Abstract: It has been proposed that the dynamical stability of topological defects in the visual cortex reflects the Euclidean symmetry of the visual world. We analyze defect stability and pattern selection in a generalized Swift-Hohenberg model of visual cortical development symmetric under the Euclidean group E(2). Euclidean symmetry strongly influences the geometry and multistability of model solutions but does not directly impact on defect stability.
The Stability Paradox of Visual Maps
In the mammalian visual cortex, neurons are organized into orientation maps where specific regions respond to edges of different angles. A defining feature of these maps is the "pinwheel"—a topological defect where neurons representing all orientations converge at a single point. While standard pattern-formation models typically predict that these defects should decay into simple stripes, biological pinwheels remain stable throughout an organism's life. This paper investigates whether the Euclidean symmetry of visual space—the invariance of visual laws under translation and rotation—is responsible for this stability.
Modeling Cortical Development
The researchers utilized a generalized Swift-Hohenberg model, incorporating $E(2)$ symmetry to reflect the geometry of visual space. A critical addition to the model is the symmetry-breaking parameter $\epsilon$, which represents "shift-twist" breaking. This allows for the exploration of how prioritizing certain spatial arrangements influences the final cortical map. The model also accounts for long-range interactions between neurons, which prove essential for favoring complex patterns over simple stripes.
Symmetry Breaking and Pattern Selection
Contrary to previous hypotheses, the study found that Euclidean symmetry influences the geometry of the maps but does not inherently stabilize pinwheels. Instead, pinwheel stability is a result of symmetry breaking.
As shown in Figure 1, at $|\epsilon|=0$, simple stripes (plane waves) are energetically favored. As $\epsilon$ increases, the system transitions through intermediate states until pinwheel-rich crystals and aperiodic patterns become the stable, lower-energy configurations. This suggests that the brain's preference for complex maps is driven by a specific level of symmetry breaking that filters out simpler patterns.
The Phase Landscape of Pinwheel Stability
The stability of these patterns is highly dependent on long-range interactions ($g$) and the spatial scale of these interactions ($\sigma/\Lambda$). The authors mapped these relationships using phase diagrams to identify where different modes ($n$) are stable.
Figure 4 maps the stability regions across different values of $\epsilon$. In the absence of symmetry breaking ($\epsilon=0$), stripes dominate the landscape. However, as $\epsilon$ increases (e.g., to 0.2), the "multistability" of the system—where many different patterns could potentially exist—collapses. Complex, pinwheel-rich patterns occupy a much larger portion of the parameter space, effectively forcing the system into a unique, biologically realistic organization.
Conclusion
The research demonstrates that the stability of cortical pinwheels is not a direct consequence of the world's symmetry, but rather a result of how that symmetry is broken during development. By including long-range interactions and shift-twist symmetry breaking, the model successfully predicts the stable, aperiodic orientation maps observed in nature. This work provides a robust mathematical foundation for understanding how the brain self-organizes into complex functional structures.
Summary updated