Preview

PDF, English - main document Download (26MB) | Lizenz: Influence of arousal and attention on the perception of auditory near-threshold stimuli by Doll, Laura underlies the terms of Creative Commons Attribution-NonCommercial-NoDerivatives 4.0

Abstract

This thesis investigated the influence of attention and arousal on the processing of near-threshold auditory stimuli. In addition, the role of the decision criterion was examined along with the question of whether perception is graded, or an all-or- nothing phenomenon. For this purpose, three experiments were conducted with different neuroimaging techniques: Two measurements with magnetoencephalography, electroencephalography, and pupillometry, and one measurement combining pupillometry and functional magnetic resonance imaging. The pupil size was recorded as a proxy for the activity of the locus coeruleus and other brainstem nuclei regulating arousal. In Experiment 1, continuous white noise with near-threshold tones and more salient, transient amplitude-modulations of the noise were presented to the listeners. During three runs with the same stimulation, the participants had to detect the noise modulations in the first and third run, and the near-threshold tones in the middle run. They were unaware of the exact contents of the stimulation, and only informed about the respective target stimuli. The tones were perceived only in the second run, and by half of the listeners occasionally in the last run, after they got used to detecting them previously. In line with this, significant amounts of neuronal activity evoked by those tones was only present when they were task-relevant. But in this case, even the undetected tones evoked a pupil dilation response and activity in the auditory cortex. In contrast, the task had less influence on the more salient noise modulations, which were likely perceived also in passive runs and consequently evoked auditory activity in all runs. However, the pupil dilation response was not present for task-irrelevant noise modulations. In line with the expectations, the decision-related P3b was only present for detected task-relevant stimuli. Experiment 2 used the same near-threshold tones, but with short intervals of noise that contained either one or none tone. After each interval, listeners indicated via a button press, how confident they were that a tone was present. For this, six different response options were provided, three each for signal present or absent: signal certainly/likely/uncertainly present or absent. This approach provided the opportunity to investigate, whether the auditory activity and pupil dilation for undetected targets in the first experiment was caused by wrong classification in the context of a bimodal model of perception, where tones perceived with low confidence might have been classified as miss, or whether the neuronal activity is instead coupled to the strength of perception and decreases gradually across the ratings as suggested by a unimodal model of perception. Even though the overall decision criterion was lower than in Experiment 1 (1.6 0.4 vs. 0.6 0.4, mean and standard deviation across participants), the waveforms for hits and misses combined across the three ratings reproduced the previous results. But across the individual ratings, the amplitudes of the auditory cortex activity and the pupil dilation response decreased gradually. Consequently, the evoked responses for missed tones were rather not caused by a misclassification of detected tones. The P3b appeared more bimodal, as it was mainly present for detected tones, but a large variability in latency might have influenced this. All three responses were well explained by a model based on signal detection theory, which postulates a continuous relationship between neural activity and perception. A supplementary psychoacoustic experiment compared the confidence rating to an audibility rating used in a previous publication supporting a bifurcation of perception. The present results did not support this finding. Furthermore, listeners on average estimated the number of signal-to-noise ratios at which the tones were presented to 5-6, even though only one constant ratio was used. This indicates that for physically identical stimuli, the strength of perception varies across different levels, not only between perceived and unperceived. Experiment 3 explored the influence of attention on auditory perception and identified related brain networks. The stimulation was similar to the first experiment, with the exception that instead of noise modulations, supra-threshold amplitude-modulated tones were used, which could be easily perceived despite the scanner noise. The tasks for the three runs changed from passive listening, to detection of the supra-threshold, and finally detection of the near-threshold tones. With the increasing amount of attention directed to the auditory stream, activity in the auditory cortex increased for the near-threshold stimuli already during the second run, even though participants did not perceive those tones yet. Overall, auditory activity was similar to those of Experiments 1 and 2, with significant activation for detected and missed targets, and the salient distractors. Apart from auditory cortex, increased activation for missed targets was also observed in the cingulo-opercular network. Additionally, enhanced pre-stimulus activity for hits compared to misses was observed in this network. This effect is similar to the relationship of pupil size and performance, although the former is presumably monotonic, while the latter is inverted-U-shaped. Thus, pre-stimulus brain activation and pupil size are not directly related. A correlation of pupil size and the blood oxygenation level-dependent signal revealed a widespread interaction between arousal and cortical activity. This activity included the cingulo-opercular network, suggesting that the relationship between pupil dilation and activity within this network is more complex. A preliminary analysis of subcortical activation showed similar modulations by task and similar correlations with pupil size fluctuations as in the cortex. Future measurements with improved spatial resolution could further investigate how these activation patterns differ across the brainstem, and how the different nuclei are connected to phasic and tonic arousal. In summary, the present work confirmed the importance of arousal and attention on the processing of different auditory stimuli, and provided evidence for a differential influence of attention on perception, depending on the type of stimuli. For near-threshold stimuli, these data showed that neural activity is directly connected to the strength of a perception, and that perception is graded rather than bifurcated.