, Arthur Wunderlich
, Reinhard Tomczak, Henrik Walter, Matthias Wilhelm Riepe
University of Ulm, Department of Psychiatry, Ulm, Germany
University of Ulm, Department of Radiology, Ulm, Germany
University of Ulm, Department of Neurology, Ulm, Germany
Introduction
Tasks to measure visual memory usually require complex motor functions for the response process. Due to technical limitations during fMRI recognition rather than recall is usually used to measure behavioral performance. However, recognition engages different cortical regions than recall in visual memory (1-3). Therefore, we designed a novel task and a novel response procedure allowing free recall of abstract visual designs during functional MRI.
Method
Subjects (n=6; 2 females) were asked to memorize 10 abstract figures each presented for 5 seconds during a learning phase. During the control condition subjects attended to a screen image which was different in form and color to avoid interference with the figures to memorize. Free recall was measured subsequently after each learning phase. Active recall responses were given by means of a specially designed keyboard. During recall a simple visual reaction task served as control condition with the identical number of button presses. Functional MRI was performed with a 1.5 Tesla Siemens Magnetom Vision scanner with EPI-booster. Continuous multi-slice T2-weighted images were obtained using orientation parallel to the AC-PC-line (effective TR=3.3seconds/volume; TE=66ms). The matrix size was 128 x 128 pixels with 16 slices each. Slice thickness was 5 mm with a gap of 1 mm resulting in an anisotropic voxel size of 1.8 x 1.8 x 6 mm. For anatomical reference, T2w TSE-images were obtained at identical positions. Image processing and data analysis was performed by means of Statistical Parametric Mapping, SPM96 (Wellcome Department of Cognitive Neurology, London). Z-statistics for each and every voxel were thresholded at p <.001. The level of significance for inference of activated voxel clusters was p<.05 (corrected for multiple comparisons).
Results
Behavioral results showed a clear increase of correctly learned figures. Subjects recalled 23.3% of the figures after the first learning phase and 91.7% after the fifth learning phase. Group analysis of brain activation during learning revealed significant activations mainly of the lateral visual areas as well as posterior midline structures: Right occipital gyrus (BA 19), right medial occipital gyrus (BA 19), cuneus and precuneus, bilaterally, as well as right superior parietal lobule (BA 7). In about 30% of the subjects, additional analysis of each individual data set showed significantly activated frontal regions in the right hemisphere. There was no significant difference of brain activation between the first and fifth learning phase. During the active recall of the figures activation was again mainly bilateral within the posterior medial and lateral visual areas (BA 18,19: occipital gyrus, medial occipital gyrus and cunueus). In addition significant activation of the right superior parietal lobule (BA 7) was observed. As in the learning task, there were no significant differences between activated brain regions during the first and fifth reproduction phase despite the clear increase in correctly recalled figures. Qualitatively, a clearly greater extension of significantly activated brain tissue emerged.
Conclusions
Locations of significant activations are in accordance with other studies on visual memory (4,5). In contrast to our expectation group-analyses did not reveal medial temporal activation as a result of the free recall procedure. Nevertheless, in three of our six subjects individual analysis of data sets showed significant activation of the right fusiform gyrus. Interestingly, these subjects did also show the best behavioral results. We conclude that this newly designed test allows investigation of free recall visual memory under conditions of fMRI.
References
1 Mandler, G., Psych. Rev., 1980, 87:252-271
2 Janowsky, J.S., et al., Behav. Neurosci., 1989, 103:548-560
3 Schacter, D. L., et al., Proc. Natl. Acad. Sci. USA, 1996, 93:321-325
4 Roland, P.E. and Gulyas, B., Cereb. Cortex, 1995, 1:79-93
5 Stern, C., et al., Proc. Natl. Acad. Sci. USA, 1996, 93:8660-8665
