The Prefrontal Cortex in Top-Down Visual Modulation and Memory

1. Causal role of the prefrontal cortex in top-down modulation of visual processing and working memory Zanto, T. P., Rubens, M. T., Thangavel, A. & Gazzaley, A. (2011). Nature Neuroscience, 14, 656-661

2. Introduction • Attended items are more likely to be remembered than ignored items. • Selective attention seems to utilize an internal template or attentional trace that is based on working memory to resolve competition among multiple elements in the environment.

3. Question 1 A direct causal link via a common control region and underlying neural process has yet to be established.

4. Introduction • Top-down modulation is a bi-directional process, accomplished by differentially enhancing and suppressing the neural activity. • Top-down signals are thought to bias the likelihood of successfully representing and maintaining relevant information. • Activity in visual cortices during early visual processing stages.

5. Question 2 Direct evidence that selective attention and working memory are bound by this common neural mechanism has not yet be documented.

6. Introduction • Top-down modulation during visual processing engages the frontal-parietal-visual network. • The prefrontal cortex (PFC) is a source of this activity modulation.

7. Question 3 The mechanism underlying this long-distance communication and its influence on working memory performance has not been evaluated from a causal perspective.

8. Introduction • A region in the PFC, the inferior frontal junction (IFJ) may serve as a source of the top-down modulation underlying attention to visual features. • EEG data suggest that the IFJ may exert an influence on visual processing as early as 100 ms post-stimulus onset. (P1 component) • Also, top-down modulation may be subserved by long-distance alpha-band phase coherence.

9. Question 4 This hypothesis has not been directly addressed with a methodology that permits causal inferences.

10. Research goal • Determine the extent to which early top-down activity modulation in visual cortices was driven by the IFJ and how this affected working memory performance. • Identify neural mechanisms underlying the long-distance modulatory influences.

11. Method • Directly addressed both the causal role and mechanisms of PFC-mediated top-down modulation, as driven by selective attention, on subsequent working memory performance. • fMRI: neural network • EEG: posterior P1 amplitude and fronto-posterior alpha phase-coherence modulation • rTMS: causal role determined • Behavior: working memory accuracy

12. Method • 20 individuals (1 excluded) • Features: color (V4) and motion (V5) • Staircase procedure, and the procedure continued until a ‘just 100%’ level of performance reached. • 6 colors and 12 possible directions • Questionnaire for strategy

13. Experiment design • Condition: attend to motion or colors • 2 blocks each condition • 6 practice + 30 trials per condition (60 trials per EEG condition)

14. Method & design • 2 sessions: fMRI & EEG+rTMS • fMRI: • V4(color) & V5(motion) • V4 attend network , V4 ignore network, V5 attend network & V5 ignore network • EEG: • P1 • Negativity 200-400 ms post-stimulus onset • Lower v.s. higher left IFJ-V5 functional connectivity in attend network • rTMS: • 1-Hz for 1-min to the right IFJ (color: right IFJ only v.s. motion: bilateral) • rTMSv.s. sham rTMS, 2 blocks each • First half v.s. second half

15. Procedure

16. Result – neural networks subserving top-down modulation • Main effect: ROI and relevance • ROI x relevance interaction

17. Result – neural networks subserving top-down modulation • Fronto-parietal neural networks were engaged during top-down modulation for both features. • Right IFJ was a node in both color and motion neural network. • Target site for rTMS

18. Result – effects of rTMS on top-down modulation and working memory • Main effect: condition • rTMS x block half interaction • 3 way interaction

19. Result – effects of rTMS on top-down modulation and working memory • Main effect: rTMS in color • 3-way interaction in color • Comparable to the working memory accuracy findings

20. Result – effects of rTMS on top-down modulation and working memory • rTMS reduced the degree to which ignored color stimuli were suppressed. • An rTMS-related decrease in the amount relevant color stimuli was enhanced strongly trended toward significance • A single region in the PFC drives increases and decreases in cortical activity of a distant sensory region solely on the basis of task goals.

21. Result – effects of rTMS on top-down modulation and working memory • Subgroup x rTMS interaction • For actual rTMS, lower left IFJ-V5 functional connectivity was associated with diminished P1 modulation compared with higher left IFJ-V5 connectivity • Bilateral IFJ control compensates for the disruptive effects of rTMS to the right IFJ for motion processing

22. Result – effects of rTMS on top-down modulation and working memory • No negativity main effects or interaction was found. • The right IFJ rTMS effects on activity modulation appear to be limited to early visual processing.

23. Result – correlations between EEG, fMRI and behavioral data • Color: • Motion: no such relation ship (r=0.12, p>0.6)

24. Result – phase coherence subserving top-down modulation • Annticipatory alpha-band phase coherence may serve as a mechanism for top-down modulation to operate over long cortical distances.

25. Conclusion • Color processing showed an rTMS-related decline in both P1 modulation recovered with time, so did working memory performance. • As the P1 modulation recovered with time, so did working memory performance.

26. Conclusion • On an individual participant basis, the rTMS-induced effect on P1 modulation during color processing predicted changes in working memory accuracy. • Motion processing did not have rTMS-related effects on P1 modulation or working memory accuracy.

27. Conclusion • Early top-down activity modulation during stimulus processing driven by attentional demands is causally related to subsequent working memory performance.

28. Discussion • Previous research suggested that enhancement and suppression rely on distinct mechanisms. • Function segregation of the IFJ may exist at a scale smaller than the area affected by rTMS, consistent with the notion of topographical organization and functional specialization of this region.

29. Discussion • IFJ is involved in task switching, interference control and working memory. • It has been hypothesized that the IFJ is involved in updating relevant task representation.  support

30. Discussion • Alpha band activity appears to be a prime candidate for such process, as it is modulated by attention and linked to fronto-parietal networks subservingattentional and working memory. • The research show direct evidence.

31. Discussion • Although the observations were made during the encoding period, this does not preclude the idea that top-down modulation during other stages of the task influences working memory performance. • IFJ is only one region identified, future research will assess the necessity of other putative control regions in frontal and parietal cortex.

32. Thanks for your listening