CT,PET,Diffusion-MR画像を用いた消化器癌転移陽性リンパ節の特定

概要

論文の詳細を見る
The quantitative assessment of abdominal lymph nodes in the context of malign tumors requires an automatic detection technique in multi-modality datasets. The region of interest was set with multi-modality datasets. We performed template matching to do rough extraction with a spherical teacher image. The quantity of shape characteristic of detected object was estimated precisely. Our method was evaluated with lymph nodes from clinical multi-modality datasets and could be proved to be comparable to the manual detection results of experienced users. This detection technique is helpful when a doctor examines a lymphadenectomy range by the operation.
近畿大学工業高等専門学校の論文