Texture-based image feature analysis for the classification of oral squamous cell carcinoma using machine learning approach

Description

Salma Bahaa, Omneya Ramadan, Zeinab Elsayed Darwish, Eslam El-Fiky, and Mai Saleh used image texture to sort oral cancer slides in twenty twenty-five. Oral squamous cell carcinoma is often caught late, with poor outcomes, and slide reading can be slow and subjective. The team took one hundred fifteen cases, forty-nine cancer and sixty-six normal, at two magnifications, and drew twelve texture features from each image.

Cancer images showed finer texture, more local variation, more complexity, and less uniformity than normal ones. Several machine learning models were trained on these features, alone and fused with clinical data. Many reached perfect scores on catch rate, exactness, and overall accuracy across the test tasks. The authors say texture-based aids could give pathologists a faster, more standard route to diagnosis and lift survival chances.

Catalog record

Source: https://europepmc.org/articles/PMC12809959