Geodesic registration for cervical cancer radiotherapy
Authors: Roy, S., Totman, J.J., Ng, J., Low, J. and Choo, B.A.
Journal: Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume: 9958 LNCS
Pages: 52-69
eISSN: 1611-3349
ISBN: 9783319464718
ISSN: 0302-9743
DOI: 10.1007/978-3-319-46472-5_7
Abstract:Uterus, bladder and rectum are the maximally exposed organs during cervical cancer radiotherapy and are at high risk of radiation exposure. Estimation of dose accumulation in these organs across multiple fractions of external beam radiotherapy (EBRT) and brachytherapy (BT) is extremely challenging due to structural mis-correspondences and complex anatomical deformations between the EBRT and BT images. This paper proposes a unified registration framework that aligns multiple EBRT and BT images of a patient to a single coordinate frame for a cumulative dose assessment. The proposed method transforms the radiotherapy anatomical images to their distance maps from the critical organs (uterus, bladder, and rectum) and registers the distance maps. A Markov random field model is used to fuse the resulting dense deformations and transform the anatomical image. Registration accuracy is evaluated on 42 clinical image pairs and it is shown that the proposed system outperforms existing methods in the literature.
Source: Scopus
Geodesic Registration for Cervical Cancer Radiotherapy
Authors: Roy, S., Totman, J.J., Ng, J., Low, J. and Choo, B.A.
Journal: Workshop on Clinical Image-Based Procedures
Pages: 52-59
Source: Manual