Visualization of Density Variation in Lung Nodules

We propose a method for visualize lung nodule, in order to emphasize their internal structures. This provides an additional tool for diagnosis, which is traditionally based on the analysis of shape, growth and location. The method is divided into three stages: the first stage is semi-automatic segmentation, using the regiongrowing algorithm; the second stage, based on the previous segmentation, identifies the internal structures of the nodule using moment theory to determine several thresholds; finally, in the third stage, to help visualizing these structures, a pseudocolor is associated to each threshold determined in the previous stage. The tool developed is under tests, undergoing clinical evaluation, and its initial results are promising.