Decoding eye-of-origin outside of awareness

Area of Science:

• Neuroscience
• Visual Perception
• Computational Neuroscience

Background:

• Ocular dominance columns in the visual cortex segregate information from each eye.
• Despite this segregation, humans often cannot consciously identify the eye of origin for visual stimuli.

Purpose of the Study:

• To investigate if eye-of-origin information is present in electrophysiological activity patterns.
• To determine if this information can be decoded using multivariate pattern analysis of EEG data.

Main Methods:

• Electroencephalography (EEG) was used to record brain activity from 24 observers.
• Stimuli (gratings, plaids) were presented to the left, right, or both eyes.
• A support vector machine (SVM) was employed for multivariate pattern analysis to decode eye-of-origin.

Main Results:

• Eye-of-origin was successfully decoded from EEG signals at approximately 140ms and 220ms post-stimulus onset, significantly above chance.
• Human observers performed at chance level when asked to report the eye of origin.
• Other stimulus features like orientation and binocularity were also decodable, but eye-of-origin decoding persisted even when these were not consciously perceived.

Conclusions:

• Electrophysiological signals contain eye-of-origin information that is not consciously accessible.
• Multivariate pattern analysis can decode information organized in cortical columns, irrespective of conscious awareness.
• This supports a structural basis for information processing in the visual cortex, even when awareness is absent.