What question did this study set out to answer?

This research aims to evaluate a deep learning model for predicting eye gaze from fMRI signals during movie viewing.

June 28, 2026PsychoradiologyOpen Access

fMRI-Based Prediction of Eye Gaze During Naturalistic Movie Viewing Reveals Eye-Movement–Related Brain Activity

View Full Paper

Ask AI

Bookmark

Authors

Randomized trial demonstrates eye-movement-related brain activity in neuroimaging datasets, suggesting limitations for individual gaze predictions.

This research aims to evaluate a deep learning model for predicting eye gaze from fMRI signals during movie viewing.
Applied pre-trained model from DeepMReye framework in zero-shot setting on three fMRI datasets.
Evaluated model accuracy against camera-based eye-tracking data and inter-subject correlations.
Derived eye-movement-related time series to map associated brain activation.
Predicted gaze at the individual level showed modest correspondence with ground-truth data (r ≈ 0.24–0.37).
Group-averaged gaze predictions exhibited higher reliability (r ≈ 0.73–0.84) and revealed widespread activation in oculomotor control regions.
Exploratory analyses indicated inconsistent age-related effects on gaze prediction and brain activity across datasets.

Gao et al. (2026) studied this question.

View Full Paper

Ask AI

Bookmark

Randomized trial demonstrates eye-movement-related brain activity in neuroimaging datasets, suggesting limitations for individual gaze predictions.