On the origin of sensory errors: Contrast discrimination under temporal constraint

Area of Science:

• Neuroscience
• Cognitive Science
• Computational Neuroscience

Background:

• Perceptual variable estimation is inherently imprecise and error-prone.
• The physiological underpinnings of these perceptual errors remain largely unknown.
• Trial-by-trial variability in sensory neuron responses is a hypothesized source of perceptual error.

Purpose of the Study:

• To develop a mathematical framework for characterizing the physiological basis of perceptual errors.
• To determine the time scale of visual information integration.
• To test the hypothesis that sensory neuron variability causes perceptual errors.

Main Methods:

• Developed a novel mathematical formalism to analyze perceptual errors.
• Estimated statistical properties of the underlying physiological system from behavioral data.
• Quantified the time window for perceptual evidence integration.

Main Results:

• Identified a constant level of noise contributing to perceptual errors, independent of stimulus intensity and duration.
• Demonstrated that perceptual errors are unlikely to originate from variability during the neural encoding stage.
• Determined that visual information is integrated over a time window of approximately 230 ms.

Conclusions:

• Perceptual errors are not solely explained by variability in sensory neuron responses during encoding.
• The integration of perceptual evidence occurs within a relatively short time window.
• Further investigation into alternative sources of error consistent with behavioral data is warranted.