Back

iSleep: Continuous, binocular pupil tracking in sleep and reduced consciousness for physiological monitoring, predictions and interventions

Yuzgec, O.; Legendre, G.; Asadian, A.; Gloria, Y.; Bayer, L.; Tourdot, Q.; Pellat, S.; Novozhilova, S.; D'Ancona, C.; Ulanova, M.; Kecik, M.; Maccari, A.; Lopez, D.; Licini, L.; Nguyen Tang, G.; Van Delden, S.; Bravetti, G. E.; Vo, T.; Kropp, M.; Boex, C.; Jakus, L.; Schaller, K.; Sollander, F.; Vutskits, L.; Thumann, G.; Schwartz, S.; Huber, D.

2025-06-15 neuroscience
10.1101/2025.06.13.659058 bioRxiv
Show abstract

Monitoring pupil dynamics is a key tool in understanding arousal. Pupil size can serve as a biomarker for the autonomic nervous system balance as well as for identifying brain states. While internal states can also be self-reported when awake, automated detection and non-invasive monitoring is crucial during sleep and reduced consciousness. Here, we introduce iSleep, an innovative pupil tracking and analysis framework for sleep in humans. It features comfortable, humidified eye-tracking goggles and a platform for integrated analysis and prediction capabilities. We show that iSleep allows safe and continuous access to binocular pupil size and ocular dynamics during sleep and anesthesia. iSleep reveals that pupillary fluctuations correlate tightly with brain activity, heartbeat, and breathing, and can reliably predict brain states. Pupil constrictions reflect parasympathetic drive and likely serve a protective function for deep sleep stability; while dilations indicate arousals. Unexpectedly, we observed a decoupling of binocular movements during periods of sleep, indicating alterations in reflexes which usually govern voluntary eye movements. Finally, iSleep was tested in surgery patients under general anesthesia, revealing dynamic pupil changes to noxious stimuli, suggesting the potential for nociception monitoring during surgeries. In summary, iSleep offers an easy-to-use, robust alternative to read out brain states during sleep and anesthesia, opening new avenues in monitoring, diagnostics, and treatments, previously obscured by closed eyelids.

Matching journals

The top 5 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.