Coding visual images of objects in the inferotemporal cortex of the macaque monkey

Area of Science:

• Neuroscience
• Computational Neuroscience
• Visual Perception

Background:

• The inferotemporal cortex (IT) is crucial for visual object recognition and discrimination.
• Lesions in the monkey IT impair visual learning tasks, highlighting its importance.

Purpose of the Study:

• To investigate the cellular mechanisms underlying object recognition in the inferotemporal cortex.
• To determine the optimal stimuli that activate individual IT neurons.

Main Methods:

• Recording neuronal activity from anesthetized, immobilized monkeys.
• Presenting visual stimuli (bars, disks, complex shapes) to determine optimal cell activation.
• Analyzing the features that maximally activate individual IT cells.

Main Results:

• Posterior IT cells respond to simple features like bars or disks (size, orientation, color).
• Anterior IT cells require more complex features for maximal activation, varying between cells (e.g., shape, texture, color combinations).
• These complex features represent partial object attributes, with sharp selectivity.

Conclusions:

• Anterior IT neurons are tuned to moderately complex partial features of objects.
• Object images are likely coded in anterior IT by combinations of neurons, each representing a specific partial feature.
• This provides insight into the neural basis of visual object recognition.