Investigating the neural correlates of a streaming percept in an informational-masking paradigm

Abstract

Humans routinely segregate a complex acoustic scene into different auditory streams, through the extraction of bottom-up perceptual cues and the use of top-down selective attention. To determine the neural mechanisms underlying this process, neural responses obtained through magnetoencephalography (MEG) were correlated with behavioral performance in the context of an informational masking paradigm. In half the trials, subjects were asked to detect frequency deviants in a target stream, consisting of a rhythmic tone sequence, embedded in a separate masker stream composed of a random cloud of tones. In the other half of the trials, subjects were exposed to identical stimuli but asked to perform a different task—to detect tone-length changes in the random cloud of tones. In order to verify that the normalized neural response to the target sequence served as an indicator of streaming, we correlated neural responses with behavioral performance under a variety of stimulus parameters (target tone rate, target tone frequency, and the "protection zone", that is, the spectral area with no tones around the target frequency) and attentional states (changing task objective while maintaining the same stimuli). In all conditions that facilitated target/masker streaming behaviorally, MEG normalized neural responses also changed in a manner consistent with the behavior. Thus, attending to the target stream caused a significant increase in power and phase coherence of the responses in recording channels correlated with an increase in the behavioral performance of the listeners. Normalized neural target responses also increased as the protection zone widened and as the frequency of the target tones increased. Finally, when the target sequence rate increased, the buildup of the normalized neural responses was significantly faster, mirroring the accelerated buildup of the streaming percepts. Our data thus support close links between the perceptual and neural consequences of the auditory stream segregation.

Figure 1. Stimulus paradigm and behavioral performance.

(A) Schematic representation of the stimulus design. A rhythmic sequence of pure tones (target sequence, red) is placed within a background of randomly distributed (in time and frequency) tones (maskers, yellow) and ‘protected’ by a spectral zone with no stimulus energy (green region). In the target task, subjects detected a randomly occurring frequency-shifted tone (red arrow). In the masker task, subjects were instructed to detect an elongation of all constituent tones of the masker in a 0.5 s time window (blue arrows). Each trial contained only one type of deviant, both, or none. Subjects performed the tasks in separate blocks, with the order counterbalanced across subjects. (B) Behavioral performance in the target task as a function of protection zone in a range from 0 to 16 semitones (Psychoacoustic experiment A, N = 14) (C) Behavioral build-up of detection in the target task. Histogram of time constants obtained from exponential fitting to the buildup curves of the behavioral responses as a function of the size of the protection zone (0 to16 semitones). The inset shows behavioral buildup of target task detection for a sample subject illustrating the changes in the buildup speed as a function of different protection zone sizes.

Figure 2. Behavioral performance improvement with target sequence rate reflected in neural build-up curve.

(A) Behavioral performance (Psychoacoustic experiment B, N = 12) as a function of target sequence rate for an expanded range from 2 to 10 Hz, in steps of 2 Hz. Overall performance increased with presentation rate, eventually reaching the ceiling value of d' = 4.1 (for 200 trials) (B) Build-up of the behavioral performance as a function of presentation rate. The time for achieving ceiling detection performance is reduced for faster presentation rates. Results are depicted as median and [25,75]% percentiles.

Figure 3. Attention modulates the normalized neural response.

(A) The power at the target sequence rate is larger in the target task compared to the masker task (MEG experiment D, N = 12, 20 best channels selected for each participant, see Methods for details). (B) Normalized neural response to the target sequence for each participant is plotted in target-masker normalized response space for each participant. The normalized neural response is computed as the ratio of the neural response power at the target sequence rate (7 Hz) to the average power of the background neural activity (from 6–8 Hz). Error bars represent the standard error for the target task (red, orthogonal bars) and the masker task (blue, horizontal bars). Inset: the MEG magnetic field distributions of the target rhythm response component for a single participant, with red and green representing the target magnetic field strength projected onto a line with constant phase.

Figure 4. Larger protection zones ease the target task, but not the masker task.

(A) Behavioral performance and neural results (MEG experiment C, N = 12) for the target task (left panel) and the masker task (right panel), as a function of protection zone. (B) Analysis of neural and behavioral build-up over time for the target task. Behavioral performance (left panel) and normalized neural responses, normalized with respect to the masker task neural response power (right panel) are plotted as a function of time for both the 4 and 7 Hz target sequence rate (orange and green curves, respectively), averaged over participants. Data shown for the 4 Hz target rate is obtained from the study by Elhilali et al. 2009. Neural responses and corresponding behavioral performances are acquired only for the 8 semitones protection zone.

Figure 5. Bottom-up saliency of the target sequence increases for higher target frequencies.

(A) Behavioral and neural responses (MEG experiment C, N = 12) as a function of target frequency. In the left panel, the red/orange line corresponds to behavioral and neural responses for the target task with respect to the low and high target frequency. In the left panel dark/light blue corresponds to behavioral and neural responses for the masker task. Error hull represent 1 SEM. (B) Correlation of the behavioral and neural responses as a function of target frequency. The ratio of the neural to behavioral response differences as a function of target frequency is averaged across participants. A mean slope angle of 42.4° for target (left plot) task and −29.8° for masker (right plot) task (yellow line) were obtained in this analysis. As detailed in Methods, the slope angle corresponds to the strength of correlation between neural and behavioral data. Bootstrap estimates (overlying green lines) and their 95% confidence intervals (pink and blue background for the target and masker task, respectively) confirm the positive/negative correlations for target/masker task.

Figure 6. Stream formation recruits more widespread brain areas at the target sequence rate.

Power enhancement during the target task. The difference between the normalized neural responses in the target task versus the masker task (MEG experiment D, N = 12) shows a significant and highly precise enhancement at the frequency of the target sequence (7 Hz, circled in red). Error bars represent 1 SEM in each graph.