Biases in hand perception are driven by somatosensory computations, not a distorted hand model

Area of Science:

• Neuroscience
• Cognitive Science
• Human Perception

Background:

• Accurate hand and finger geometry perception is crucial for environmental interaction.
• Existing research indicates significant biases in human finger geometry perception.
• These biases have been interpreted as evidence of distorted internal body representations.

Purpose of the Study:

• To propose and test an alternative explanation for finger geometry perception biases.
• To investigate whether biases result from Bayesian integration of somatosensory signals.
• To differentiate between sensory-based and space-based coordinate models of integration.

Main Methods:

• Bayesian reverse engineering approach.
• Behavioral experimentation on index finger joint and fingertip localization.
• Modeling Bayesian integration in sensory versus space-based coordinates.

Main Results:

• A space-based coordinate model explained perceptual biases without distorting finger length representation.
• Behavioral data aligned with the space-based model, not the sensory-based model.
• The space-based model outperformed models with inherent geometric biases.

Conclusions:

• Finger geometry perception biases stem from near-optimal Bayesian inference on noisy somatosensory signals.
• These distortions do not reflect an inaccurate internal hand model.
• The findings challenge previous interpretations of distorted body representations.