Detection of rounded opacities on chest radiographs usingconvergence index filter

摘要

This paper presents a method to detect rounded opacities ondigital chest radiographs. Based on a model of rounded opacities, threefilters that evaluate the convergence degree of gradient vectors in theneighborhood of the pixel of interest are investigated to detect roundedopacities in the complex lung area. These filters do not depend on thecontrast of the rounded opacity to its background. Their outputs dependon only the distribution of the gradient vector orientation. Experimentsto test their performance have been performed. Their results show theeffectiveness of the proposed filters. Especially, the iris filter andthe adaptive ring filter are shown to be very effective in detectingcancerous tumor candidates. Another experiment to compare theperformance of the iris filter and the Min-ΦDD filter which is oneof the filters effective in detecting rounded opacities have beenperformed and the superiority of the proposed filters has been shown