美国数学建模竞赛特等奖论文——2003 B O Shelling Tumors with Caution and Wiggles【数学建模】.pdfVIP

Shelling Tumors with Caution and Wiggles 365 Shelling Tumors with Caution and Wiggles Luke Winstrom Sam Coskey Mark Blunk University of Washington Seattle, WA Advisor: James Allen Morrow Abstract We discuss a simple model for tumor growth, which results in realistic tumor shapes. We then address the question of efficient irradiation of tumors in terms of packing their volume with disjoint spheres of given radii. We develop three algorithms, based on a shelling method, which give recommendations for optimal sphere-packings. We interpret these packings as treatment plans, and we examine the corresponding radiation dosages for various qualities. We analyze the effectiveness of our algorithms in recommending optimal dosage plans. Several problems with the algorithms are addressed. Finally, we conclude that all of our algorithms have a common difficulty covering very large tumors, and that while one of our algorithms is effective as a means of packing the volume with spheres, it does not translate into an efficient treatment plan. Introduction The gamma knife is a highly effective means of destroying tumor tissue. A collection of 201 beams, which send ionizing radiation produced by cobalt-60 sources, converge and deliver to a specific region a powerful dose of radiation without harming the surrounding tissue. It is necessary to determine where the beams should converge, so the pa- tient’s brain is scanned with an MRI or CT imaging device to produce a three- dimensional image of the tumor. An optimal treatment plan is determined by analyzing this image and deriving a collection of points inside the tumor. Doses concentrated at those points are administered by the gamma knife. Selecting an optimum treatment plan can be lengthy and difficult. We seek to automate the process of selecting an op