Brain activity underlying stereotyped and non-stereotyped retrieval of learned stimulus-response associations

When humans retrieve learned stimulus-response associations in a stereotyped manner the necessary brain structures may differ from those required when the same associations must be retrieved and adapted to new circumstances. We tested this hypothesis by means of tasks that resembled those employed in monkeys, using positron emission tomography (PET). Stimuli consisted of abstract two-dimensional shapes. Stereotyped retrieval of learned stimulus-response associations was studied by the use of a concurrent discrimination task with fixed pairing. This was contrasted with conditions requiring retrieval and adaptation of learned associations: forced-choice recognition, response reversal and concurrent discrimination with random pairing. Visuomotor control, passive viewing and fixation conditions were also included. During concurrent discrimination with fixed pairing, the left lower precentral gyrus and rostral anterior cingulate demonstrated higher blood flow levels in comparison with recognition, concurrent discrimination with random pairing, and to a lesser degree, response reversal. In the left lower precentral gyrus these blood flow levels were also higher in comparison with control conditions. Conversely, during recognition, concurrent discrimination with random pairing and reversal, a single region within the right inferior frontal gyrus demonstrated higher blood flow levels in comparison with concurrent discrimination with fixed pairing and control conditions. This right inferior frontal gyrus activation did not depend on the need for active familiarity judgements or response inhibition. To conclude, the left lower precentral gyrus is more active during stereotyped retrieval of learned stimulus-response associations and the right inferior prefrontal cortex is more active when a learned stimulus-response association must be retrieved and adapted to new circumstances.