肝臓分割と血管抽出による肝臓情報の三次元可視化と手術支援システム

摘要

Computer Assisted Diagnosis/Surgery systems provide an indisputable effect on the diagnosis and surgery of hepatic diseases. In this paper, we propose a novel simulation system for liver surgical planning, which combined image processing and computer vision techniques. The proposed system is composed of three modules, liver segmentation, vessel extraction, and visualization / interaction. We first segment liver region from CT volume by using K-Means Clustering and Geodesic Active Contour Algorithms. And then we extract vessels by using multiscale filters. The third visualization / interaction module visualizes them, and prepares the user a virtual environment to perform surgical operation on the liver image. The system has potentials to be used for treatment/surgical planning, and even as a guided surgery system.%肝臓は個人においてもその大きさ，形状及び脈管分布においてバリエーションが広く，各患者の肝臓情報に応じた手術支援は重要な課題である.本研究は各患者のCT画像からクラスタリング法による肝臓分割とFrangi法に基づいた血管情報抽出により詳細な患者肝臓情報の三次元可視化を行う.得られた情報を基に，計算機上で疑似手術環境を設定し，その擬似手術環境上で執刀医が仮想メスを用いて対話的に肝臓内侵聾しながら直観的に侵襲状況をディスプレイ上で確認を行うことが可能な手術支援システムを構築した.