Abstract
Statistical learning of transition patterns between sounds—a striking capability of the auditory system—plays an essential role in animals’ survival (e.g., detect deviant sounds that signal danger). However, the neural mechanisms underlying this capability are still not fully understood. We recorded extracellular multi-unit and single-unit activity in the auditory forebrain of awake male zebra finches while presenting rare repetitions of a single sound in a long sequence of sounds (canary and zebra finch song syllables) patterned in either an alternating or random order at different inter-stimulus intervals (ISI). When preceding stimuli were regularly alternating (alternating condition), a repeated stimulus violated the preceding transition pattern and was a deviant. When preceding stimuli were in random order (control condition), a repeated stimulus did not violate any regularities and was not a deviant. At all ISIs tested (1 s, 3 s, or jittered at 0.8–1.2 s), deviant repetition enhanced neural responses in the alternating condition in a secondary auditory area (caudomedial nidopallium, NCM) but not in the primary auditory area (Field L2); in contrast, repetition suppressed responses in the control condition in both Field L2 and NCM. When stimuli were presented in the classical oddball paradigm at jittered ISI (0.8–1.2 s), neural responses in both NCM and Field L2 were stronger when a stimulus occurred as deviant with low probability than when the same stimulus occurred as standard with high probability. Together, these results demonstrate: (1) classical oddball effect exists even when ISI is jittered and the onset of a stimulus is not fully predictable; (2) neurons in NCM can learn transition patterns between sounds at multiple ISIs and detect violation of these transition patterns; (3) sensitivity to deviant sounds increases from Field L2 to NCM in the songbird auditory forebrain. Further studies using the current paradigms may help us understand the neural substrate of statistical learning and even speech comprehension.
Highlights
To study whether the auditory system can detect an oddball stimulus when its onset is not fully predictable, we conducted the classical oddball experiment at a jittered inter-stimulus interval (ISI)
Based on previous results about repetition suppression and deviance detection[9,20,21], we expect: 1) in the classical oddball condition at jittered ISI, a stimulus elicits larger neural responses as oddball than as standard; 2) in the control condition, a stimulus elicits smaller neural responses as deviant (2nd stimulus in the repetition) than as standard (1st stimulus in the repetition); 3) in the alternating condition, the neural responses to the deviant are larger than expected in NCM but not in Field L2 at all tested ISIs because previous studies have shown higher auditory areas are more sensitive to regularities in the sound sequence than primary auditory areas[8,22]
The surprise index (SI) was significantly larger in NCM than in Field L2 (t = 12.209, p < 0.001, n1 = 115, n2 = 150; independent sample t-test), suggesting that neural responses to a sound are more sensitive to the occurrence probability of a sound in NCM than in Field L2
Summary
To study whether the auditory system can detect an oddball stimulus when its onset is not fully predictable, we conducted the classical oddball experiment at a jittered ISI. Based on previous results about repetition suppression and deviance detection[9,20,21], we expect: 1) in the classical oddball condition at jittered ISI, a stimulus elicits larger neural responses as oddball than as standard; 2) in the control condition, a stimulus elicits smaller neural responses as deviant (2nd stimulus in the repetition) than as standard (1st stimulus in the repetition); 3) in the alternating condition, the neural responses to the deviant are larger than expected in NCM but not in Field L2 at all tested ISIs because previous studies have shown higher auditory areas are more sensitive to regularities in the sound sequence than primary auditory areas[8,22]. Because prediction for future stimuli can help reduce the uncertainty of a sound (e.g., word) in a rapid series of sounds (e.g., speech) when individual sounds are noisy (e.g., in a noisy environment), our results may provide insight into the neural mechanisms of rapid speech processing
Sign-in/Register to view highlights & summary Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
