Investigating common coding of observed and executed actions in the monkey brain using cross-modal multi-variate fMRI...

Area of Science:

• Neuroscience
• Primate Cognition
• Motor Control

Background:

• Mirror neurons are hypothesized to underpin action understanding by mirroring observed actions onto the observer's motor system.
• Investigating human mirror neuron systems often relies on functional magnetic resonance imaging (fMRI) due to limitations in single-neuron recordings.
• Cross-modal multivariate pattern analysis (MVPA) of fMRI data is proposed as a method to infer mirror neuron activity.

Purpose of the Study:

• To validate the use of cross-modal MVPA of fMRI data in non-human primates to study the mirror neuron system.
• To bridge the gap between electrophysiological findings in monkeys and fMRI studies in humans regarding action representation.

Main Methods:

• Rhesus monkeys performed and observed reach-and-grasp and reach-and-touch actions.
• fMRI data were analyzed using unimodal and cross-modal MVPA to decode executed and observed actions.
• Decoding accuracy was assessed in various brain regions, focusing on areas known to contain mirror neurons.

Main Results:

• Both executed and observed actions were decodable from numerous brain regions using unimodal analysis.
• Specific regions, including rostral parietal, premotor, motor, and somatosensory cortices, showed significant asymmetric action-specific cross-modal decoding.
• These findings indicate shared action representations between self-execution and observation.

Conclusions:

• Cross-modal MVPA of fMRI data is a valid technique for investigating action representations in the primate brain.
• The results support the hypothesis that observed actions are mapped onto the observer's own motor system, consistent with mirror neuron function.
• This study validates a crucial methodology for future research on the neural basis of action understanding across species.