Gradient vector flow based active shape model for lung field segmentation in chest radiographs

Abstract

Accurate lung field segmentation is crucial to computer-aided diagnosis (CAD) of lung diseases such as lung cancer and tuberculosis (TB). In this paper, we propose a modified gradient vector flow based active shape model (GVF-ASM) for lung field extraction from chest radiographs. Experimental results show that the proposed technique provides around 3-5% improvement over the ASM techniques.