BRAIN PROCESSES ASSOCIATED WITH COGNITIVE CONTROL
ICON XII, Brisbane, Australia, 7/30/2014
Speakers
Diane Beck (U. of Illinois) The role of feedback in visual processing
Paul Corballis (U. of Auckland) Lateralisation o the ERP reveals neural correlates of
attention, distractor suppression, and visual short term memory
Gabriele Gratton (U. of Illinois) Investigating brain networks in task preparation
Pauline Baniqued (U. of Illinois) A functional and structural view of task-switching
dynamics in ageing
INVESTIGATING BRAIN NETWORKS IN TASK PREPARATION
Gabriele Gratton & Monica Fabiani Psychology Department & Beckman Institute University of Illinois
Cognitive control
Set of operations that prepares the brain to perform a particular cognitive task The same stimulus information can be processed in
different ways
It involves setting up the information processing system so as to weigh information appropriately for the particular task context at hand Notion of “prepared reflexes” (Allport, A&P, 1980)
Interaction between top-down and bottom-up processes
From Dosenbach et al.
PNAS 2007;104:11073-11078
Brain areas involved in cognitive control
From fMRI work
Cingulo-opercular network (black): Long-term goal setting
Dorsal attention network (yellow): Trial-to-trial adaptation
How are these areas related to each other?
Preparation paradigm
Gratton et al., 2009; Baniqued et al., 2013; Leaver et al., submitted; Low et al., in preparation
Recording period
Precue Stimulus
Reaction Stimulus
Precue Stimulus
400 ms
1600 ms
400 ms 400 ms
1600 ms 1600 ms
• Indicates the “rule” or task for that trial
• Use the default or old rule
Little activity
OR
• Use a different rule More activity
• Prompts when and how to perform the task
• Some level of conflict
• Indicates the “rule” or task for that trial
Recording period
1600 ms
EROS: A tool for studying the time course of preparatory activity
Reviews: Gratton & Fabiani, TICS, 2001 Gratton & Fabiani, Frontiers in Human Neuroscience, 2010
EROS (Fast, neuronal)
ms
Signal averaging by stimulus
64 ms
128 ms
sec
Δ C
on
cen
tra
tio
n
NIRS (Slow, hemodynamic)
Derive: [HbO2] [HbR] Signal averaging by block
HbO2
HbR
10 sec
Optical Recording
Study 1 Auditory/Visual
Precue: Auditory Visual
A V
RS: Lft Hand Rt Hand
I O
Conflict: Hear “I” + See “O”
or
Hear “O” + See “I”
Study 2 Global/Local
Precue: Big Little
B L
RS: Lft Hand Rt Hand
S H
Conflict:
Study 3 Left/Right
Precue: Left Right
RS: Lft Hand Rt Hand
S T
Conflict:
S + T or
T + S
Study 4 Manual/Vocal
Precue: Hand Voice
H V
RS: Left Right
L R
Conflict:
Response
modality
unknown without
precue
H H H H
H
H H H H
H
H H H H
S S
S S
S S S S S
S S
S S
EROS: N=16/study
Task-general EROS activity
-3.00 0.00 +3.00 Z-scores
Time (ms)
230-330
345-560
Switch vs. No-Switch
Left/Right Manual/Vocal Auditory/Visual Global/Local
-3.00 0.00 +3.00 Z-scores
Local Global 280 ms 255 ms
130 ms 230 ms
Visual Auditory Left Right 170 ms 170 ms
560 ms Manual Vocal
510 ms
Task-specific EROS activity
Preparation for Global/Local Processing
Leaver et al., submitted
Behavioral Results
480
490
500
510
520
530
540
550
560
570
Local Global
RT
(m
s)
Congruent
Incongruent
Local task is harder than global task More conflict in local than in global task
Region of Interest (left)
Ph
ase
de
lay
(ps)
-100 100 300 500 700 900
Time (ms)
BA8
BA9
BA7
BA40
BA19
BA37
Interval of Interest
200-300 ms
300-400 ms
400-500 ms
500-600 ms
-3 0 +3 Z score
600-700 ms
Local - Global
0 ms 128 ms 307 ms 358 ms 51 ms
Y=-11 Z=43 z=3.091
Y=-26 Z=21 z=3.809
Y=-68 Z=43 z=4.383
Y=-11 Z=43 z=2.785
0 ms 153 ms 256 ms 333 ms 26 ms
Y=12 Z=29 z=2.882
Y=-26 Z=33 z=3.812
Y=-11 Z=23 z=3.433
Y=-21 Z=-18 z=2.911
Cross-correlation analysis
Left Hemisphere
Right Hemisphere
-10
-8
-6
-4
-2
0
2
4
6
8
10
-60 -40 -20 0 20 40 60
Loca
l Pre
par
atio
n -
Glo
bal
Pre
par
atio
n
(ps)
Local Conflict Cost - Global Conflict Cost (ms)
BA9
BA7
Preparation helps reduce conflict
r=-.52, p<.05
r=-.57, p<.05
How do top-down processes influence bottom-up processing?
A flourishing of papers in the last five years indicate that processing of sensory stimuli is influenced by the amplitude and phase of oscillatory activity (alpha) in sensory cortex
E.g., Mathewson et al., JoN, 2009
Do attentional networks influence these oscillatory activities?
E.g., Thut & Miniussi, TICS, 2009
Meta-contrast masking
+
Fixation 247 ms
Blank 400 ms
Target 11.7 ms
ISI 46.8 ms
Mask 23.4 ms
Response ~1520 ms
Trial 2220 ms
1o
1o
2o
+
.5o
Mathewson et al., JoN, 2009
Brain states and detection
Averaged evoked potential for detected and undetected targets
Probability of detection for trials with large and small alpha power
Probability of detection for trials with alpha phase in “high” and “low” mode
Mathewson et al., JoN, 2009
77
75
73
71
69
67
65
Low High α Power
n. s. ** α Phase
45-225°
225-45°
77
75
73
71
69
67
65
α Power Quartile 1 3 4 2
De
tect
ion
Ra
te (
%)
-0.3
0.6
0.3
0
Vo
lta
ge
(μV
)
Time (ms) -100 -50 50 100
Target onset
0
All
Detected
Undetected
De
tect
ion
Ra
te (
%)
Cortical excitability and alpha oscillations
Sensory Input
High
Low
Small Amplitude Alpha Large Amplitude Alpha
Cortical Excitability Unconscious
Conscious
Brain activity prior to targets EROS alpha power map Detected - Undetected
-332 ms -255 ms
3 -3 0 5
10
15
20
25
30
Time (ms)
Fre
qu
en
cy (
Hz)
-1000 -600 -400 -200 0 200 400
5
10
15
20
25
30
Fixation Onset
Target Onset
-1
-0.8
-0.6
-0.4
-0.2
0
0.2
0.4
0.6
0.8
1
Power (dB) .75
-.75
EEG time frequency Detected - Undetected
Time-locked activity Detected vs. Undetected
Mathewson, Beck, Ro, Fabiani, & Gratton, 2014
-20 -10 -5 5 10 20
EROS 10 Hz Alpha Detected - Undetected
-255 ms
3 -3 0
500 1000 1500 2000 2500
50
100
150
200
250
300
Backward-Lagged Cross Correlation Seed in cuneus (alpha)
Alpha
D
-205 ms -179 ms -154 ms -128 ms -102 ms -77 ms -51 ms -26 ms 0 ms
lags
mPFC SFG IPS FEF
Alpha in Cuneus
Model
mPFC
SFG
IPS
FEF
Cuneus
3 -3 0
Discussion Interaction between task-general areas (involved in top-down
regulation) and task-specific areas (involved in bottom-up processing)
Task general areas include DAN and CON
Within DAN, frontal areas are activated before parietal ones
What is their respective role?
How are action plans represented here?
Task specific areas include visual, auditory, and motor networks During preparation, they are activated during the foreperiod
Regulation of these areas may involve up- or down-regulation of rhythmic activity
The phase of the rhythmic activity may be involved in gating information processing
It may represent the excitability of particular cortical regions
Acknowledgements
CNL (present): Monica Fabiani , Ph.D, Pauline Baniqued Courtney Burton Antonio Chiarelli, Ph.D. Antoine De Jong Mark Fletcher Tania Kong Christina Koury Kathy Low, Ph.D. Ed Maclin, Ph.D. Kyle Mathewson, Ph.D. Nils Schneider-Garces Chin Hong Tan John Walker Ben Zimmerman
Former students and postdocs Jason Agran, Vanderbilt Chandramallika Basak, UT Dallas Carrie Brumback, UC-Irvine Bruce Bartholow, UNC Paul Corballis, U of Auckland Corby Dale, UC-SF Brian Gordon, Washington Univ. Erika Henry, UMC Echo Leaver, Salisbury Univ. Nate Parks, U Arkansas Melanie Pearson, Emory Elena Rykhlevskaia, Stanford Univ. Jeff Sable, Christian Brothers Univ. Eunsam Shin, Korea Chun Yu Tse, CUHK Emily Wee, UIUC Christopher Whalen, Champaign Eddie Wlotko, Tufts
Collaborators Renee Baillargeon, UIUC Diane Beck, UIUC Gary Dell, UIUC Matt Dye, UIUC Kara Federmeier, UIUC Cynthia Fisher, UIUC Susan Garnsey, UIUC Enrico Gratton, UC Irvine Dan Hyde, UIUC Vicki Kazmerski, PSU-Erie Art Kramer, UIUC Alejandro Lleras, UIUC Eddie McAuley, UIUC Trevor Penney, NSU Tony Ro, CUNY Brad Sutton, UIUC Fabrice Wallois, U de Picardie
Extramural Funding Agencies Abbott/CNLM ABMRF Carle Foundation DARPA McDonnell-Pew Missouri ADRD NIA NIBIB NIMH NRCC ONR
Thank you
Dynamics of cognitive control
How does cognitive control operate?
What are the relationships between different cognitive control regions?
How do they influence each other?
How do they influence perceptual (bottom-up) areas?
What happens in the perceptual areas that influences stimulus processing?
Example: Conflict effects
Gratton et al., JEP: Gen., 1992
Feat.
Anal.
Conj.
Anal.
p(Feature Analysis) = acc(comp) – acc(incomp) p(Conjunction Analysis) = 0.5*(acc(comp)+acc(incomp))
HHHHH SSHSS
SSSSS HHSHH
Compatible Incompatible
Respond Left/Right
Respond Right/Left
0.00
0.02
0.04
0.06
0.08
0.10
Noise Type
Error Rate
Compatible Incompatible
460
480
500
520
540
560
Noise Type
Reaction Time (ms)
Compatible Incompatible
Conflict adaptation: Effects of changes in expectation
Exp. 1: Sequential effect Exp. 2: Blocked probability effect
Exp. 3: Cued probability effect
Large P3 Small P3
Gratton et al., JEP: General, 1992
* *
** **
Interpreting conflict adaptation
Strategy selection can be influenced by varying expectancy for compatible and incompatible noise
Precue A (predict compatible)
Precue B (predict incompatible) HHHHH SSHSS
SSSSS HHSHH
Compatible Incompatible
Feature Anal.
Conjunct. Anal.
Gratton et al., JEP: General, 1992
Predict compatible Predict incompatible
LRP
Comp.
Inc.
Summary
In RT tasks, subjects prepare for incoming stimuli by preparing appropriate stimulus-response plans ideomotor function
The front0-parietal network (FPN) exerts an important role in preparation Activation occurs first in frontal and then in parietal areas
Activation in FPN precedes that occurring in task-specific areas
The amount of preparatory activity is predictive of subsequent behavioral advantages
EROS provides a tool for tracking the time course of preparatory activity