Three-dimensional multicriterion automatic segmentation of pulmonary nodules of helical computed tomography images

A 3-D multicriterion automatic segmentation algorithm is developed to impro ve accuracy of delineation of pulmonary nodules on helical computed tomogra phy (CT) images by removing their adjacent structures. The algorithm applie s multiple gray-value thresholds to a nodule region of interest (ROI). At e ach threshold level, the nodule candidate in the ROI is automatically detec ted by labeling 3-D connected components, followed by a 3-D morphologic ope ning operation. Once the nodule candidate is found, its two specific parame ters, gradient strength of the nodule surface and a 3-D shape compactness f actor, can be computed. The optimal threshold can be determined by analyzin g these parameters. Our experiments with in vivo nodules demonstrate the fe asibility of employing this algorithm to improve the accuracy of nodule del ineation, especially for small nodules less than 1 cm in diameter. This dis closes the potential of the algorithm for accurate characterizations of nod ules (e.g., volume, change in volume over time) at an early stage, which ca n help to provide valuable guidance for further clinical management. (C) 19 99 Society of Photo-Optical Instrumentation Engineers.