7.2. [ FRACTAL ] ln (N(r)) Df: fractal dimension = degre of similarity • Df = ———— with N(r): number of pattenrs at scale r ln (1/r) r: scale [Length] |→ If Df increases ⇒ Self-similarity increases
7.3. [ MICROSCOPE USE ] • Bright field image conerted to binary ⇒ Graph of Df plotted ⇒ Slope = self-similarity degree
8. Results
8.1. Pancreas results: Df(I) = Df(normal) +4% Df(II) = Df(normal) +6% Df(III) = Df(normal) +9% (Same increase shapes for breast, colon and prostate) ⇒ As expected, Df increases with tumor stage
9. Conclusion
9.1. → Not the only study on the subject → Can detect QUANTITATIVELY and also lead to a large international OBJECTIVE database ⇒ Refinement of the method ⇒ Study of the demographic evolution of cancers