助聽器降噪演算法透過提升相位鎖定的大腦反應對聲學邊緣之時頻一致性，支持神經語音追蹤

// Hearing Aid Noise Reduction Algorithms Support Neural Speech Tracking by Enhancing Spectrotemporal Consistency of Phase-Locked Brain Responses to Acoustic Edges

26位聽損者（戴雙側助聽器）在安靜與噪音下聽連續語音，開/關噪音抑制(NR)並記錄EEG｜噪音降低P1振幅，NR部分恢復且與試次間頻時一致性提升相關｜顯示NR透過恢復對聲學邊緣（聲音突變）的神經反應，增強語音追蹤，支持以神經原理優化NR。

// The study tested whether hearing-aid noise reduction (NR) restores neural responses to acoustic edges and thereby improves EEG speech tracking; 26 hearing-impaired listeners with bilateral hearing aids heard continuous speech in quiet and in noise with and without NR while EEG was recorded, and analyses focused on event-related potentials to speech onset edges, trial-by-trial spectrotemporal consistency, and stimulus reconstruction of the speech onset envelope. P1 amplitudes in auditory late responses were reduced in noise versus quiet but partially recovered with NR, and this recovery correlated with increased spectrotemporal consistency across trials. Stimulus reconstruction produced qualitatively identical, highly correlated results. These findings indicate that NR partly enhances neural speech tracking by restoring neural responses to acoustic edges, supporting NR designs that leverage neural principles.

Published 2026年9月3日