A computational relationship between thalamic sensory neural responses and contrast perception

Area of Science:

• Neuroscience
• Computational Neuroscience
• Sensory Perception

Background:

• Perceptual decisions integrate signals from neural populations, typically studied in the sensory cortex.
• Neural activity in the sensory thalamus during behavior is less understood, leaving gaps in how thalamic information forms perception.
• Previous work showed individual LGN neurons correlate with behavior despite poor sensitivity and lack of interneuronal correlation.

Purpose of the Study:

• To computationally investigate how lateral geniculate nucleus (LGN) neural responses relate to contrast detection behavior.
• To determine the number of LGN neurons involved in perceptual decisions and how their information is pooled.
• To test hypotheses about the role of thalamic neural activity in contrast perception.

Main Methods:

• Computational modeling of perceptual decisions using varying numbers of LGN neurons.
• Analysis of neural responses from the macaque LGN during a contrast detection task.
• Differential weighting of individual neurons based on signal-to-noise ratios (d-primes) in pooling models.

Main Results:

• The most successful computational model for perceptual decisions involved 50-200 LGN neurons.
• Differential weighting of neurons, based on their signal-to-noise ratios, best explained the behavioral data.
• Individual LGN neuron response fluctuations were found to influence contrast perception.

Conclusions:

• The perceptual decision pool for contrast detection likely comprises multiple thalamic neurons.
• Thalamic neurons, particularly those with higher sensitivity (d-primes), play a crucial role in shaping contrast perception.
• Understanding thalamic contributions is key to deciphering the neural basis of sensory perception.