used in more places than fMRI/PET, as these techniques require bulky and immobile equipment (fMRI requires the use of a mulN-ton magnet in a shielded room) Millisecond temporal resoluNon RelaNvely tolerant of subject movement, unlike most other neuroimaging techniques (kids!)

Silent, which allows for beUer study of the responses to auditory sNmuli. Does not aggravate claustrophobia, unlike fMRI

Does not involve exposure to radioligands, unlike positron emission tomography Weaknesses Poor spaNal resoluNon, parNcularly for deep structures CorrelaNonal (not mechanisNc), like all brain imaging techniques

Strengths & Weaknesses of EEG/ERP Strengths Cheap

EEG can be used in more places than fMRI/PET, as these techniques require bulky and immobile equipment (fMRI requires the use of a mulN-ton magnet in a shielded room) Millisecond temporal resoluNon RelaNvely tolerant of subject movement, unlike most other neuroimaging techniques (kids!)

Silent, which allows for beUer study of the responses to auditory sNmuli. Does not aggravate claustrophobia, unlike fMRI

Does not involve exposure to radioligands, unlike positron emission tomography Weaknesses Poor spaNal resoluNon, parNcularly for deep structures CorrelaNonal (not mechanisNc), like all brain imaging techniques

Strengths & Weaknesses of EEG/ERP Strengths Cheap

EEG can be used in more places than fMRI/PET, as these techniques require bulky and immobile equipment (fMRI requires the use of a mulN-ton magnet in a shielded room) Millisecond temporal resoluNon RelaNvely tolerant of subject movement, unlike most other neuroimaging techniques (kids!)

Silent, which allows for beUer study of the responses to auditory sNmuli. Does not aggravate claustrophobia, unlike fMRI

Does not involve exposure to radioligands, unlike positron emission tomography Weaknesses Poor spaNal resoluNon, parNcularly for deep structures CorrelaNonal (not mechanisNc), like all brain imaging techniques

Strengths & Weaknesses of EEG/ERP Strengths Cheap

EEG can be used in more places than fMRI/PET, as these techniques require bulky and immobile equipment (fMRI requires the use of a mulN-ton magnet in a shielded room) Millisecond temporal resoluNon RelaNvely tolerant of subject movement, unlike most other neuroimaging techniques (kids!)

Silent, which allows for beUer study of the responses to auditory sNmuli. Does not aggravate claustrophobia, unlike fMRI

Does not involve exposure to radioligands, unlike positron emission tomography Weaknesses Poor spaNal resoluNon, parNcularly for deep structures CorrelaNonal (not mechanisNc), like all brain imaging techniques

Magnetic Resonance Imaging (MRI) Scanner

Skip video clip

MRI Intro Video (6 min; stop at 5:55) https://www.youtube.com/watch?v=BmQR57V5TVU

How Does MRI Work?

You will not be responsible for the details

structural MRI vs. functional MRI

MRI fMRI

One 3D volume

series of 3D volumes (i.e., 4D data) (e.g., every 2 sec for 5 mins)

high spatial resolution

(1 mm)

lower spatial resolution (~3 mm)

structural MRI vs. functional MRI

fMRI Does NOT Measure Neuronal Activity

It Measures Blood Oxygenation

(Details Not Important)

How do we estimate fMRI activation?

Conceptually Similar to ERPs

Events (Dog)

Events (Dog)

fMRI Signal

Events (Dog)

fMRI Signal

Average Response to Event (in arbitrary units!)

But complex psychological phenomena (T&P) reect the

coordinated acAvity of distributed neural circuits

How do we assess the connecAvity of

funcAonal circuits in the brain?

Dummies Guide to Functional Connectivity

fMRI Caveats

fMRI Caveats Measuring blood oxygenaNon not neurons!

Although recent work suggests that, unlike EEG, the BOLD signal is strongly correlated with neuronal spiking (Lima et al J Neuro 2014)

Not enNrely clear whether fMRI signal reects neuronal acNvaNon

We also know, however, that the fMRI signal is o6en unable to dieren;ate between func;on-specic processing and neuromodula;on, between boAom-up and top-down signals, and it may occasionally confuse excita;on and inhibi;on (Logothe;s, 2008). [LogotheNs Neuroimage 2012]

Not clear whether funcNonal connecNvity (cross-correlaNon of Nme-series) reects true coupling between regions (e.g., regulaNon or the transmission of informaNon)

fMRI Caveats Measuring blood oxygenaNon not neurons!

Although recent work suggests that, unlike EEG, the BOLD signal is strongly correlated with neuronal spiking (Lima et al J Neuro 2014)

Not enNrely clear whether fMRI signal reects neuronal acNvaNon

We also know, however, that the fMRI signal is o6en unable to dieren;ate between func;on-specic processing and neuromodula;on, between boAom-up and top-down signals, and it may occasionally confuse excita;on and inhibi;on (Logothe;s, 2008). [LogotheNs Neuroimage 2012]

Not clear whether funcNonal connecNvity (cross-correlaNon of Nme-series) reects true coupling between regions (e.g., regulaNon or the transmission of informaNon)

fMRI Caveats Measuring blood oxygenaNon not neurons!

Although recent work suggests that, unlike EEG, the BOLD signal is strongly correlated with neuronal spiking (Lima et al J Neuro 2014)

Not enNrely clear whether fMRI signal reects neuronal acNvaNon

We also know, however, that the fMRI signal is o6en unable to dieren;ate between func;on-specic processing and neuromodula;on, between boAom-up and top-down signals, and it may occasionally confuse excita;on and inhibi;on (Logothe;s, 2008). [LogotheNs Neuroimage 2012]

Not clear whether funcNonal connecNvity (cross-correlaNon of Nme-series) reects true coupling between regions (e.g., regulaNon or the transmission of informaNon)

Strengths & Weaknesses of fMRI Strengths

Very good spaNal resoluNon Non-invasive (no radiaNon) Common at most universiNes and larger hospitals

Weaknesses

Poor temporal resoluNon (seconds) Noisy, unnatural Expensive ($600/hour @ UMCP) Requires parNcipants to remain completely sNll! Neuronal source of signal is not enNrely clear (but correlated with spiking)

CorrelaNonal, not causal

Strengths & Weaknesses of fMRI Strengths

Very good spaNal resoluNon Non-invasive (no radiaNon) Common at most universiNes and larger hospitals

Weaknesses

Poor temporal resoluNon (seconds) Noisy, unnatural Expensive ($600/hour @ UMCP) Requires parNcipants to remain completely sNll! Neuronal source of signal is not enNrely clear (but correlated with spiking)

CorrelaNonal, not causal

Strengths & Weaknesses of fMRI Strengths

Very good spaNal resoluNon Non-invasive (no radiaNon) Common at most universiNes and larger hospitals

Weaknesses

Poor temporal resoluNon (seconds) Noisy, unnatural Expensive ($600/hour @ UMCP) Requires parNcipants to remain completely sNll! Neuronal source of signal is not enNrely clear (but correlated with spiking)

CorrelaNonal, not causal

Some Things Are Just Plain Hard to Do with fMRI

Anything requiring subject to speak or move One word or sound can be OK Requires censoring out data during subject speech jaw moNon is bad for images

Anything that uses subtle sounds (music, language)

Scanner is very loud Workaround: silent period between scans

Very long duraBon tasks (learning; drug infusion)

Hard to tell long acNvaNon changes from MRI signal drijing up or down (e.g., head drij)

Not impossible, but requires special analysis or pulse sequence (ASL)

Some Things Are Just Plain Hard to Do with fMRI

Anything requiring subject to speak or move One word or sound can be OK Requires censoring out data during subject speech jaw moNon is bad for images

Anything that uses subtle sounds (music, language)

Scanner is very loud Workaround: silent period between scans

Very long duraBon tasks (learning; drug infusion)

Hard to tell long acNvaNon changes from MRI signal drijing up or down (e.g., head drij)

Not impossible, but requires special analysis or pulse sequence (ASL)

Some Things Are Just Plain Hard to Do with fMRI

Anything requiring subject to speak or move One word or sound can be OK Requires censoring out data during subject speech jaw moNon is bad for images

Anything that uses subtle sounds (music, language)

Scanner is very loud Workaround: silent period between scans

Very long duraBon tasks (learning; drug infusion)

Hard to tell long acNvaNon changes from MRI signal drijing up or down (e.g., head drij)

Not impossible, but requires special analysis or pulse sequence (ASL)

Some Things Are Just Plain Hard to Do with fMRI

Anything requiring subject to speak or move One word or sound can be OK Requires censoring out data during subject speech jaw moNon is bad for images

Anything that uses subtle sounds (music, language)

Scanner is very loud Workaround: silent period between scans

Very long duraBon tasks (learning; drug infusion)

Hard to tell long acNvaNon changes from MRI signal drijing up or down (e.g., head drij)

Not impossible, but requires special analysis or pulse sequence (ASL)

Reality Check: Whats Lurking in a Voxel

Reality Check: Whats Lurking in a Voxel A complex mesh of vessels, dierent kinds of neurons

and glia; all contribuAng to the fMRI signal

A single voxel contains

* 5.5 million neurons * 22 km of dendrites * 220 km of axons * 4 x 1010 (40B) synapses

Reality Check: Whats Lurking in a Voxel A complex mesh of vessels, dierent kinds of neurons

and glia; all contribuAng to the fMRI signal

A single voxel contains

* 5.5 million neurons * 22 km of dendrites * 220 km of axons * 4 x 1010 (40B) synapses

40B synapses is a lot

x 4

This aint just theoreAcal

Pare & Duvarci Curr Opin Neurobiol 2012

e.g., Amygdala contains complex microcircuits that gate and modulate

the expression of fear and anxiety

DETAILS ARE NOT IMPORTANT!

These microcircuits can not be resolved using fMRI

CriBcal Thinking QuesBons

Please pick 2 of 3

CriBcal Thinking QuesBons 1. Watch Joe LeDouxs BigThink interviews: hUps://www.youtube.com/watch?v=n8hpf4dC8H0 hUps://www.youtube.com/watch?v=jZ3_-Z3Jycw

-OR- 2. Watch June Grubers interview of Tor Wager: hUps://www.youtube.com/watch?v=z-8XCK9P430 Briey describe the 2 most important or interes;ng things that you learned and 2 key challenges for future research.

CriBcal Thinking QuesBons 3. The brains structural connectome. Briey describe: A) What is diusion tensor imaging (DTI) and how can it be used to map the brains wiring?

For some Nps, see these handy references - Overview @ hUp://www.diusion-imaging.com/2009/05/lectures-on-dN-basics-and-analysis.html

- Wiki @ hUps://en.wikipedia.org/wiki/Diusion_MRI

- Film clip @ hUps://www.youtube.com/watch?v=i6ZqP04RhNI

- Film clip @ hUps://www.youtube.com/watch?v=TjbkepJRrZc

B) How can DTI be used to understand a facet of T&P?

- For some Nps, see - hUp://www.annualreviews.org/doi/full/10.1146/annurev-clinpsy-032210-104507

The End (No Review QuesBons)

Material to Consider IncorporaBng as a CriBcal Thinking Q

in the Future

Extra Slides

EEG Intro Video, Version #1 (1:12 to end)

https://www.youtube.com/watch?v=kGuayTwqieM

Summary EEG fMRI

Strengths excellent temporal excellent spatial

Weaknesses poor spatial poor

temporal

Measuring? electrical currents oxygen in

blood

Upper wave in each pair = signal; lower = residual noise est

DTI Video (60 sec)

https://www.youtube.com/watch?v=i6ZqP04RhNI And this one is 30 sec https://www.youtube.com/watch?v=TjbkepJRrZc

fMRI Experiment Stages: Anatomicals 4) 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

-79-

Some FMRI Data

Left = decent looking single subject activation map From 300 s of data (150 time points)

Right = data time series that gives activation map This is good data [strong activation, little head movement]

Hemodynamic Response FuncNon (HRF)

Time course of fMRI is ~6 seconds (blood ow is slow!)

Dip because neurons begin to extract oxygen before oxyHb arrives

Increase in oxyHb because new oxyHb has overcompensated

Back to baseline because neurons extract oxygen so oxyHb becomes deoxyHb

10 seconds

4-6 seconds

FMRI Experiment

F

S

F

F

F

S

S

S

S

Faces

Scrambled

Faces

fMRI Analysis

Condition I Condition S (If significant)

Voxel

I = Intact face

S = Scrambled face

fMRI relies on subtracNon method

fMRI Experiment Stages: Functionals 5) 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

Reverse Engineering Marion Barry

Confusions & Distinctions Brain vs. Mind

Neuroscience vs. Cognitive science Mass level vs. Micro-circuitry

Connecting blobs to cell-level actions? Excitation & Inhibition both consume energy

What does active mean? Active vs. Necessary (e.g., lesion studies) Modulated here? Or there? MVPA vs. Specificity Resting State: Function vs. Physiology

-86-

BART Procedure

Computer simulated sequence of balloons $0.05 per balloon pump (in temporary bank) Must hit collect $$$ button to earn money Balloon explosion = temporary money lost Explosion point for balloons (1 128) Earnings after session in gift certificates

Earn

ings

($)

MIN MAX

Not Just Psychiatry Other syndromes with complex eBology and overlapping symptoms E.g., Diabetes, hypercholesterolemia, hypertension

Clinical presentaNon: PaNent complains of faNgue; whats the Dx?

DierenNal Diagnosis: challenges (stress test) and assays to determine a primary eNology and Dx

E.g., glucose tolerance test, serum cholesterol, BP

Similar symptoms, dierent endophenotypes, dierent geneNc underpinnings

GoUesman & Gould Amer J Psychiatry 2003

Promises to enhance: EAology (cause)

What is the proximal biological cause(s) of the disorder/T&P?

Nosology (Dx) Unlike many physical diseases, psychiatric disorders/T&P are dened by symptoms not

pathophysiology; understanding the eNology would lead to a restructuring of nosology (or the Big 3 or 5 Factors)

Challenging to know how disorders/factors should be clumped together what if symptoms reect similar vs. dierent causes?

E.g., growing biological evidence that SZ and BPD are related Prognosis

If we understood the eNology, we could beUer predict the likely outcome (cf. MoU) Treatment (Personalized or Precision Medicine)

Most disorders resist treatment; extant treatments work for a subset of paNents If we understood the eNology, we could target treatments, enhancing cost/benet

Early intervenAon for high-risk populaAons (e.g., cholesterol test staAns) Screen idenNfy intervene before the deleterious secondary consequences emerge

The Promise

MRI vs. fMRI

fMRI Simplified

Time

~2s

~ 5 min

Time

fMRI Signal

Intensity

ROI Time Course

Condition

fMRI Setup