Ya-Ping Chen 陳雅苹

(A) Average time-generalized Spearman’s R matrices relating CI ears and non-CI ears to NHCs were computed per session. (B) Correlation coefficients within the 600-1200 ms reveal that the correlation between non-CI ears and NHCs improves at 6 and 12 months post-operation. (6/7)

To track the neural recovery in each ear of CI users, we then used RDMs to correlate word degradation representations of the CI user's CI ear and non-CI ear with those of NHCs. (5/7)

The decoding results from postoperative non-CI ears along with normal hearing controls (NHCs) showed greater-than-chance decoding accuracy. On the other hand, CI ears did not show greater-than-chance decoding accuracy in any session. (4/7)

To see how the brain discriminates among various degraded speech via CI and non-CI ears, we applied SVM to classify 4 degradation levels of words. Outcomes at each time point were aggregated into a representational dissimilarity matrix (RDM) and a decoding accuracy curve. (3/7)

The subjective comprehension rating increased in both CI and non-CI ears at 6 months post-operation. Furthermore, the improvement in the non-CI ears was greater than that observed in the CI ears. (2/7)

#SfN2022 Drop by my poster this afternoon if you are interested in how speech intelligibility modulates neural speech tracking within 400 ms! Keywords: vocoded speech, temporal response functions (TRF), coherence, FOOOF, MEG @SalzburgL @PLUS_1622

Relating TRF effects to FOOOF components of coherence spectra suggests that M50TRF and M200TRF reflect shifts in which speech features (acoustic or syllabic information) are tracked over time.

To complement our results in TRF, we also computed speech-brain coherence and parameterized coherence spectra using FOOOF.

Three TRF components were identified: M50, M200, and M350. Only M200 decreased with speech intelligibility.