Integration of distributed cortical systems by reentry: a computer simulation of interactive functionally segregated...

Area of Science:

• Computational neuroscience
• Visual cortex modeling
• Neural network simulation

Background:

• Understanding how segregated cortical areas integrate information for unified perception is a key challenge.
• The proposed mechanism for cortical integration is reentry, involving recursive signaling between separate neural maps.

Purpose of the Study:

• To develop and test a computer model of reentrant cortical integration (RCI) for analyzing perceptual coordination.
• To simulate interactions between simplified models of macaque visual areas (VOR, VOC, VMO) based on V1, V3, and V5.

Main Methods:

• Construction of a computer model simulating three interconnected cortical areas (VOR, VOC, VMO) with over 222,000 units and 8.5 million connections.
• Simulation of responses to visual stimuli including illusory contours and structure-from-motion.
• Analysis of functional integration based on three reentrant processes: competitive elimination, mutual output utilization, and self-reentry.

Main Results:

• The RCI model successfully simulated responses to illusory contours and structure-from-motion, consistent with empirical observations.
• A novel combined illusion elicited responses supporting the necessity of recursive reentrant processing.
• Disruption of reentrant connections abolished figural synthesis, highlighting the critical role of reentry.

Conclusions:

• Reentry provides a viable mechanism for functional integration across multiple cortical areas.
• The RCI model offers a framework for understanding how distributed visual processing leads to unified perception.
• This model has implications for understanding integration in broader cortical systems beyond vision.