Improved bolus shaping accuracy using the surface segmentation and spectral clustering. (4th July 2021)

• Record Type: Journal Article
• Authors: Li, Rui; Peng, Qingjin; Ingleby, Harry; Sasaki, David
• Is Part Of: International journal of modelling & simulation. Volume 41:Number 4(2021)
• Page Start: 299
• Page End: 310
• Publication Date: 2021-07-04
• Subjects: Surface segmentation -- clustering -- surface unfolding -- bolus; Mathematical models -- Periodicals; Simulation methods -- Periodicals; Mathematical models; Simulation methods; Periodicals; 003.3
• DOI: 10.1080/02286203.2020.1756036
• Languages: English
• ISSNs: 0228-6203

Abstract

ABSTRACT: The bolus is used to cover tumors in the shallow skin of cancer patients for a desired dose distribution in the high-energy radiotherapy. This research develops methods to improve the accuracy of bolus shaping by unfolding the human surface model for refolding. A surface segmentation method is proposed to separate a non-developable surface into a few pieces based on a clustering method to increase flattenability. In the segmentation process, a 3D surface model is embedded into a spectral space using a Laplacian matrix to combine the surface segmentation saliency and geometric characteristics. A smooth common boundary is designed based on clustering results to simplify the shape of each part. The 3D surface model is unfolded into 2D patches by the coordinate transformation and optimized using a mass-spring model. Case studies are conducted to verify the effectiveness of the proposed method in the bolus shaping.