Fuzzy graph and nonlinear models for medical image segmentation
Irma Ibrišimović, Nebojša Ralević, Bratislav Iričanin, Andrija Blesić · Applicable Analysis and Discrete Mathematics · 2025
Nonlinear analysis and graph theory form a unified framework for examining complex radiological image patterns. Pixel intensities are modeled as graph nodes, while edges represent spatial relations. Fuzzy graphs handle uncertainty, enabling flexible and robust segmentation. Entropy-based nonlinear methods enhance boundary detection and highlight structural irregularities. The framework improves identification of pathological regions and supports advanced image interpretation.