Short-term audio-tactile training affects cortical speech-envelope tracking but not speech intelligibility

Humans are adept at integrating auditory and visual information to improve intelligibility, which can improve neural tracking of the speech envelope. Speech-relevant tactile information also improves tracking yet effects on intelligibility are mixed, possibly because we are not regularly exposed to speech-relevant tactile information. Here, we used a short-term audio-tactile training protocol to fundamentally advance understanding of audio-tactile integration capabilities during speech perception. Data from 64 younger adult participants (aged 18-29) were collected over five EEG sessions. Participants completed a speech-in-noise task, with audio-tactile and audio-only stimuli. Participants then received training with either tactile information that was congruent with sentences heard (trained group) or incongruent (pseudo-trained group). After three training sessions, participants completed the speech-in-noise task again in a fourth session. Measures of neural speech-envelope tracking were recorded. Two weeks later, participants returned for a follow-up session. Results indicated a significant effect of session (pre- or post-training) on speech intelligibility, but no significant effect of group (trained, or pseudo-trained) or stimulus (audio-only, or audio-tactile). At baseline, there was a significant increase in speech-envelope tracking accuracy with audio-tactile stimuli relative to audio-only, suggesting more accurate neural representation during perception of audio-tactile stimuli. After training, there was no benefit of congruent training to audio-tactile tracking, but there was an enhancement of audio-only speech-envelope tracking following incongruent training. This suggests that speech intelligibility and tracking of audio-tactile speech are not enhanced by short-term audio-tactile training. This work challenges prevailing assumptions by demonstrating a lack of support for the assumed link between tracking and intelligibility. These data provide critical evidence that will inform future innovation of speech understanding in technology.