Search Abstracts | Symposia | Slide Sessions | Poster Sessions
Lesion-symptom mapping of phonological working memory load using rhyme judgments in people with chronic post-stroke aphasia
Poster Session D, Saturday, October 26, 10:30 am - 12:00 pm, Great Hall 3 and 4
William Matchin1, Dirk den Ouden1, Zeinab Khoshhal Mollasaraei1, Julius Fridriksson1; 1University of South Carolina
Working memory (WM) is a critical supporting mechanism for sentence processing. Phonological WM deficits are pervasive in aphasia, which makes it difficult to determine the extent to which a patient struggles with syntactic and semantic aspects of sentence processing, potentially limiting effective interventions. Brain research has converged on several regions, including inferior frontal lobe, inferior parietal lobe, and posterior superior temporal gyrus as critical regions for phonological WM, potentially distinct from the hypothesized brain bases of syntactic processing in the temporal lobe. This suggests that, in principle, syntactic deficits can be distinguished from phonological WM deficits in aphasia. However, it is difficult to distinguish regions implicated in difficulties with WM load as opposed to general aspects of task performance and sensory-motor abilities. Here we performed lesion mapping analyses people with chronic post-stroke aphasia of two different measures of phonological WM using rhyme judgments, using covariates to isolate the effects of WM load independent of the sensory, motor, and executive function aspects of the task by using covariates to control for these confounding variables. 103 patients with chronic post-stroke aphasia were recruited for a clinical trial conducted at the University of South Carolina and the Medical University of South Carolina. All patients were administered with the Western Aphasia Battery-Revised as well as several subtests of the Psycholinguistic Assessments of Language Processing in Aphasia (PALPA) and the Temple Assessment of Language and Short-term Memory in Aphasia (TALSA). Here analyzed performance on the TALSA triplets word rhyming task with heavy WM load using the PALPA auditory word rhyming task to control for auditory processing and executive task demands and the WAB-R auditory word recognition task to control for visual image recognition based on auditory stimulus. In addition, a subset of 78 patients completed the NAVS sentence comprehension test; we therefore related performance on our composite measure of phonological WM load to noncanonical sentence comprehension. We performed lesion mapping analyses using a mass univariate approach with regions of interest from the Johns Hopkins University atlas, analyzing proportion damage to each ROI for each participant and a permutation correction for multiple comparisons across regions (4,000 permutations). We also calculated uncorrected voxel-wise maps in order to spatially visualize the lesion maps associated with each of our primary analyses. The TALSA triplets task (with PALPA auditory word rhyme and WAB-R auditory word recognition as covariates) revealed significant lesion correlates in the supramarginal gyrus, inferior longitudinal fasciculus, and posterior insula. Voxel-based analyses revealed primary lesion correlates extending into the inferior parietal lobe. When including our composite measure of phonological WM as a covariate, lesion-symptom mapping analyses of noncanonical sentence comprehension were highly robust, implicating temporal lobe damage that did not extent into parietal lobe. The results suggest that the lesion bases of phonological WM load and those for syntactic comprehension deficits established in the literature at least partially diverge. In addition, measures of phonological WM load using rhyme judgment can be a useful tool for assessing WM abilities in people with post-stroke aphasia.
Topic Areas: Disorders: Acquired, Phonology