Cell density features from histopathological images to differentiate non-small cell lung cancer subtypes

摘要

Histopathological evaluation plays a crucial role in the process of understanding lung cancer biology. Suchevaluation consists in analyzing patterns related with tissue structure and cell morphology to identify the presenceof cancer and the associated subtype. This investigation presents a multi-level texture approach to differentiatethe two main lung cancer subtypes, adenocarcinoma (ADC) and squamous cell carcinoma (SCC), by estimatingglobal spatial patterns in terms of cell density. Such patterns correspond to texture features computed from celldensity distribution in a co-occurrence frame. Results using the proposed approach achieved an accuracy of 0:72and F-score of 0:72.