Premium
SU‐E‐T‐327: Tracking Palladium Seeds Implanted in Prostate Using Shape Analysis Approach
Author(s) -
Zhai Y,
Jiang Y,
Papademetris X,
Roberts K,
Nath R,
Chen Z
Publication year - 2012
Publication title -
medical physics
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.473
H-Index - 180
eISSN - 2473-4209
pISSN - 0094-2405
DOI - 10.1118/1.4735414
Subject(s) - prostate , prostate cancer , consistency (knowledge bases) , tracking (education) , artificial intelligence , computer science , computer vision , deformation (meteorology) , medicine , nuclear medicine , materials science , cancer , psychology , pedagogy , composite material
Purpose: To develop and test a shape‐based image registration algorithm to register and analyze serial CT image datasets of a given patient, and then to track the motions of the implanted seeds for a quantitative characterization of prostate deformation and edema. Methods: Sequential 3D CT image datasets for a prostate cancer patient, acquired before palladium source implantation and after implantation, were used to perform deformable image registrations with shape‐based algorithm to track the sources motions. The seed centers were detected in a group‐wise fashion after extracting the seed surfaces, i.e. using consistency across the sequence of images acquired from the same subject as an extra constraint, to maximize the reliability of seed detections. Results: The shape analysis approach was applied to the series of CT scans at day 0, 1, and 30 for one prostate cancer patient. The motion of seeds was visualized in plots, from which the non‐rigid deformation of prostate was interpolated. It was observed that the prostate volume was shrinking from Day 0 to Day 30, while it expanded from Day 0 to Day 1. Our data also showed that the prostate deformation caused by the resolution of prostate edema was anisotropic. Conclusions: A shape analysis approach was developed for quantitative characterization of post‐implant seed tracking and the prostate deformation. The advantage over other methods is that the constraints of consistency among different images made our method more reliable than other seed localization methods that estimate the seed centers for each image individually. The precise localization of implanted seeds as function of time would enable a more accurate calculation and proper summation of radiation dose delivered to each tumor sub‐volume in post‐implant dosimetry. This research project is currently funded by the NIH grant R01CA134627‐01 “Prostate Edema in Permanent Interstitial Brachytherapyâ€.