Robust population orientation encoding by orientation-untuned neurons in macaque V1.

Area of Science:

• Neuroscience
• Computational Neuroscience
• Visual Perception

Background:

• The primary visual cortex (V1) contains orientation-selective neurons, but a subset lacks this selectivity.
• The functional role of these orientation-untuned V1 neurons in population orientation discrimination remains poorly understood.

Purpose of the Study:

• To investigate the role of orientation-untuned V1 neurons in encoding stimulus orientation and enhancing coding robustness.
• To examine the impact of orientation-untuned units in artificial neural network models of the primate visual pathway.

Main Methods:

• Analysis of 2-photon calcium imaging data from thousands of V1 neurons in awake macaques responding to grating stimuli.
• Population analysis to decode stimulus orientation from neuronal responses.
• Modeling the primate ventral visual pathway using artificial neural networks.

Main Results:

• Orientation-untuned V1 neurons independently decode stimulus orientation with accuracy comparable to tuned neurons.
• Orientation-untuned neuronal populations enhance coding robustness by reducing noise sensitivity.
• Removing orientation-untuned units in artificial neural network models impaired natural object recognition.

Conclusions:

• Orientation-untuned neurons encode orientation information and are vital for primate visual perception.
• These neurons contribute significantly to coding robustness and noise reduction.
• The findings support a continuous distribution of visual features across neurons, challenging notions of strict specialization.