Combined fMRI- and eye movement-based decoding of bistable plaid motion perception

Area of Science:

• Neuroscience
• Cognitive Science
• Psychology

Background:

• Bistable perception, where perception shifts spontaneously under constant stimulation, is crucial for studying conscious perception.
• Current methods rely on active observer reports, which can introduce bias and confound results.
• Alternative methods using eye or brain signals aim to bypass subjective reporting.

Purpose of the Study:

• To optimize the decoding of perceptual alternations in bistable perception by combining eye and brain signals.
• To develop a more objective measure of perceptual dynamics, reducing reliance on observer reports.

Main Methods:

• Eye-tracking and functional magnetic resonance imaging (fMRI) were used in 20 participants viewing a bistable visual plaid motion stimulus.
• Multivoxel pattern analysis (MVPA) of fMRI data was combined with eye-tracking data using a Support Vector Machine (SVM).
• The SVM was trained to decode participants' perceptual time courses from the combined signals.

Main Results:

• Individual measures (eye-tracking: 86% accuracy; fMRI: 88% accuracy) showed high decoding performance.
• Combining eye-tracking and fMRI data significantly improved decoding accuracy to 91%.
• This demonstrates the enhanced ability to track perceptual dynamics using multimodal data.

Conclusions:

• Combining fMRI and eye movement data offers a more accurate method for decoding bistable perception.
• This multimodal approach facilitates the development of optimized no-report paradigms for studying consciousness.
• Improved decodability advances our understanding of the neural correlates of conscious perception.