| Popis: |
Abstract Background Assessments of gaze direction (eye movements), pupil size, and the pupillary light reflex (PLR) are critical for neurological examination and neuroscience research and constitute a powerful tool in diverse clinical settings ranging from critical care through endocrinology and drug addiction to cardiology and psychiatry. However, current bedside pupillometry is typically intermittent, qualitative, manual, and limited to open-eye cases, restricting its use in sleep medicine, anesthesia, and intensive care. Methods We combined short-wave infrared (SWIR, ~0.9-1.7μm) imaging with image processing algorithms to perform rapid (~30 ms) pupillometry and eye tracking behind closed eyelids. Forty-three healthy volunteers participated in two experiments with PLR evoked by visible light stimuli or directing eye movements towards screen targets. Imaging was performed simultaneously on one eye closed, and the other open eye serving as ground truth. Data analysis was performed with a custom approach quantifying changes in brightness around the pupil area or with a deep learning U-NET-based procedure. Results Here we show that analysis of SWIR imaging data can successfully measure stimulus-evoked PLR in closed-eye conditions, revealing PLR events in single trials and significant PLRs in nearly all individual subjects, as well as estimating gaze direction. The neural net-based analysis could successfully use closed-eye SWIR data to recreate estimates of open-eye images and assess pupil size. Conclusions Continuous touchless monitoring of rapid dynamics in pupil size and gaze direction through closed eyes paves the way for developing devices with wide-ranging applications, fulfilling long-standing goals in clinical and research fields. |
Full text from SpringerLink