Multifractal Analysis of H&E Stained Nuclear Images for Differentiation of Early-Grade Renal Cell Carcinoma

Renal cell carcinoma (RCC) is the most common type of kidney cancer, and early-grade detection is crucial for improving patient outcomes. Nuclear morphology plays a vital role in identifying early-grade RCC, as nuclear abnormalities such as structural irregularities and heterogeneity are key indicators of malignancy. In this study, analysis of nuclear morphology is done using multifractal analysis to differentiate RCC from normal renal tissue in hematoxylineosin-stained Whole Slide Images (WSI). Nuclear segmentation is carried out using Detectron2, generating grayscale masks of individual nuclei. The multifractal spectrum is extracted using Multifractal Detrended Fluctuation Analysis (MFDFA), and key multifractal features are evaluated. Results demonstrate that tumor nuclei exhibit a broader multifractal spectrum with higher variability in singularity strength compared to normal nuclei. Statistical analysis confirmed significant differences in multifractal features (Δα and Δf) between tumor and normal nuclei (p<0.05), indicating increased structural heterogeneity in RCC. These findings highlight the potential of multifractal analysis as a quantitative approach for early-grade RCC classification and automated histopathological diagnosisClinical relevance- This study demonstrates that multifractal analysis of nuclear structure can serve as an objective, automated tool for early RCC detection and grading, potentially improving diagnostic precision and informing patient management.