3次元適応的 CP 写像を用いた脳組織の自動抽出

概要

論文の詳細を見る
In medical image processing, it is desirable to extract interested regions from image signals for reconstructing 3-dimensional structures. For this purpose, a novel algorithm called the 3-dimensional Clumsy Painter method (3D-CP) were proposed. Although we have already applied it to 3-dimensional gray scale images without binarization, there still exists a problem which is impossible to adapt the shape of clumsy painter templates to the variation of intensity among different brain tissues. This problem was caused by painting only by a single cp template for any positions in an image. For two-dimensional MRI, we solved the problem by considering the upper and lower movement conditions with graylevel information of three MR images, T1-weighted(Tl), T2-weighted(T2) and Proton density-weighted(PD)images. In this paper, we propose the method to adapt shapes of maps to a local change of brain contour regions by deciding a shape of maps from the graylevel information of three-dimensional MR images.
社団法人電子情報通信学会の論文
1997-08-13