Steven Phillips and Kazuhisa Niki
Neuroscience Research Institute, National Institute of
Advanced Industrial Science and Technology (AIST)
Previous research (Phillips & Niki, 2002) identified the parietal lobes with increased number of item relationships, but not number of items in pair recognition memory tasks. The increased parietal activity was due to either increased item associates; or increased index length - fewest number of positions at which every pair has a unique combination of items. For example, AB BC CD is a unary (length one) indexed list of three pairs, because all items are unique at the first (or second) position; AB AD CB is a binary indexed list, because only pairs of items are unique. In this experiment, the numbers of items and associates were varied independently of index length. Subjects were asked to make a recognition judgment for each three-pair list: Was the test pair in the previous list? A contrast of binary minus unary indexed lists revealed occipital and parietal activity (bilaterally) during the retention-only period. This result is not explained by item load or item fan effects, because the numbers of items and associates were the same for both types of lists. Binary indexed lists require attention to items in both positions to maintain unique identifiability of each pair memory trace. Hence, this result is interpreted in terms of increased shifts of attention, a function often attributed to the parietal lobes.
Twelve Japanese university students participated in the experiment, after providing informed consent according to AIST safety and ethics guidelines. Subjects were given 50 trials with the following sequence structure: First-pair (1170ms), Clear-screen (2000ms), Second-pair, Clear-screen, Third-pair [Encode phase], Clear-screen (12000ms) [Retention phase], Test-pair (2000ms), End-of-trial-marker (5000ms) [Test phase]. Five types of three-pair lists were used: AB BC CA (unary, three); AB BC CD (unary, four); AB AC CB (binary, three); AB AD CB (binary, four); and AB CD EF (unary, six).
Scanning was performed on a 3.0-T MRI Scanner (GE 3T Signa) with EPI capability. 18 axial slices (5.5 mm thick, interleaved) were set to cover the entire brain. A T2* weighted gradient echo EPI was employed. The imaging parameters were TR=2 sec, TE=32 ms, FA=70°, FOV=20x20 (64x64 mesh). The data were preprocessed and fitted by a general linear model using SPM99. Five encoding and retention-only events (one for each list type) were defined. Encode events were modeled as a canonical hemodynamic response function convolved with a step function covering the presentation of target pairs. Retention events were modeled as a step function coinciding with the duration of the retention-only period. There were 5 [List] x 2 [Test] + 1 [Error] recognition events modeled as the canonical hemodynamic response function with the onset coinciding with the onset of the test pair. A high-pass filter with a cutoff of 120s was used. Random effects analyses were performed to make inferences about the contrasts of interest. The voxel threshold was set at P<.001, uncorrected.
Reference:
Phillips and Niki (2002). NeuroImage, 17(2), 1031-1055.