Neural Mechanism of Implicit and Explicit Memory Retrieval: Functional MR Imaging

Abstract

Purpose: To identify, using functional MR imaging, distinct cerebral centers and to evaluate the neural mechanism associated with implicit and explicit retrieval of words during conceptual processing. Materials and Methods: Seven healthy volunteers aged 21-25 (mean, 22) years underwent BOLD-based fMR imaging using a 1.5T Signa Horizon Echospeed MR system. To activate the cerebral cortices, a series of tasks was performed as follows: the encoding of two-syllable words, and implicit and explicit retrieval of previously learned words during conceptual processing. The activation paradigm consisted of a cycle of alternating periods of 30 seconds of stimulation and 30 seconds of rest. Stimulation was accomplished by encoding eight twosyllable words and the retrieval of previously presented words, while the control condition was a white screen with a small fixed cross. During the tasks we acquired ten slices (6 mm slice thickness, 1 mm gap) parallel to the AC-PC line, and the resulting functional activation maps were reconstructed using a statistical parametric mapping program (SPM 99). Results: A comparison of activation ratios (percentages), based on the number of volunteers, showed that activation of Rhs-35, PoCiG-23 and ICiG-2630 was associated with explicit retrieval only; other brain areas were activated during the performance of both implicit and explicit retrieval tasks. Activation ratios were higher for explicit tasks than for implicit; in the cingulate gyrus and temporal lobe they were 30% and 10% greater, respectively. During explicit retrieval, a distinct brain activation index (percentage) was seen in the temporal, parietal, and occipital lobe and cingulate gyrus, and PrCeG-4, Pr/PoCeG-43 in the frontal lobe. During implicit retrieval, on the other hand, activity was greater in the frontal lobe, including the areas of SCA-25, SFG/MFG-10, IFG-4445, OrbG-1147, SFG-68, and MFG-946. Overall, activation was lateralized mainly in the left hemisphere during both implicit and explicit retrieval tasks. For explicit retrieval, the lateralization index was more than twice as high as for implicit retrieval. Conclusion: Our findings indicate that there is neuro-anatomical dissociation between implicit and explicit retrieval of words during conceptual processing, suggesting, on the basis of cognitive neuroscience, that the performance of implicit and explicit memory-related tasks involves different mechanisms.