Elastic image mapping for 4-D dose estimation in thoracic radiotherapy

Abstract

Demonstrate the path integration of a four-dimensional (4-D) dose distribution onto the 3-D anatomy. A computer-generated 4-D thoracic phantom with a lung tumour was constructed. Eight respiratory phases were generated. A radiotherapy treatment plan was applied to all the phases resulting in a 4-D dose distribution. An elastic image registration algorithm was used to find the vector displacement between all the image elements and the end expiration phase. The path-integrated tissue dose distribution and each component dose distribution were compared with the planned dose distribution. Numerical path integration was performed to calculate the tissue dose distribution. Loss of tumour coverage was the predominant effect observed with tumour motion in this study. The loss was asymmetric and dependent on the tumour trajectory. The elastic image registration allowed an accurate path integration through a 4-D data set to produce an accurate 3-D tissue dose estimate.