A morphological filter with region expansion to extract minute convex regions and its applications to medical image processing
Yoshihiro Hagihara, Hidefumi Kobatake · Systems and Computers in Japan · 2001
Abstract In recent years morphological filters (M‐filters) have been used in a wide range of fields since they offer various capabilities through the use of the shape of structural elements in the image. In this paper the authors propose an M‐filter which can minimize the effect of the background for the purpose of extracting undefined minute convex regions such as lesions in medical X‐ray images. The proposed method is based on expanding the region by padding for each brightness level. In addition, the authors consider an approach to implement their proposed method effectively, then describe the resulting algorithm in detail. Experiments comparing the proposed method and a method using only conventional opening calculations are performed. In experiments using binary images and multibit simulation data and in experiments on the extraction of minute calcification images in a mammogram or kidney–lung shadows from real chest CTs, the proposed method was shown to be superior to the conventional method using opening calculations. © 2001 Scripta Technica, Syst Comp Jpn, 32(9): 22–31, 2001