Synthetic brain imaging: grasping, mirror neurons and imitation

Area of Science:

• Neuroscience
• Computational Neuroscience
• Cognitive Neuroscience

Background:

• Current brain imaging models often oversimplify neural activity by focusing on isolated regions.
• A gap exists in directly relating global brain imaging data (PET, fMRI) to underlying neural network computations.
• Integrating diverse neurophysiological data across species presents significant challenges.

Purpose of the Study:

• To develop and extend computational models for analyzing brain imaging data (PET, fMRI).
• To bridge the gap between macroscopic brain activity and microscopic neural circuitry.
• To establish a framework for Synthetic Brain Imaging applicable to various cognitive functions.

Main Methods:

• Utilizing schema-based models for direct analysis of brain imaging data.
• Employing the Synthetic PET imaging method, which uses computational models of neural circuitry based on animal data.
• Extending the Synthetic PET method to functional Magnetic Resonance Imaging (fMRI).

Main Results:

• Demonstrated the application of Synthetic PET for visuo-motor processing in grasping.
• Showcased the extension to Synthetic fMRI for analyzing motor skill imitation, including mirror system data.
• Highlighted the correlation between regional cerebral blood flow (rCBF) and integrated synaptic activity.

Conclusions:

• Synthetic Brain Imaging offers a powerful method to link brain-wide data to neural network function.
• The approach is extendable to fMRI and various cognitive functions, including those without direct animal data.
• Comparative neuroscience and evolutionary arguments are crucial for advancing Synthetic Brain Imaging.