Pinpointing the neural signatures of single-exposure visual recognition memory

Area of Science:

• Neuroscience
• Cognitive Science
• Visual Perception

Background:

• Visual recognition memory is theorized to involve reduced neural responses in high-level visual areas like the inferotemporal (IT) cortex, a phenomenon termed repetition suppression (RS).
• Stimulus contrast is known to modulate neural responses in the IT cortex, potentially confounding interpretations of RS.
• Existing models of RS do not fully account for the impact of stimulus variations like contrast on neural activity.

Purpose of the Study:

• To investigate whether repetition suppression (RS) or an alternative neural mechanism underlies visual recognition memory.
• To differentiate the neural correlates of memory from the effects of stimulus contrast modulation in the inferotemporal (IT) cortex.
• To determine the neural activity pattern that best predicts behavioral performance in a visual recognition memory task.

Main Methods:

• Rhesus monkeys performed a task requiring them to distinguish novel from repeated images, while image contrast was systematically varied.
• Neural responses in the IT cortex were recorded and analyzed in relation to behavioral performance and stimulus properties.
• Computational models inspired by RS and sensory-referenced suppression were used to decode behavioral performance from neural data.

Main Results:

• Monkey behavior was largely invariant to contrast changes, contradicting predictions from a standard RS-inspired decoder.
• An RS-inspired decoder failed to distinguish between responses to repeated images and responses to lower-contrast novel images.
• A modified decoder, incorporating contrast modulation correction (sensory-referenced suppression), accurately predicted behavioral patterns.

Conclusions:

• The findings challenge the hypothesis that repetition suppression (RS) is the primary neural mechanism for visual recognition memory.
• Sensory-referenced suppression, characterized by contrast-corrected reductions in IT population response magnitude, provides a better explanation for neural activity related to single-exposure visual recognition memory.
• Neural representations of visual memory are modulated by sensory factors, necessitating contrast correction for accurate interpretation.