Intra-Patient Anatomic Statistical Models for Adaptive Radiotherapy
Stephen M. Pizer, Robert E. Broadhurst, Gregg S. Tracton, Ja-Yeon Jeong, Rohit R. Saboo, Qiong Han, Joshua V. Stough, Edward L. Chaney · 2006
Abstract — A statistical issue of clinical importance is intrapatient variation from day to day. We use these probability densities for segmentation of daily images by posterior optimization of deformable models. However, the information on intra-patient variation is only available after the multiple days of imaging; yet the densities are needed for segmentation on each day. Still, each patient’s anatomy and image properties are distinct. We describe an approach of using sample means over the days so far to describe a Fréchet mean of the patient. We assume intra-patient variation is stationary across patients, so one can pool training statistics on residues from the mean of the respective patient. The approach is applied both to principal geodesic analysis of m-rep residues describing anatomic variation and to PCA of intensity quantile residues from model-relative regions. In trials to date, application of these statistics in segmentations of male pelvic organs from CT in adaptive radiotherapy yields results competitive with human segmentations and with segmentations based fully on intra-patient statistics. Index Terms — segmentation, adaptive radiotherapy, deformable model segmentation by posterior optimization of deformable models using the combination of within-patient sample means and other-patient statistics of variations. Section 2 summarizes work by others on estimating interpatient probability densities both on anatomic geometry and on intensities. It then describes our work on estimating these densities by Principal Geodesic Analysis (PGA) on m-rep geometric models and by PCA on regional intensity quantile functions (RIQFs), respectively. Section 3 describes the method for estimating intra-patient probability densities. Section 4 describes the application of these densities in the segmentation of intra-patient male pelvic organs from CT and gives results that show that segmentation by posterior optimization based on these probability densities gives results that are not only good but also as good as those given when the probability densities are estimated from all days of the particular patient. I.