Search Abstracts | Symposia | Slide Sessions | Poster Sessions
Lost in Transmission? - Morphological Processing with Cochlear Implants
Poster Session F, Friday, October 2, 2:45 - 4:45 pm, Wangari Maathai
Maxi Wollenberg1, Dr. Anja Hahne1; 1University Hospital Carl Gustav Carus at TUD Dresden University of Technology
Introduction: Cochlear implants (CIs) enable individuals with severe sensorineural hearing loss to (re)gain access to spoken language. Nevertheless, the auditory input provided by CIs differs markedly from natural hearing, posing challenges for language comprehension. Syntactically complex structures and non-salient speech cues are especially difficult for most individuals with CI (IwCI). The aim of the present study was to investigate the neurocognitive processes underlying sentence comprehension in postlingually deaf IwCI using event related potentials (ERPs). Methods: Twenty-nine IwCI (age 18–76 years) and 29 age-matched typically hearing participants (age 22–75 years) listened to sentences while EEG was recorded. Sentences were either grammatically correct (congruent condition: e.g., “The teacher tests the student”) or contained a morphological violation (incongruent condition: e.g., “*The teacher test the student”). Participants performed a sentence judgement task. Afterwards, ERPs were calculated time locked to the critical verb and analyzed in successive 200 ms windows from 400 to 1600 ms after stimulus onset in a parietal region of interest (ROI). Results: A mixed design ANOVA revealed significant main effects of condition, group, and a condition × group interaction (also moderated by time window). Follow up simple effects analyses showed a robust positive shift between 600 and 1200 ms in the incongruent relative to the congruent condition in the control group (600-800ms: F(1, 56) = 38.45, p < .001, η²ₚ = .41, 800-1000ms: F(1, 56) = 46.80, p < .001, η²ₚ = .46, 1000-1200ms: F(1, 56) = 13.28, p < .001, η²ₚ = .19), reflecting a robust P600 component. The IwCI group showed a weaker and shorter P600 effect (800-1000 ms: F(1, 56) = 4.56, p = .037, η²ₚ = .08). When only trials with correct behavioral responses were included, the group differences decreased markedly. Further exploratory analyses indicated qualitative differences within the IwCI group: high performing participants exhibited a clear P600 effect, whereas low performing participants showed a N400 like negativity. No comparable subgroup differences were observed in the control group. Conclusion: The P600 in the control group reflects the allocation of additional processing resources for integrating the morphologically incorrect verb suffix and re-analyzing these incongruent sentences. Overall, the IwCI group showed a weaker positive amplitude. According to our additional exploratory analyses, this appears to stem from two distinct, overlapping mechanisms within the IwCI group: (1) a P600 in some users and (2) a N400 type response in other participants, suggesting reliance on lexical semantic processing. These ERP patterns closely correspond to performance on the sentence judgement task, highlighting individual differences in neurocognitive processing among CI users and providing direct evidence for a link between neurocognitive EEG markers and behavioral language outcomes in adult IwCI.
Topic Areas: Speech Perception, Morphology