Abstracts

Rationale: Impairments in visuoperceptual ability exist in patients with temporal lobe epilepsy (TLE), but these deficits are subtle and cannot always be demonstrated with standard tests. This may be because the available tests are not challenging enough to reveal mild impairments or because they may depend on other cognitive skills, such as naming or memory. We developed a new test in four versions to measure visual perception. We expect this test to be sensitive to perceptual deficits in patients with TLE. As recent functional imaging studies have shown preferential involvement of the right hemisphere in visuoperceptual processing, we also expect greater impairments in patients with right TLE than in those with left. Methods: Two types of stimuli were used: complex abstract designs and real scenes. We paired them such that each pair was identical except for five differences that consisted of the appearance, disappearance, modification, rotation, or displacement of elements. We constructed 36 pairs of each type and, for piloting, divided them into three equal sets containing 12 abstract and 12 scene pairs. As these puzzles are meant to vary in level of difficulty and to engender little or no between-test practice effects, we tested healthy subjects over three sessions so that we could organize the items into four tests of similar difficulty. Patients with unilateral TLE were given just one test containing 12 puzzles (6 abstract, 6 scenes). In all cases the task was to circle the 5 differences between each pair as quickly as possible. Response measures were time to completion and number of differences found. Results: Healthy subjects found differences in the abstract puzzles significantly faster than in the real scenes. In both types of puzzle we found a clear difference in level of difficulty based on time to completion, allowing them to be divided into four groups of 12 items, each with equal numbers of easy, medium, and hard puzzles. There were no differences related to testing session. The difficulty of the puzzle was affected by the type of change: appearances and disappearances were the easiest to find and rotations and displacements were the most difficult. Results with TLE patients showed that they were slower than the healthy subjects and made more errors. These preliminary results also suggest greater impairment in patients with right than in those with left TLE. Conclusions: Our new test promises to provide a good measure of subtle visuoperceptual deficits, with a graduated level of difficulty so that it can be used with patients with low or high intellectual functioning. Further, our four versions can be used in retest without fear of practice effects. The task does not require naming or memory, and thus should isolate visuoperceptual skills from other cognitive functions that could confound the results. Our preliminary results with patients suggest that the task is sensitive to subtle visuoperceptual deficits in TLE, and, in particular, right TLE. Funded by Canadian Institutes of Health Research