Sabri M, Campbell K B
School of Psychology, University of Ottawa, Ottawa, Canada K1N 6N5.
Brain Res Cogn Brain Res. 2001 Aug;12(1):171-80. doi: 10.1016/s0926-6410(01)00026-x.
The mismatch negativity (MMN) increases in amplitude as the probability of deviant occurrence decreases. It is unclear whether the determining variable is sequential probability (i.e. the probability of a deviant within a number of standards) or temporal probability (i.e. the probability of a deviant within a period of time). Eight subjects heard a train of frequently occurring 1000 Hz standard tones. The probability of a 1100 Hz pitch deviant was manipulated. In one condition the stimulus-onset-asynchrony (SOA) was 150 ms, with temporal probability of deviant occurrence being either 1/9.00, 1/4.50, 1/2.25, or 1/1.125 s (sequential probability being 1/60, 1/30, 1/15, or 1(deviant)/7.5(standards), respectively). In another condition the SOA was 600 ms, with temporal probability being either 1/9.00, 1/4.50, or 1/2.25 s (sequential probability being 1/15, 1/7.5, or 1/3.75, respectively). In a final condition, the SOA was 2400 ms with temporal probability being 1/9.00 s (sequential probability 1/3.75). Both sequential and temporal probabilities had a marked effect on the MMN. When a deviant occurred every 2.25, 4.50, or 9.00 s, the MMN increased as temporal probability decreased. When a deviant occurred once every 7.5 or 15 standards, the MMN was larger for lower sequential probability, but the effect was not significant. Nevertheless, when temporal probability was held constant at 1/9.00 s, the MMN increased as sequential probability decreased. At rapid rates of stimulus presentation, the MMN was largest. However, it was attenuated when the probability of deviant occurrence was very high perhaps due to the refractoriness of its generator. At the slowest rate, the MMN was diminished perhaps due to memory decay for the standard stimuli.
失配负波(MMN)的波幅会随着偏差刺激出现概率的降低而增大。目前尚不清楚起决定作用的变量是序列概率(即若干标准刺激中出现一个偏差刺激的概率)还是时间概率(即一段时间内出现一个偏差刺激的概率)。八名受试者聆听了一连串频繁出现的1000赫兹标准音调。对1100赫兹音调偏差刺激的概率进行了操控。在一种条件下,刺激起始异步时间(SOA)为150毫秒,偏差刺激出现的时间概率分别为1/9.00、1/4.50、1/2.25或1/1.125秒(序列概率分别为1/60、1/30、1/15或1(偏差刺激)/7.5(标准刺激))。在另一种条件下,SOA为600毫秒,时间概率分别为1/9.00、1/4.50或1/2.25秒(序列概率分别为1/15、1/7.5或1/3.75)。在最后一种条件下,SOA为2400毫秒,时间概率为1/9.00秒(序列概率为1/3.75)。序列概率和时间概率对MMN均有显著影响。当偏差刺激每2.25秒、4.50秒或9.00秒出现一次时,MMN会随着时间概率的降低而增大。当偏差刺激每7.5个或15个标准刺激出现一次时,序列概率越低,MMN越大,但这种效应并不显著。然而,当时间概率保持在1/9.00秒不变时,MMN会随着序列概率的降低而增大。在刺激呈现速度较快时,MMN最大。然而,当偏差刺激出现的概率非常高时,MMN会减弱,这可能是由于其产生源的不应期所致。在最慢的速度下,MMN会减小,这可能是由于对标准刺激的记忆衰退所致。