Presentation
Search Abstracts | Symposia | Slide Sessions | Poster Sessions | Lightning Talks
Leftward shift in dichotic ear advantage associated with temporal lobe lesions but not contralesional supranormal language activity in chronic post-stroke aphasia
Poster E52 in Poster Session E, Thursday, October 26, 10:15 am - 12:00 pm CEST, Espace Vieux-Port
Andrew DeMarco1, Kelly Martin1, Nikta Najib1, Peter Turkeltaub1,2; 1Georgetown University, 2MedStar National Rehab Hospital
Dichotic listening involves simultaneously presented, conflicting auditory tokens to each ear. When speech stimuli are presented, right-handers exhibit a classic right-ear advantage (REA), more frequently reporting the right-ear percept due to greater connectedness to the left, language-dominant hemisphere. Prior studies have found that lesions can cause a reversed, left-ear advantage (LEA). While some have interpreted a LEA as evidence for a hemispheric reversal in language dominance, the more common explanation involves the left-ear percept reaching left-hemisphere processors via an extra right-hemisphere auditory relay. But this explanation is difficult to reconcile with damage to those processors, and does not account for evidence suggesting that frontal lobe and attention might be involved in the interhemispheric transfer. Localization of lesions causing a LEA has been inconsistent, although most work implicates damage to Heschl’s gyrus and the geniculo-temporal pathway. Here we aimed to localize lesions associated with a LEA in chronic left-hemisphere stroke-survivors with aphasia, and test whether patients with LEA exhibit supranormal contralesional language activation. Participants were selected from patients (N=72, M/F=36/36, Age=61.2(11.5)) and controls (N=90, M/F=51/49, Age=60.7(12.1)) were selected from an ongoing cross-sectional study of stroke aphasia outcomes. All participants underwent neuroimaging including structural scans, from which lesions were traced, and an fMRI language-mapping task. The task involved decisions on written stimulus-pairs, with alternating blocks of semantic relatedness judgment on words (SEM) and identity judgment on false-font strings (VIS). All participants completed a fused CV dichotic test. Lesions associated with ear advantage were localized using multivariate lesion-symptom mapping (LSM), controlling for lesion size (voxelwise P<.005, clusterwise, P<.05, 10,000 permutations). Right-hemisphere atlas parcels were examined if they overlapped with the left-right flipped LSM result. Then a two-sample t-test measured if activation (SEM>VIS) within qualifying parcels was greater in LEA or REA patients than controls (P<.05, Bonferroni corrected). Of 64 dextral controls, 41 showed a REA and 3 showed a LEA. Of 56 reliable patients, 16 showed a REA and 22 showed a LEA. Ear-advantage was associated with one damage to the superior temporal gyrus (22.6cc; y ~ -38 to +19), underlying white matter, and posterior insula. Two of 24 right-hemisphere parcels showed supranormal activation in LEA patients, including anterior insula (t(59)=3.80, P_bf=.004) and Rolandic operculum (t(59)=3.70, P_bf=.006). The same supranormal parcels were found in the REA patient group. LEA was associated with neither aphasia severity (WAB, t(39)=-0.99, P=.33) nor lesion size (t(39)=1.49, P=.15). Patients were almost eight times more likely than controls to exhibit a left-ear advantage. Lesions associated with ear-advantage involved expected temporal lobe regions but extended beyond classic speech processing areas. Supranormal activity was observed in contralesional frontal lobe and insula in both LEA and REA patients. These findings suggest that lesions to the ascending auditory pathway and temporal lobe can cause a LEA, but do not support a right-hemisphere account for the LEA.
Topic Areas: Disorders: Acquired, Speech Perception