Optimizing MR-based gaze-decoding for eyes-closed eye-tracking in fMRI
- 1Institut de Neurosciences de la Timone, CNRS, Aix-Marseille Universite
Abstract
Eye movements provide valuable insights into human cognition and are a critical variable in numerous functional magnetic resonance imaging (fMRI) studies. Yet, when the eyes are closed, camera-based eye-tracking is unavailable, making studies of eyes-closed states challenging. Here, we address this gap using DeepMReye, a deep learning framework for camera-free gaze reconstruction from the MR-signal of the eyes. We first show that fine-tuning DeepMReye on visuomotor calibration data acquired with the eyes open significantly improves gaze decoding, and that this fine-tuning does not require simultaneous camera-based data. We next assessed whether decoding could be extended to eyes-closed states using a novel auditory-guided task, in which participants gazed at learned target positions with and without visual input, and with their eyes open, blinking, or closed. While DeepMReye was originally trained exclusively on eyes-open data, the network successfully generalized to eyes-closed periods, and this generalization was further improved through task-specific fine-tuning. Finally, fine-tuning also improved decoding of eyelid-state (open versus closed) directly from the MR-signal. These findings demonstrate that gaze and eyelid-state monitoring during eyes-closed periods is feasible, enabling broader integration of eye-tracking in fMRI research.
Related
- Dynamic Gaze Patterns Predict Performance on the Reading the Mind in the Eyes Test
- Exploring the robustness of permutation entropy analysis to differentiate between closed-eyes and open-eyes resting states
- Aesthetic preference and gaze-based visual attention in atlas-cover design: Evidence from an eye-tracking study
- Longitudinal changes in resting-state EEG eyes-closed/open delta differentiation during an Antarctic expedition
- Decoding Social Integration in Schooling Fish Using Closed-Loop Real-Virtual Interactions
BibTeX
@misc{kling2026optimizing,
title = {{Optimizing MR-based gaze-decoding for eyes-closed eye-tracking in fMRI}},
author = {S. M. Kling and U. Lascombes and M. Nau and G. S. Masson and M. Szinte},
howpublished = {bioRxiv},
year = {2026},
month = jul,
doi = {10.64898/2026.07.07.736972},
url = {https://www.biorxiv.org/content/10.64898/2026.07.07.736972v2},
note = {CC BY},
}
RIS
TY - UNPB
TI - Optimizing MR-based gaze-decoding for eyes-closed eye-tracking in fMRI
AU - S. M. Kling
AU - U. Lascombes
AU - M. Nau
AU - G. S. Masson
AU - M. Szinte
PY - 2026
DA - 2026/07/07
PB - bioRxiv
DO - 10.64898/2026.07.07.736972
UR - https://www.biorxiv.org/content/10.64898/2026.07.07.736972v2
AB - Eye movements provide valuable insights into human cognition and are a critical variable in numerous functional magnetic resonance imaging (fMRI) studies. Yet, when the eyes are closed, camera-based eye-tracking is unavailable, making studies of eyes-closed states challenging. Here, we address this gap using DeepMReye, a deep learning framework for camera-free gaze reconstruction from the MR-signal of the eyes. We first show that fine-tuning DeepMReye on visuomotor calibration data acquired with the eyes open significantly improves gaze decoding, and that this fine-tuning does not require simultaneous camera-based data. We next assessed whether decoding could be extended to eyes-closed states using a novel auditory-guided task, in which participants gazed at learned target positions with and without visual input, and with their eyes open, blinking, or closed. While DeepMReye was originally trained exclusively on eyes-open data, the network successfully generalized to eyes-closed periods, and this generalization was further improved through task-specific fine-tuning. Finally, fine-tuning also improved decoding of eyelid-state (open versus closed) directly from the MR-signal. These findings demonstrate that gaze and eyelid-state monitoring during eyes-closed periods is feasible, enabling broader integration of eye-tracking in fMRI research.
ER -