Retinal Vascular Box-Counting Fractal Dimension (Df) Skill

Zero-dependency sovereign personal agent skill for non-invasive evaluation of retinal microvascular box-counting fractal dimension (Df), arborization complexity, and capillary rarefaction.

flowchart TD
    Fundus["Retinal Fundus Image / Vessel Segmentation"] --> MultiScale["Multi-Scale Box Partitioning (Epsilon)"]
    MultiScale --> BoxCount["Occupied Box Enumeration N(Epsilon)"]
    BoxCount --> LogRegression["Linear Regression (log N vs log 1/Epsilon)"]
    LogRegression --> FractalScore["Fractal Dimension Df & Complexity Score"]
    FractalScore --> Agent["OpenClaw / Hermes Personal Agent Actions"]

Features

• Deterministic Math: Pure Python standard library implementation with zero external dependencies.
• Mathematical Fractal Geometry: Implements rigorous logarithmic box-counting and Pearson correlation validation.
• Microvascular Aging Stratification: Non-invasive detection of capillary rarefaction and cerebral small vessel health.
• Model Context Protocol (MCP): Native JSON-RPC tool integration for LLMs and autonomous agents.

Quickstart

from client import RetinalVesselFractalDimensionClient

client = RetinalVesselFractalDimensionClient()
result = client.evaluate_fractal_dimension(
    box_sizes=[4.0, 8.0, 16.0, 32.0, 64.0, 128.0],
    box_counts=[1745, 640, 235, 86, 31, 11]
)
print("Fractal Dimension Df:", result["fractal_dimension_df"])
print("Classification:", result["classification"])

License

Apache-2.0