P59Session 1 (Thursday 11 January 2024, 15:35-18:00)Word comprehension with and without noise: Longitudinal evidence in cochlear implant users using event-related brain potentials
Adult patients who receive a cochlear implant (CI) do not immediately understand spoken language. Rather, it is a complex learning process to gain access to the speech system via the implant. This is particularly true for understanding in a noisy environment, which often does not reach a satisfactory level, even in the long term. Our understanding of this learning process and its neurophysiological basis is rather limited.
In this study, objective brain potential measurements were used to investigate the development of word processing over time. The study was designed as a longitudinal study with measurement points 3 days, 6 weeks, 3 months and 12 months after initial implant activation. 23 post-lingual CI users [mean (SD): 64 (12); median 66, range: 36-81] with severe to profound contralateral hearing loss participated. In addition, a matched typical hearing control group [mean (SD): 64 (12); median 66, range: 33-80] was tested. A picture-word matching paradigm was used, i.e. a picture was shown on a screen accompanied by a spoken word. This word either matched the picture (correct condition) or did not match the picture (incorrect condition). Half of the acoustic words were accompanied by a stationary noise (ICRA1; SNR 5dB). The EEG was recorded and evoked potentials were calculated offline.
For the no noise condition, we observed a significant N400 effect in the ERP (= difference between incorrect and correct condition) already 3 days after first fitting. However, the onset latency was largely delayed. Onset latency decreased systematically over the measurement points. After one year of implant use, the ERP effect of the CI group was similar to that of the control group. In the noise condition, the first weak and late N400 effect was seen after 3 months of implant use, and even after one year the N400 effect was still later and reduced compared to the no noise condition.
Using objective methods, we were able to follow the process of learning to understand words in CI users during the first year of implant use. While comprehension of words without noise was observed within a few days after initial activation, comprehension of words in noise is much more challenging and not comparable to the control group even after one year of implant use.
Acknowledgements: We thank MEDEL for funding this study.