A new computer aided detection system for pulmonary nodule detection in chest radiography

Early detection of lung nodules is extremely important for the diagnosis and clinical management of lung cancer. In this paper, a novel computer aided detection (CAD) system is proposed. The proposed system offers several innovations. First, a computationally simple double localizing region-based active model algorithm is used for lung segmentation. Second, detection of lung nodule candidates is conceived as a filtering process that searches for any region with a spherical structure (where a potential nodule may happen to occur) in chest radiography and eignvalues based Hessian matrix is used to do such a work. Finally, Multiple Massive Training SVMs (MTSVM) classifier is proposed for FPs reduction, which is not only computationally simple, but also has the ability to generalize well even with relatively few training samples. Experimental results suggest that the proposed CAD scheme was superior to others in FPs reduction of lung nodule detection in chest radiograph.