Entropy-Based, Multiple-Portal-to-3DCT Registration for Prostate Radiotherapy Using Iteratively Estimated Segmentation

In external beam radiotherapy (EBRT), patient setup verification over the entire course of fractionated treatment is necessary for accurate delivery of specified dose to the tumor. We develop an information theoretic minimax entropy registration framework for patient setup verification using portal images and the treatment planning 3D CT data set. Within this framework we propose to simultaneously and iteratively segment the portal images and register them to the 3D CT data set to achieve robust and accurate estimation of the pose parameters. Appropriate entropies are evaluated, in an iterative fashion, to segment the portal images and to find the registration parameters. Earlier, we reported our work using a single portal image to estimate the transformation parameters. In this work, we extend the algorithm to utilize dual portal images. In addition, we show the performance of the algorithm on real patient data, analyze the performance of the algorithm under different initializations and noise conditions, and note the wide range of parameters that can be estimated. We also present a coordinate descent interpretation of the proposed algorithm to further clarify the formulation.