Neural processing in a three-choice reaction-time task: a study using cerebral evoked-potentials and single-trial analysis in normal humans.

1. Previous studies have shown that the long latency event-related potentials (ERPs) reflect certain aspects of the sensory discrimination process, although the coupling of these ERPs to the actual discrimination is variable. Indeed, we have previously shown that during a two-choice reaction time task the discrimination is accomplished as a two-stage process, with the more frequently occurring stimulus discriminated at an earlier point than the rarer stimulus. The present paper examines the hypothesis that, in a three-choice reaction time task, the discrimination is similarly organized, i.e., is accomplished with the use of a three-stage process. 2. In the present experiments, we continuously recorded the electrocerebral activity (EEG) from the scalp and the electromyogram (EMG) from the responding muscles in a three-choice reaction time task in 10 strictly right-handed subjects. EEG and EMG responses were subsequently analyzed off-line by aligning them by the onset of either the stimulus (stimulus-synchronized) or the response (response-synchronized) for both correct and incorrect responses. 3. Subjects could be classified as "fast" or "slow" responders based on the mean response-latency to the most frequently occurring of the three tones (Frequent tone). Fast responders to the Frequent tone were also fast responders to the more frequent of the rare tones (the Rare-1 tone). By contrast, the response latency to the Frequent tone did not predict the speed of response to the most rare tone (the Rare-2 tone). 4. In the response-synchronized averages well-formed premovement potentials were present for the correct responses to all three tones. In the case of the Frequent tone, these potentials were symmetrical over the two cerebral hemispheres (as expected because both hands responded to this tone). They began > or = 200 ms before the average onset of the stimulus, suggesting that the preparation to respond preceded the stimulus. In the case of the two rare tones, the amplitudes of these premovement potentials were asymmetrical over the two hemispheres. For the Rare-1 tone, these potentials were lateralized to the hemisphere contralateral to the hand moved. For the Rare-2 tone, however, these premovement potentials were initially lateralized to the ipsilateral hemisphere, indicating that, even when subjects were able to respond rapidly and correctly to this tone, they were anticipating a need to respond with the incorrect hand (in anticipation of the more frequent Rare-1 tone).(ABSTRACT TRUNCATED AT 400 WORDS)