A Model-Aided Segmentation in Urethra Identification Based on an Atlas Human Autopsy Image for Intensity Modulated Radiation Therapy
Yan Song, Boris Muller, Chandra M Burman, Borys R. Mychalczak, Yulin Song · Conference proceedings · 2007
In order to protect urethra in radiation therapy of prostate cancer, the urethra must be identified and localized as an organ at risk (OAR) for the inverse treatment planning in intensity modulated radiation therapy (IMRT). Because the prostatic urethra and its surrounding prostate tissue have similar physical characteristics, such as linear attenuation coefficient and density, it is difficult to distinct the OAR from the target in CT images. To localize the urethra without using contrast agent or additional imaging modalities other than planning CT images, a different approach was developed using a standard atlas of human anatomy image. This paper reports an investigation, in which an adult urethra was modeled based on a human anatomic image. An elastic model was build to account for a uniform tissue deformation of the prostate. This model was then applied to patients to localize their urethras and preliminary results are presented.