Precision Engineering, Tokyo University, Japan
In order to plan the least invasive surgical approach, it is demanded to grasp the boundary between the surgical site and normal tissue from biomedical images obtained in advance. For the purpose, active contour so-called snake is effective algorithm because it always enables us to obtain closed contour and it is a noise robust segmentation algorithm. However, there remained some problems to be solved for practical use. The goal of this project was to optimize an ImageJ plug-in about active contour for coronary artery in ultrasonography. The most significant problem was self-intersection. The self-intersection check which can saves time effectively with sweep line algorithm was added to the snake plug-in previously developed. The movie of ultrasonography was reconstructed as 3D model highlighting coronary artery segmented by the plug-in.