Title :
Diffeomorphic surface-based registration for MR-US fusion in prostate brachytherapy
Author_Institution :
ICTEAM Inst., Univ. Catholique de Louvain, Louvain-La-Neuve, Belgium
Abstract :
In nowadays prostate interstitial brachytherapy, transrectal ultrasound guidance is typically used to localize the prostate and the rectum. However, the ability of ultrasound to distinct normal from cancer tissues is weak. In most treatments, the organ as a whole is thus irradiated, leading to numerous side effects such as urinary or sexual dysfunction. The quality of magnetic resonance images (MRI), on the other hand, has improved over the last few years and allows for an accurate delineation of the tumor [1]. This paper proposes a novel framework for MRI-US surface registration of the prostate and the rectum. The preoperative MRI is first segmented using a multiresolution graph-cut method. The intraoperative US image is manually segmented by the surgeon. The registration is then performed in two steps. In a first step, a bi-affine registration is performed on the surfaces of the rectum and the prostate using an expectation maximization iterative closest point method (EM-ICP). In the second step, non-rigid registration is applied to the distance maps resulting from the pre-registered surfaces. Our approach has been applied on 5 MR/US pairs and shows a relative independence between prostate and rectum motions.
Keywords :
biological organs; biomedical MRI; biomedical ultrasonics; brachytherapy; cancer; expectation-maximisation algorithm; image fusion; image registration; image resolution; image segmentation; medical image processing; tumours; ultrasonic imaging; MR-US fusion; MRI; cancer tissues; diffeomorphic surface-based registration; expectation maximization iterative closest point method; image segmentation; intraoperative US image; magnetic resonance images; multiresolution graph-cut method; organ; prostate interstitial brachytherapy; rectum; sexual dysfunction; surgeon; transrectal ultrasound guidance; tumor; urinary dysfunction; Biomedical imaging; Brachytherapy; Image edge detection; Image resolution; Image segmentation; Magnetic resonance imaging; Ultrasonic imaging;
Conference_Titel :
Electrotechnical Conference (MELECON), 2012 16th IEEE Mediterranean
Conference_Location :
Yasmine Hammamet
Print_ISBN :
978-1-4673-0782-6
DOI :
10.1109/MELCON.2012.6196575