Stimulus-Change Probability and Inter-Stimulus Interval Effects on the Acoustic Change Complex

Abstract

Introduction: The Acoustic change complex (ACC) is an obligatory cortical auditory evoked potential (CAEP) evoked by a change in the acoustic properties of an ongoing stimulus. The ACC may be considered a bio-marker of the stimulus change as discriminated at the level of the primary auditory cortex. Inter-stimulus interval (ISI) and stimulus-change probability are factors that affect the CAEP P1-N1-P2 onset response latency and amplitude. Specifically, CAEP onset responses increase in amplitude as stimulus ISI increases and probability decreases. It is likely that the effects of ISI and stimulus-change probability on the ACC are similar to those for the P1-N1-P2 onset complex. The present study was undertaken to test that prediction. The overarching goal of this work is to develop valid, efficient, sensitive and specific methods for using ACC in diagnostic audiology. Methods: Participants: Twenty young adults (mean age =27 years) with normal (<20 dB HL) pure tone hearing thresholds participated in this study. Stimuli: The stimuli were 1000 Hz and 2000 Hz pure tones, each with 400 ms total duration, which were further shaped with a 20 ms Hanning-filter rise-fall envelope and presented at 80 dBA SPL. The stimulus change direction was always from 2000 Hz to 1000 Hz. Two experiments were conducted. Tokens for Experiment 1 were created by concatenating the 400 ms tokens to create an 800 ms token. The ISI was defined as the time between the offset of the 800 ms token to the onset of the next 800 ms token. The two ISIs tested were 1 and 250 ms. Experiment 2 utilized an “odd-ball” stimulus paradigm in which the 2000 Hz toneburst served as the standard and the 1000 Hz token as the deviant. The response to the deviant is the ACC. The standard vs. deviant probabilities tested were 50-50%, and 80-20% and the two ISIs tested were 1 and 250 ms. In this case, the ISI was the time between stimulus offset and stimulus onset for the 400 ms token. Control trials, for which there was no stimulus change, were obtained for each ISI and probability condition in both Experiment 1 and Experiment 2. Results: Experiment 1: ACCs were present for all stimulus change trials for both ISI conditions. Analyses of variance indicated no significant differences in the amplitudes or latencies of ACC components as a function of stimulus ISI. Mean amplitudes for the ACC P1-N1’ and N1-P2’ components were 1.15 μV and 1.19 μV, respectively. Mean latencies for the ACC components P1’, N1’ and P2’ were 72.3 ms, 133.7 ms, and 197.4 ms, respectively. Experiment 2: ACCs were present for all stimulus change trials for every combination of ISI and probability. Analysis of variance revealed that the stimulus ISI and probability had significant effects on ACC component amplitudes. P1’-N1’ amplitudes were significantly lower for the 1 ms ISI (p=0.0071) and 50% probability (p<0.0001) conditions, with no significant interaction effects. Mean amplitudes ACC P1’-N1’ were 1.71 μV for 1ms ISI-50% condition compared to 4.58 μV for 250 ms ISI-20% probability conditions. For N1’-P2’ amplitude, only probability had a significant effect, with amplitudes smaller in the 50% probability condition (p<0.0001), mean =1.95 μV, compared to the 20% probability condition, mean =4.66 μV. The analyses indicated a significant difference in the ACC P1’ latency as a function of both ISI (p=0.002) and probability (p=0.0007), with no interaction effects. Significantly shorter P1’ latencies were found for the 250 ms ISI and 20% probability conditions. Stimulus probability (p=0.0164), but not ISI, also had an effect on ACC P2’, with latencies significantly earlier for the 50% probability compared to the 20% probability. Discussion: Reduced amplitude and prolonged latencies of peaks at short ISIs may be attributed to neural refractoriness. Longer ISIs lead to availability of longer recovery periods for neurons to overcome refractory effect, thus allowing the neurons to respond with higher discharge rates, resulting in larger amplitude response. Similarly, the interval between the frequency changes would be shorter in the 50% stimulus-change probability conditions in comparison to 20% probability conditions. Neurons contributing to the ACC, would have already responded for the previous frequency change, and entered refractory state, resulting in smaller amplitude ACC. The 20% probability condition is essentially an increased ISI condition (for stimulus change) compared to the 50% probability condition. This study indicates that the ISI of the change token is the major driver of ACC amplitude, and even a change probability of 20% leads to substantial gains in ACC amplitude and therefore, detectability. This has implications for the clinical use of ACC, in that a 50% probability may be used in the first instance, as a quick screen, but if the ACC is not present at that probability, a 20% probability condition should be tested.