Helical computed tomography (CT) is the most sensitive imaging modality for
detection of pulmonary nodules, However, a single CT examination produces
a large quantity of image data. Therefore, a computerized scheme has been d
eveloped to automatically detect pulmonary nodules on CT images, This schem
e includes both two- and three-dimensional analyses. Within each section. g
ray-level thresholding methods are used to segment the thorax from the back
ground and then the lungs from the thorax. A rolling ball algorithm is appl
ied to the lung segmentation contours to avoid the loss of juxtapleural nod
ules, Multiple gray-level thresholds are applied to the volumetric lung reg
ions to identify nodule candidates. These candidates represent both nodules
and nor mal pulmonary structures. For each candidate, two- and three-dimen
sional geometric and gray-level features are computed. These features are m
erged with linear discriminant analysis to reduce the number of candidates
that correspond to normal structures. This method was applied to a 17-case
database. Receiver operating characteristic (ROC) analysis was used to eval
uate the automated classifier, Results yielded an area under the ROC curve
of 0.93 in the task of classifying candidates detected during thresholding
as nodules or nonnodules.