The timing of activation in synesthesia: a magnetoencephalography study D Brang, EM Hubbard, S Coulson, M Huang & VS Ramachandran University of CA, San Diego [email protected]

Upright vs. italic detection task 2.2° tall white letters, numbers, and non-linguistic characters⁵ Approximately 100 targets per condition 500ms presentation with 1000ms ISI

V4 Retinotopy Task

Stimuli and Procedure

Red blocks subtending 3.1 degrees of the visual field 200 trials per visual quadrant 200ms presentation with ISI ranging 200-248ms V4 localized within posterior temporal lobe areas showing retinotopy

Ventral View

A

c

0.14

0.035

0.12

0.03

0.1

0.025 0.02 0.015 0.01 0.005 0 -0.005 10

Significant activations from the Grapheme Task localized to the posterior temporal lobe. Data is overlaid on subjects’ inflated cortical reconstructions, viewed ventrally.

40

70

100

130

160

190 220 Time (ms)

250

280

310

340

370

400

0.08 0.06 0.04 0.02 0 -0.02 10

40

70

V4 ROI

130

-0.04

160

190

220

250

280

310

340

370

400

Time (ms)

Activity in synesthetes’ Grapheme and V4 ROIs 0.14 0.12

Synesthetic subject

100

Synesthetes Controls Synesthetes Controls Achromatic letters and numbers evoked significantly more activity in synesthetes compared to controls between 141-160ms in V4 t(6) = 3.09, p<.05, but not in the Grapheme Area t(6) = 0.06, p=.95.

Grapheme ROI Color area V4

0.1 0.08 0.06 0.04 0.02

Grapheme ROI

0 10

In order to functionally localize V4 within each subject, square blocks were presented randomly in one of four visual quadrants, in order to elicit retinotopy in visual area V4; while identical stimuli were presented in each quadrant, blocks are displayed here in different colors for presentation.

V4 Retinotopy

Italic non-linguistic

MEG activation between 120-200ms

Methods

4 grapheme-color synesthetes (projectors) and 4 age- and handedness-matched controls. Data collected with a whole-head Elekta Neuromag 306-channel system in an enhanced multi-layer magnetically shielded room. Data analyzed using an L1-norm (VESTAL) approach³.

Italic letter

Average activity in synesthetes’ and controls’ grapheme ROIs

0.04

nanoAmperes x mm^2

Two main theories have been proposed to account for synesthetic experiences. The cross-activation model suggests that the experience of colored letters in synesthesia reflects hyperconnectivity between posterior fusiform areas involved in grapheme processing and adjacent color area V4¹. In contrast, the disinhibited feedback theory proposes a disinhibition of pre-existing, 'latent' feedback connections among areas in the hierarchy of visual processing²; the two models thus make different predictions about the time course of neural activity in color area V4 (early vs. late). In order to resolve this debate, we used magnetoencephalography (MEG) to test whether V4 and grapheme regions activate nearly simultaneously, as predicted by the cross-activation theory, or whether V4 activation occurs only after the initial stages of grapheme processing, as predicted by the disinhibited feedback theory.

Average activity in synesthetes’ and controls’ V4 ROIs

Using methods similar to those of Hubbard and colleagues⁴, synesthetes and controls were presented with graphemes and non-graphemic stimuli in a randomly intermixed presentation format. Grapheme ROIs created from a separate run. nanoAmperes x mm^2

In synesthesia, stimulation of one processing stream (e.g. hearing) elicits concurrent experiences in a second, unstimulated stream (e.g. visual colors). For example, to synesthete JC the number 2 always appears green (2), irrespective of its actual color. Although psychophysical experiments have demonstrated the reality of synesthetic experiences, the neural mechanisms underlying this phenomenon remain a matter of contention.

Timing of activity within V4 and Grapheme ROIs

Grapheme Task

nanoAmperes x mm^2

Introduction

40

70

100

130

160

190

220

250

280

310

340

370

400

Time (ms)

Left hemisphere

Right hemisphere

Whereas controls show only significant activation within grapheme regions in the posterior temporal lobe (light blue), synesthetes showed significant co-activation of both grapheme and V4 color ROIs (dark blue).

Control subject

Inflated Ventral View

V4 ROI

Grapheme ROI

Left hemisphere

Right hemisphere

Activity in synesthetes’ grapheme ROI reached significance between 135 and 139ms t(3) = 3.14, p=.05, with activity in the V4 ROI reaching significance between 140 and 144ms t(3) = 3.61, p<.05, suggesting V4 engagement occurs near simultaneously with processing of graphemes.

Discussion These data provide the strongest evidence to date that grapheme-color synesthesia involves direct communication between V4 and grapheme processing areas in the posterior temporal lobe (PTGA). The near simultaneous activation of color area V4 and PTGA between 135-145ms argues strongly against the cortical disinhibited feedback model of synesthesia which predicts activation of V4 only after substantial processing has occurred. The similar onset latencies we observed for increased activity in synesthetes' grapheme (135-139ms) and color (140-144ms) ROIs suggest the rapid exchange of information between these areas, in keeping with the report of increased connectivity between V4 and the posterior fusiform in synesthetes⁶. References 1. Ramachandran, V.S. & Hubbard, E.M. (2001). Proc Royal Soc Lond B 268, 979-983. 2. Grossenbacher, P.G. & Lovelace, C.T. (2001). Trends in Cognitive Sciences 5, 36-41. 3. Huang, M., Dale, A.M., Song, T., Halgren, E., Harrington, D.L., Podgorny, I., Canive, J.M., Lewis, S., Lee, R.R. (2006). NeuroImage. 31(3):1025-1037 4. Hubbard, E.M., Arman, A.C., Ramachandran, V.S. & Boynton, G.M. (2005). Neuron 45, 975-985. 5. Pesenti, M., Thioux, M., Seron, X., and De Volder, A. (2000). J. Cogn. Neurosci. 12, 461–479. 6. Rouw, R. & Scholte, H.S. (2007). Nature Neuroscience 10, 792-797.