Pupillometric decoding of high-level musical imagery

Area of Science:

• Cognitive Neuroscience
• Auditory Perception and Imagery
• Psychophysiology

Background:

• Humans frequently engage in auditory imagery, mentally hearing sounds like music or speech without external stimuli.
• While neural correlates of auditory imagery exist, decoding the specific content of these mental sounds has remained challenging.

Purpose of the Study:

• To introduce and validate a novel pupillometric method for decoding the content of auditory imagery.
• To investigate if pupil size dynamics (PSDs) can serve as a neural signature for imagined sounds.

Main Methods:

• Participants listened to music and then silently imagined it while their eye movements were tracked.
• Pupil size dynamics (PSDs) were recorded, independent of ambient light, linked to noradrenergic activity.
• Machine learning algorithms were employed to decode imagined songs from PSDs, trained on both perceived and imagined music.

Main Results:

• The study successfully decoded imagined songs using PSDs, both within and across participants.
• Decoding accuracy correlated positively with self-reported imagery strength, similar to findings in visual imagery.
• Identifiable patterns in pupil size dynamics were observed during auditory imagination.

Conclusions:

• Light-independent pupil size dynamics (PSDs) provide a sensitive neural signature for decoding auditory imagination.
• This pupillometric approach offers a new window into understanding the contents of mental auditory experiences.