Back

Optimizing MR-based gaze-decoding for eyes-closed eye-tracking in fMRI

Kling, S. M.; Lascombes, U.; Nau, M.; Masson, G. S.; Szinte, M.

2026-07-11 neuroscience
10.64898/2026.07.07.736972 bioRxiv
Show 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 through MR-based gaze decoding with DeepMReye, a deep learning framework for camera-free reconstruction of gaze behavior from the MR-signal of the eyes. We first show that fine-tuning DeepMReye using visuomotor calibration data acquired when the eyes were open significantly improves gaze decoding, and that this fine-tuning does not require simultaneous camera-based data. We next assessed whether model performance could be further improved by incorporating data acquired while participants gazed at known positions with both eyes open and closed. Notably, while DeepMReye was originally trained exclusively on eyes-open data, the network successfully generalized eyes-closed periods, with performance improving significantly through fine-tuning on the eyes-closed data. These findings demonstrate that reliable gaze monitoring during eyes-closed periods is feasible, enabling a more effective integration of eye-tracking in fMRI research and, consequently, advancing our understanding of human cognition.

Matching journals

The top 7 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.