Neuroplasticity underlies improved speech in noise comprehension following a gamified training app
Refereed conference paper presented and published in conference proceedings
CUHK Authors
Full Text
There are no full text file(s) associated with this record. |
Other information
AbstractAbstract: The use of telemedicine applications among hearing healthcare providers has grown dramatically in the last few years. Many of these applications train hearing in noisy everyday situations, especially for children and the elderly. A promising development as comprehension of speech in noisy situations is a cornerstone to a high quality of life. Several training programs have successfully demonstrated significant improvements in speech in noise comprehension among normal hearing, the elderly, and cochlear implant users.
Motivation: No study has yet examined if functional neuroplasticity underlies improvements in speech in noise comprehension. This study thus aimed to identify the neural markers that underlie these improvements.
Methods: Twenty young normal hearing adults were assigned to either Treatment or Control group. The Treatment group received 15 minutes of daily speech in noise training through a smartphone app for 12 straight days; the Control group received none. Hemodynamic responses to stimuli were measured using functional near-infrared spectroscopy (fNIRS) while subjects performed a sentence in noise comprehension task at different noise levels (10, 6, 2 dB SNR). Both fNIRS and comprehension measures were twice taken: before and immediately following training.
Results: All subjects completed training and follow-up. A significant group effect was observed for improvement in both accuracy (p < .01) and reaction time (p < .05). The fNIRS data supported these behavioral results with a significant group effect of oxyhemoglobin measures in the bilateral superior and middle temporal gyrus (SMTG, left and right: p < .05) and left inferior frontal gyrus (IFG, left: p < .01). Further correlation analyses revealed that improvement in accuracy of the Treatment group appear correlated with pre-training hemodynamic response measures in both SMTG (left and right: r = -0.7, p < .05) and likely in IFG (left: r = -0.6, p = .073).
Discussion: These findings corroborate previous literature demonstrating SMTG and left IFG’s association with auditory processing and verbal working memory. Subsequent work on this study’s complete dataset will focus on predicting treatment efficacy from labeled pretreatment fNIRS data through conventional machine learning models using time, spectral, and wavelet domain features. It is expected this study will have implications for both the use of speech in noise training and the clinical management of hearing loss more broadly.
Motivation: No study has yet examined if functional neuroplasticity underlies improvements in speech in noise comprehension. This study thus aimed to identify the neural markers that underlie these improvements.
Methods: Twenty young normal hearing adults were assigned to either Treatment or Control group. The Treatment group received 15 minutes of daily speech in noise training through a smartphone app for 12 straight days; the Control group received none. Hemodynamic responses to stimuli were measured using functional near-infrared spectroscopy (fNIRS) while subjects performed a sentence in noise comprehension task at different noise levels (10, 6, 2 dB SNR). Both fNIRS and comprehension measures were twice taken: before and immediately following training.
Results: All subjects completed training and follow-up. A significant group effect was observed for improvement in both accuracy (p < .01) and reaction time (p < .05). The fNIRS data supported these behavioral results with a significant group effect of oxyhemoglobin measures in the bilateral superior and middle temporal gyrus (SMTG, left and right: p < .05) and left inferior frontal gyrus (IFG, left: p < .01). Further correlation analyses revealed that improvement in accuracy of the Treatment group appear correlated with pre-training hemodynamic response measures in both SMTG (left and right: r = -0.7, p < .05) and likely in IFG (left: r = -0.6, p = .073).
Discussion: These findings corroborate previous literature demonstrating SMTG and left IFG’s association with auditory processing and verbal working memory. Subsequent work on this study’s complete dataset will focus on predicting treatment efficacy from labeled pretreatment fNIRS data through conventional machine learning models using time, spectral, and wavelet domain features. It is expected this study will have implications for both the use of speech in noise training and the clinical management of hearing loss more broadly.
All Author(s) ListBurghardt RE, Ng IHY, Lee KYS, Wong PCM, Tong MCF
Name of ConferenceSociety of fNIRS Virtual Conference 2021
Start Date of Conference18/10/2021
End Date of Conference22/10/2021
Place of ConferenceLondon
Country/Region of ConferenceGreat Britain
Proceedings TitleSociety of fNIRS Virtual Conference 2021 Program Booklet
Year2021
Month10
Day18
Place of PublicationGreat Britain
LanguagesEnglish-United States