Presentation
Search Abstracts | Symposia | Slide Sessions | Poster Sessions | Lightning Talks
Variability in Online Speech Comprehension in Aphasia as Evidenced by Event-Related Potentials
There is a Poster PDF for this presentation, but you must be a current member or registered to attend SNL 2023 to view it. Please go to your Account Home page to register.
Poster E53 in Poster Session E, Thursday, October 26, 10:15 am - 12:00 pm CEST, Espace Vieux-Port
Anna Krason1, Erica L. Middleton1, Matthew E. P. Ambrogi1, Malathi Thothathiri2; 1Moss Rehabilitation Research Institute, 2The George Washington University
Comprehension deficits in aphasia are common but less widely studied and understood than production deficits. Behavioral responses alone may provide an incomplete picture, with different individuals potentially showing decrements in accuracy for different reasons. We show here–in both healthy controls and persons with aphasia (PWA)–that similar behavioral performance is accompanied by distinct event-related potential (ERP) patterns, suggesting underlying variability in speech processing. Two adults with agrammatic (A1,A2) or mild aphasia (M1,M2), and two controls (C1,C2) participated in two (M1,M2,C1,C2) or three (A1,A2) EEG sessions. Participants completed two auditory acceptability judgment tasks with (i) 50 syntactic (e.g., "The press often compares she to Marilyn Monroe") and (ii) 50 semantic (e.g., "The cowboy saddled his necklace") anomalous or acceptable sentences per session. We measured accuracy and mean amplitudes at posterior electrodes (A8, O1, Pz, Oz, Cz) within 600-800ms for (i) and (A8, Pz, POz, Cz) within 400-600ms for (ii) following critical anomalous/acceptable words. We used LMER with mean amplitude as the dependent variable and sentence (acceptable vs. anomalous) as the independent variable. Prior literature would predict a positivity (P600) and a negativity (N400) for syntactic and semantic anomalies, respectively in group analyses. However, we evaluated individual-level results to uncover potential variability in underlying mechanisms. Behaviorally, mild aphasics and controls performed at ceiling in both tasks (>88%). A1 was 57% and 72% accurate and A2 was 80% and 82% accurate in (i) and (ii), respectively. Trial rejection rate due to artifacts was within conventional limits across participants: mean=0.08 (SD=0.10) for (i) and mean=0.14 (SD=0.13) for (ii). As expected, group analysis showed a significant P600 for (i) (β=1.31,SE=0.55,t=2.37,p=.02) and N400 for (ii) (β=-1.07,SE=0.45,t=-2.38,p=.02). Looking at individuals, for (i), there was a significant P600 for C1 (β=4.05,SE=1.08,t=3.74,p=.001) and M1 (β=3.47,SE=0.97,t=3.59,p=.002) only. For (ii), there was a significant N400 for A1 (β=-2.45,SE=1.05,t=-2.32,p=.02); a marginal N400 for M1 (β=-1.56,SE=0.81,t=-1.91,p=.06); a significant N400 for C2 but in a later time window of 600-800ms (β=-3.11,SE=1.47,t=-2.12,p=.04); and a significant positivity for A2 (β=1.29,SE=0.62,t=2.08,p=.04) that extended to the 600-800ms interval (β=3.09,SE=0.72,t=4.29,p<.001). In controls and participants with mild aphasia, ERP differences despite similarly high accuracy suggest variability in sentence processing (lack of P600 for C2 and N400 for C1, lack of P600 and N400 effects for M2 but not M1). In individuals with agrammatism, neither participant showed a P600, suggesting difficulty in online syntactic processing even though A2 performed better than A1. For semantic anomalies, A1 showed an N400 and A2 did not and yet the former did worse behaviorally. These discrepancies between behavioral and ERP results in each individual suggest distinct underlying mechanisms of speech comprehension, particularly when one needs to make a judgment decision requiring intact cognitive control. If instead cognitive control is impaired, it can contribute to comprehension deficits in aphasia. In sum, despite finding predicted group-level P600 and N400 effects, individual analyses unveiled significant variations among adults who look similar on the surface. Uncovering and understanding these differences can provide valuable theoretical and clinical insights into mechanistic accounts of speech comprehension, predicting language outcomes, and personalized treatments.
Topic Areas: Disorders: Acquired, Speech Perception