Geometrical analysis and predictive modeling of head and neck tumors

Most of the current radiation treatment planning systems uses pre-treatment CT images to detect the tumor location and then plan the radiation therapy is delivered during the treatment period. They assume that the tumor geometry will not change throughout the treatment course; however, tumor geometry is shown to be changing over time. This article presents results of an ongoing research in 3D modeling and reconstruction of head and neck cancer tumors. The results from this phase of the project will be used in developing a prediction model for tumor deformation during radiation treatment of cancer patients. By using CT scan data in the 3D ASCII format, the tumor´s progressive geometric changes during the treatment period are quantified. After constructing slice contours, both triangular and rectangular patch approaches are applied to map and analyze the tumor surface and volume. The changes in tumor location are calculated based on a reference feature on the top of the spine canal. MATLAB routines are developed to perform the required calculations. A set of prototype mockups of different stages are used for the purpose of validation and verification of the proposed methodology. The proposed method is applied to calculate volume, surface, and displacement of the tumor, using patients´ data obtained from the University of Texas- MD Anderson Cancer Center. The results are consistent with the actual data.