Functional anatomical correlates of controlled and automatic processing

Behavioral Studies have shown that consistent practice of a cognitive task can increase the speed of performance and reduce variability of responses a nd error rate, reflecting a shift from controlled to automatic processing. This study examines how the shift from controlled to automatic processing c hanges brain activity. A verbal Sternberg task was used with continuously c hanging tat-gets (novel task, NT) and with constant, practiced targets (pra cticed task, PT). NT and PT were presented in a blocked design and contrast ed to a choice reaction time (RT) control task (CT) to isolate working memo ry (WM)-related activity. The three-dimensional (3-D) PRESTO functional mag netic resonance imaging (fMRI) sequence was used to measure hemodynamic res ponses. Behavioral data revealed that task processing became automated afte r practice. as responses were faster. less variable, and more accurate. Thi s was accompanied specifically by a decrease in activation in regions relat ed to WM (bilateral but predominantly left dorsolateral prefrontal cortex ( DLPFC), right Superior frontal cortex (SFC). and right frontopolar area) an d the supplementary motor area. Results showed no evidence for a shift of f oci of activity within or across regions of the brain, The findings have th eoretical implications for understanding the functional anatomic-it substra tes of automatic and controlled processing, indicating that these types of information processing have the same functional anatomical substrate, but d iffer in efficiency. In addition, there are practical implications for inte rpreting activity as a measure for task performance, such as in patient stu dies. Whereas reduced activity can reflect poor performance if a task is no t sensitive to practice effects, it can reflect good performance if a task is sensitive to practice effects.