Attentive tracking of multiple objects modulates neuronal ...jude/VSS_2006_MOT_MONKEY.pdf · Pause Resume F r e q u e n c y (H z) (2 of 4 Attended – Distracter) – (1 of 4 Attended

Attentive tracking of multiple objects modulates neuronalresponses in area V4 of macaque.

879

Conclusions

Multiple-Object Tracking Task:

1) Firing rate is enhanced (~18%) in V4 for tracked stimuli.2) The fano factor is decreased (~7.5%) for tracked stimuli.3) The modulation of rate is similar in tracking 1 or 2 items.4) The signal to noise ratio is better in tracking 1 item though.5) Both spike and LFP power spectra are modulated similarly.6) For tracked items, low frequency(5-25Hz) power decreases and high frequency (35-95Hz) power increases.

Jude F. Mitchell, Kristy A. Sundberg & John H. Reynolds

* Systems Neurobiology Laboratory, The Salk Institute for Biological Studies. Funding provided by NIH Training Grant in Cognitive Neuroscience (J.M.), NSF Graduate Research Fellowship (K.S.), and NEI grant 1R01EY13802 (J.R.).

Are V4 Unit’s Rates Modulated?

4 identical stimuli appear

(200ms)

1 or 2 targets cued by luminance increment

(100ms)

Stimuli move to pause position

(750 ms)

Stimuli Pause(1000 ms)

Stimuli move tofinal position

(750 ms)Cue off(100ms)

Fixation point off: saccadeto ALL targets

RFRF

RFRF

RFRF

RFRFRFRF

RFRFRF

RFRF

Monkeys track either 1 or 2 of 4 identical items. Stimuluscolor and orientation matched the preferences of a singleunit isolated in V4. During Tracking, a single item entered and paused inside the receptive field for 1000ms. Monkeys indicated the tracked items by making a saccade to each at the end of the trial.

Does the Signal to Noise Ratio Improve for Tracked Items?

H74

Do power spectra of spikes andLFPs change for tracked items?

Attended Target Entering Receptive Field in 2 of 4 Tracking

20

40

60

80-0.2

0.0

+0.2

+0.4

Pause Resume

Fre

quency

(Hz)

500 1000 1500 2000 2500 3000 3500Time (ms)

20

40

60

80 -0.2

0.0

+0.2

+0.4

Distracter Entering Receptive Field in 2 of 4 TrackingPause Resume

Fre

quency

(Hz)

500 1000 1500 2000 2500 3000 3500Time (ms)

20

40

60

80 -0.1

0.0

0.1

500 1000 1500 2000 2500 3000 3500Time (ms)

Pause ResumeAttended Response - Distracter Response in 2 of 4 Tracking

Fre

quency

(Hz)

20

40

60

80 -0.1

0.0

0.1

Attended Response - Distracter Response in 1 of 4 TrackingPause Resume

Fre

quency

(Hz)

500 1000 1500 2000 2500 3000 3500Time (ms)

-0.1

0.0

0.120

40

60

80

500 1000 1500 2000 2500 3000 3500Time (ms)

Pause Resume

Fre

quency

(Hz)

(2 of 4 Attended – Distracter) – (1 of 4 Attended – Distracter)


20

40

60

80-0.2

0.0

+0.2

+0.4

Pause Resume

Fre

quency

(Hz)

500 1000 1500 2000 2500 3000 3500Time (ms)


20

40

60

80-0.2

0.0

+0.2

+0.4

-0.2

0.0

+0.2

+0.4

Pause Resume

Fre

quency

(Hz)

500 1000 1500 2000 2500 3000 3500Time (ms)

500 1000 1500 2000 2500 3000 3500Time (ms)

20

40

60

80 -0.2

0.0

+0.2

+0.4


Fre

quency

(Hz)

500 1000 1500 2000 2500 3000 3500Time (ms)

20

40

60

80 -0.2

0.0

+0.2

+0.4

-0.2

0.0

+0.2

+0.4


Fre

quency

(Hz)

500 1000 1500 2000 2500 3000 3500Time (ms)

500 1000 1500 2000 2500 3000 3500Time (ms)

20

40

60

80 -0.1

0.0

0.1

500 1000 1500 2000 2500 3000 3500Time (ms)


Fre

quency

(Hz)

20

40

60

80 -0.1

0.0

0.1

-0.1

0.0

0.1

500 1000 1500 2000 2500 3000 3500Time (ms)

500 1000 1500 2000 2500 3000 3500Time (ms)


Fre

quency

(Hz)

20

40

60

80 -0.1

0.0

0.1


Fre

quency

(Hz)

500 1000 1500 2000 2500 3000 3500Time (ms)

20

40

60

80 -0.1

0.0

0.1

-0.1

0.0

0.1


Fre

quency

(Hz)

500 1000 1500 2000 2500 3000 3500Time (ms)

500 1000 1500 2000 2500 3000 3500Time (ms)

-0.1

0.0

0.120

40

60

80

500 1000 1500 2000 2500 3000 3500Time (ms)

Pause Resume

Fre

quency

(Hz)


-0.1

0.0

0.1

-0.1

0.0

0.120

40

60

80

500 1000 1500 2000 2500 3000 3500Time (ms)

500 1000 1500 2000 2500 3000 3500Time (ms)

Pause Resume

Fre

quency

(Hz)


What is the time course of changein the power spectrum of LFPs?

YES

The Fano Factor Decreasesfor Tracked Items

Changes in Power are not Different Between 1 or 2 Target Tracking

4 identical stimuli appear

(200ms)

1 or 2 targets cued by luminance increment

(100ms)

Stimuli move to pause position

(750 ms)

Stimuli Pause(1000 ms)

Stimuli move tofinal position

(750 ms)Cue off(100ms)

Fixation point off: saccadeto ALL targets

RFRF

RFRF

RFRF

RFRFRFRF

RFRFRF

RFRF

Monkeys tracked either 1 or 2 of 4 identical items. Stimuluscolor and orientation were matched to the preferences of anisolated unit in V4. During tracking, an attended or distracteritem entered and paused inside the receptive field for 1000ms. At the end of each trial, the monkey indicated whichitems were tracked by making a saccade to each.

Power spectra were computed using a sliding window of 200mswith a step of 50ms with five Slepian Tapers (W = 6Hz) using Chronux software. Powerwas then normalized at eachfrequency across trial duration and all tracking conditions. Significant attention differences (black dots) were computed, with a Bonferonni correction.

References1. Pylyshyn ZW and Storm RW (1988) Tracking multiple independent targets: evidence for a parallel tracking mechanism. Spat Vis. 3(3):179-97.2. Sears CR and Pylyshyn ZW (2000) Multiple object tracking and attentional processing. Can J Exp Psychol. 54(1):1-14.3. Fries P, Reynolds JH, Rorie AE, and Desimone R (2001) Modulation of oscillatory neuronal synchronization by selective visual attention. Science 291:1560-63.

1 2-10

0

10

20

30

Items Tracked

Fir

ing

Ra

teM

od

ula

tion

1 2-10

0

10

20

30

Items Tracked

Fir

ing

Ra

teM

od

ula

tion

Population Summary:

YES

Firing RateModulation is Similarfor Tracking1 or 2 Items

Population Summary:

0 20 40 60 80 1000.95

1.00

1.05

Frequency (Hz)

No

rma

lize

dP

ow

er

Population Spike Spectra in 1 of 4 Tracking

0 20 40 60 80 100

Frequency (Hz)

Population Spike Spectrain 2 of 4 Tracking

AttendedDistracter

AttendedDistracter

1 SEM 1 SEM

N = 118 N = 118Low High Low High

2 1 2-4

-2

0

2

Items Tracked

Mo

du

latio

no

fP

ow

er

1

LowFreq.

HighFreq.

0 20 40 60 80 100

0.6

0.8

1.0

1.2

Frequency (Hz)

No

rma

lize

dP

ow

er

0 20 40 60 80 100

Frequency (Hz)

Population LFP Spectra in 1 of 4 Tracking

Population LFP Spectrain 2 of 4 Tracking

AttendedDistracter

AttendedDistracter

1 SEM

N = 118Low High

1 SEM

N = 118Low High

-30

-20

-10

0

10

20

30

Mo

du

latio

no

fP

ow

er

2 1 2

Items Tracked1

LowFreq.

HighFreq.

*

0 20 40 60 80 1000.95

1.00

1.05

Frequency (Hz)

No

rma

lize

dP

ow

er


0 20 40 60 80 100

Frequency (Hz)


AttendedDistracter

AttendedDistracter

1 SEM 1 SEM


2 1 2-4

-2

0

2

Items Tracked

Mo

du

latio

no

fP

ow

er

1

LowFreq.

HighFreq.

0 20 40 60 80 1000.95

1.00

1.05

Frequency (Hz)

No

rma

lize

dP

ow

er


0 20 40 60 80 100

Frequency (Hz)


AttendedDistracter

AttendedDistracter

1 SEM 1 SEM


0 20 40 60 80 1000.95

1.00

1.05

Frequency (Hz)

No

rma

lize

dP

ow

er


0 20 40 60 80 100

Frequency (Hz)


AttendedDistracter

AttendedDistracter

1 SEM 1 SEM


2 1 2-4

-2

0

2

Items Tracked

Mo

du

latio

no

fP

ow

er

1

LowFreq.

HighFreq.

2 1 2-4

-2

0

2

Items Tracked

Mo

du

latio

no

fP

ow

er

1

LowFreq.

HighFreq.

0 20 40 60 80 100

0.6

0.8

1.0

1.2

Frequency (Hz)

No

rma

lize

dP

ow

er

0 20 40 60 80 100

Frequency (Hz)



AttendedDistracter

AttendedDistracter

1 SEM

N = 118Low High

1 SEM

N = 118Low High

-30

-20

-10

0

10

20

30

Mo

du

latio

no

fP

ow

er

2 1 2

Items Tracked1

LowFreq.

HighFreq.

*

0 20 40 60 80 100

0.6

0.8

1.0

1.2

Frequency (Hz)

No

rma

lize

dP

ow

er

0 20 40 60 80 100

Frequency (Hz)



AttendedDistracter

AttendedDistracter

1 SEM

N = 118Low High

1 SEM

N = 118Low High

0 20 40 60 80 100

0.6

0.8

1.0

1.2

Frequency (Hz)

No

rma

lize

dP

ow

er

0 20 40 60 80 100

Frequency (Hz)



AttendedDistracter

AttendedDistracter

1 SEM

N = 129Low High

1 SEM

N = 129Low High

-30

-20

-10

0

10

20

30

Mo

du

latio

no

fP

ow

er

2 1 2

Items Tracked1

LowFreq.

HighFreq.

*

-30

-20

-10

0

10

20

30

Mo

du

latio

no

fP

ow

er

2 1 2

Items Tracked1

LowFreq.

HighFreq.

*

YES

BothSpike and LFP Power Spectraare Modulated:High Freq. Power Increases andLow Freq. Power Decreases

Population Summary:

Distracter

Attended

Low Frequency Modulation

Distracter

Attended

High Frequency Modulation

0.5 1 1.50.5

1

1.5

N=129

- 1 of 4 Tracking

- 2 of 4 Tracking

0.5 1 1.50.5

1

1.5

- 1 of 4 Tracking

- 2 of 4 Tracking

N=129

Distracter

Attended

Low Frequency Modulation

Distracter

Attended

High Frequency Modulation

0.5 1 1.50.5

1

1.5

N=129

- 1 of 4 Tracking

- 2 of 4 Tracking

0.5 1 1.50.5

1

1.5

0.5 1 1.50.5

1

1.5

N=129

- 1 of 4 Tracking

- 2 of 4 Tracking

- 1 of 4 Tracking

- 2 of 4 Tracking

0.5 1 1.50.5

1

1.5

- 1 of 4 Tracking

- 2 of 4 Tracking

N=129

0.5 1 1.50.5

1

1.5

- 1 of 4 Tracking

- 2 of 4 Tracking

- 1 of 4 Tracking

- 2 of 4 Tracking

N=129

1000 1500 2000 2500 30000

20

40

60

Time (ms)

Pe

ak

No

rma

lize

dR

es

po

ns

e

1000 1500 2000 2500 30000

20

40

60

Time (ms)

Pe

ak

No

rma

lize

dR

es

po

ns

e

Population Mean in 2 of 4 Tracking

Pause Resume


Pause Resume

AttendedDistracter

AttendedDistracter

N=81* N=81*

1 SEM 1 SEM

1000 1500 2000 2500 30000

20

40

60

Time (ms)

Pe

ak

No

rma

lize

dR

es

po

ns

e

1000 1500 2000 2500 30000

20

40

60

Time (ms)

Pe

ak

No

rma

lize

dR

es

po

ns

e


Pause Resume


Pause Resume

AttendedDistracter

AttendedDistracter

N=81* N=81*

1 SEM 1 SEM

1000 1500 2000 2500 3000

0.8

1

1.2

1.4

1.6

1.8

Time (ms)

1000 1500 2000 2500 3000

0.8

1

1.2

1.4

1.6

1.8

Time (ms)

Sp

ike

Va

ria

nce

/M

ea

n

Population Fano Factor in 2 of 4 Tracking

Pause Resume

Sp

ike

Va

rain

ce/

Me

an


Pause Resume

AttendedDistracter

AttendedDistracter

N=81* N=81*

1 SEM 1 SEM

1000 1500 2000 2500 3000

0.8

1

1.2

1.4

1.6

1.8

Time (ms)

1000 1500 2000 2500 3000

0.8

1

1.2

1.4

1.6

1.8

Time (ms)

Sp

ike

Va

ria

nce

/M

ea

n


Pause Resume

Sp

ike

Va

rain

ce/

Me

an


Pause Resume

AttendedDistracter

AttendedDistracter

N=81* N=81*

1 SEM 1 SEM

0 20 40 600

20

40

60

Population Rate Modulationin 1 of 4 Tracking

Attended

(Pe

ak

No

rma

lize

d)

Distracter(Peak Normalized Firing)

N=118(all units)

Monkey A 18.8%Monkey B 16.8%

0 20 40 600

20

40

60


Attended

(Pe

ak

No

rma

lize

d)


N=118(all units)


-50 0 50

-50

0

50

1 of 4 Percent Modulation

2of4

Perc

entM

odula

tion

Correlation of Rate Modulation

N=118(all units)

Monkey A r = 0.91

Monkey B r = 0.70

0 20 40 600

20

40

60


Attended

(Pe

ak

No

rma

lize

d)


N=118(all units)


0 20 40 600

20

40

60


Attended

(Pe

ak

No

rma

lize

d)


N=118(all units)


0 20 40 600

20

40

60


Attended

(Pe

ak

No

rma

lize

d)


N=118(all units)


0 20 40 600

20

40

60


Attended

(Pe

ak

No

rma

lize

d)


N=118(all units)


-50 0 50

-50

0

50


2of4

Perc

entM

odula

tion

Correlation of Rate Modulation

N=118(all units)

Monkey A r = 0.91

Monkey B r = 0.70

1 2

1

2

Population Fano Factorin 1 of 4 Tracking

Attended

(Sp

ike

Va

ria

nce

/M

ea

n)

Distracter(Spike Variance / Mean)

N=118(all units)


1 2

1

2


Attended

(Sp

ike

Va

ria

nce

/M

ea

n)


N=118(all units)


1 2

1

2

Attended

(Sp

ike

Va

ria

nce

/M

ea

n)




N=118(all units)

1 2

1

2

Attended

(Sp

ike

Va

ria

nce

/M

ea

n)




N=118(all units)

-40 0 40 80-40

0

40

80Monkey A

r = 0.25Monkey B

r = 0.08

Correlation of Fano FactorModulation

N=118(all units)

2of4

Perc

entM

odula

tion


-40 0 40 80-40

0

40

80Monkey A

r = 0.25Monkey B

r = 0.08

Correlation of Fano FactorModulation

N=118(all units)

2of4

Perc

entM

odula

tion


1 2-15

-10

-5

0

5

Items Tracked

Fano

Facto

rM

odula

tion

1 2-15

-10

-5

0

5

Items Tracked

Fano

Facto

rM

odula

tion

Documents

Attentive tracking of multiple objects modulates neuronal ...jude/VSS_2006_MOT_MONKEY.pdf · Pause Resume F r e q u e n c y (H z) (2 of 4 Attended – Distracter) – (1 of 4 Attended