Neural coding of multiple motion speeds in visual cortical area MT.

Area of Science:

• Neuroscience
• Computational Neuroscience
• Vision Science

Background:

• Motion speed is crucial for visual segmentation.
• The neural representation of multiple speeds is not well understood.

Purpose of the Study:

• Investigate the encoding and decoding of multiple speeds in the visual system.
• Characterize human and macaque perceptual abilities in segmenting stimuli with different speeds.
• Determine how neurons in area MT represent multiple speeds.

Main Methods:

• Behavioral experiments with human and macaque subjects to assess perceptual segmentation of stimuli with varying speeds.
• Electrophysiological recordings from neurons in area MT of macaque monkeys.
• Development and application of a modified divisive normalization model to explain neural responses.
• Machine learning classifiers to decode speed information from neural population activity.

Main Results:

• MT neuron responses to two speeds exhibited a bias towards the faster speed component, particularly at lower speeds.
• This faster-speed bias diminished as stimulus speed increased.
• A classifier could differentiate MT responses to two speeds versus a single log-mean speed.
• Decoding two speeds from MT population responses correlated with perceptual performance for large speed separations (4x) but not small ones (2x).

Conclusions:

• Findings suggest a specific neural coding rule for multiple speeds, involving a faster-speed bias in area MT.
• The observed bias may enhance behavioral tasks like figure-ground segregation.
• A modified divisive normalization model can explain the neural representation of multiple speeds.