Association for the Scientific Study of Consciousness
CalTech, Pasadena CA
June 24-27, 2005
Arnaud Delorme-1, Marissa Westerfield-2, Scott Makeig-1. , 1. Swartz Center, Institute of Neural Computation, UCSD, La Jolla CA.
2. Department of Neurosciences, UCSD, La Jolla CA.
Neural dynamics of thought-full versus thought-free responses in orbitofrontal cortex.
According to Wundt (1832-1920) impulsive responses are triggered without the subject thinking about it. Indeed, it has been shown that complex tasks involving written language or category discrimination can be performed unconsciously (Dehaene et al, Nature, 1998, 395, 597-600; Fabre-Thorpe et al., J Cog Neuroscience, 2001, 13, 171-80). On the other hand, before slower non-impulsive responses, subjects might likely base their behavioral decisions on their more fully conscious experience. We studied the neural correlate of faster (thought-less) responses versus slower thoughtful responses in a visual selective attention paradigm. Following visual target stimuli presented infrequently at a covertly attended location, faster thumb button responses tended to follow a larger far-frontal evoked positivity (P3f or P2a) in the average event-related potential (ERP) (Makeig et al., J Neurosci, 1999, 19, 2665-80).
Closer study of these data revealed a linear relationship (r=0.96, p<10-11) between median RT and the upslope to the P3f peak. This result is compatible with a model in which the behavioral response is triggered when the area under the ERP, beginning at a relatively fixed latency near 200 ms, reaches a threshold. Consistent with this hypothesis, the area under the P3f was not correlated with response time.
We then attempted to determine how this effect was related to the EEG activity in single trials. Interestingly, spectral decomposition of the EEG signal revealed that a 1.5-cycle, 4.5-Hz theta burst preceded and was weakly phase-locked to the button press. We found that the amplitude of this theta wave, as well as its phase, covaried significantly with response time (p<10^-33), a higher theta-power wavelet tending to immediately precede fast responses. We also observed that this effect persisted even after regressing the average response-locked P3f peak out of every trial. On the other hand, regressing the theta wavelet out of every trial removed the P3f and its relation to RT. Thus it seems that wavelet analysis in the theta band better characterizes the event-related brain dynamics in this case than ERP analysis.
Finally, independent component analysis identified inferior frontal processes with similar theta dynamics in most subjects. Comparable components from two 256-channel recording sessions localized to orbitofrontal cortex. Overall, slower responders produced less P3f/theta activation, possibly reflecting these subjects' inability to respond reflexively, in line with Wundt's hypothesis. The orbitofrontal theta response complex may index inferior frontal processing of early limbic signals to facilitate or inhibit reflexive responses in speeded response tasks.