Quantitative Characterization of Brain Tissue Alterations in Brain Cancer Using Fractal, Multifractal, and IPR Metrics

TL;DR

This study introduces a multiparametric framework combining fractal, multifractal, and IPR metrics to differentiate healthy and cancerous brain tissues, enhancing early diagnosis capabilities.

Contribution

The paper presents a novel integrated approach using fractal, multifractal, and IPR analyses for detailed characterization of brain tissue microstructure in cancer detection.

Findings

Fractal dimension metrics distinguish healthy from cancer tissues.

Multifractal analysis indicates higher heterogeneity in cancer tissues.

IPR analysis shows increased nanoscale structural disorder in cancer tissues.

Abstract

We studied the structural alterations between healthy and diseased brain tissues using a multiparametric framework combining fractal analysis, fractal functional transformation, multifractal analysis, and the Inverse Participation Ratio (IPR) analysis. Accurate characterization of brain tissue microstructure is crucial for early detection and diagnosis of cancer. By applying box-counting methods on brightfield microscopy images, we estimated the fractal dimension (Df) and its logarithmic (ln(Df)) and functional (ln(Dtf)) forms to highlight spatial irregularities in the tissue architecture. While Df and ln(Df) exhibited long-tailed distributions distinguishing healthy from cancer tissues, ln(Dtf) provided significantly improved differentiation by emphasizing local structural variations. Additionally, multifractal analysis revealed broader f({\alpha}) vs {\alpha} curves in cancerous samples, reflecting higher heterogeneity. IPR analysis based on light localization further demonstrated increased nanoscale variations in mass density, reflecting higher structural disorder in cancer tissues. Combining these complementary approaches creates a robust framework for measuring tissue complexity and holds great potential to improve microscopic diagnostic methods for brain cancer detection.