Whiting K A, Martin B A, Stapells D R
Auditory Evoked Potential Research Laboratory, Albert Einstein College of Medicine, Bronx, New York, USA.
Ear Hear. 1998 Jun;19(3):218-31. doi: 10.1097/00003446-199806000-00005.
To systematically investigate in normal-hearing listeners the effects of decreased audibility produced by broadband noise masking on the cortical event-related potentials (ERPs) N1, N2, and P3 to the speech sounds /ba/ and /da/.
Ten normal-hearing adult listeners actively (button-press response) discriminated the speech sounds /ba/ and /da/ presented in quiet (no masking) or with broadband masking noise (BBN), using an ERP oddball paradigm. The BBN was presented at 50, 60, and 70 dB SPL when speech sounds were presented at 65 dB ppe SPL and at 60, 70 and, 80 dB SPL when speech sounds were presented at 80 dB ppe SPL.
On average, the 50, 60, 70, and 80 dB SPL BBN maskers produced behavioral threshold elevations of 18, 25, 35, and 48 dB (average for 250 to 4000 Hz), respectively. The BBN maskers produced significant decreases (relative to quiet condition) in ERP amplitudes and behavioral discriminability. These decreases did not occur, however, until the noise masker intensity (in dB SPL) was equal to or greater than the speech stimulus intensity (in dB ppe SPL), that is, until speech to noise ratios (SNRs) were < or = 0 dB. N1 remained present even after N2, P3, and behavioral discriminability were absent. In contrast to amplitudes, ERP and behavioral latencies showed significant decreases at higher (better) SNRs. Significant latency increases occurred when the noise maskers were within 10 to 20 dB of the stimuli (i.e., SNR < or = 20 dB). The effects of masking were greater for responses to /da/ compared with /ba/. Latency increases occurred with less masking for N1 than for P3 or behavioral reaction time, with N2 falling in between.
These results indicate that decreased audibility as a result of masking affects the various ERP peaks in a differential manner and that latencies are more sensitive indicators of these masking effects than are amplitudes.
系统研究在听力正常的受试者中，宽带噪声掩蔽导致的可听度降低对皮质事件相关电位（ERP）中N1、N2和P3成分以及对语音/ba/和/da/的影响。
10名听力正常的成年受试者采用ERP奇偶数范式，在安静环境（无掩蔽）或有宽带掩蔽噪声（BBN）的情况下，主动（按键反应）辨别呈现的语音/ba/和/da/。当语音以65 dB ppe SPL呈现时，BBN以50、60和70 dB SPL呈现；当语音以80 dB ppe SPL呈现时，BBN以60、70和80 dB SPL呈现。
平均而言，50、60、70和80 dB SPL的BBN掩蔽器分别使行为阈值升高18、25、35和48 dB（250至4000 Hz的平均值）。BBN掩蔽器使ERP波幅和行为辨别力显著降低（相对于安静条件）。然而，这些降低直到噪声掩蔽器强度（以dB SPL为单位）等于或大于语音刺激强度（以dB ppe SPL为单位）时才出现，即直到语音噪声比（SNR）≤0 dB时才出现。即使在N2、P3和行为辨别力消失后，N1仍然存在。与波幅相反，ERP和行为潜伏期在较高（较好）SNR时显著缩短。当噪声掩蔽器与刺激的强度相差10至20 dB时（即SNR≤20 dB），潜伏期显著延长。与/ba/相比，掩蔽对/da/反应的影响更大。N1潜伏期延长所需的掩蔽比P3或行为反应时间少，N2介于两者之间。
这些结果表明，掩蔽导致的可听度降低以不同方式影响各种ERP波峰，并且潜伏期比波幅更能灵敏地反映这些掩蔽效应。
