Automatic Segmentation of the Bladder Using Deformable Models

Abstract

We are interested in the fully automatic delineation of the bladder in CT images in the frame of dose calculation for conformational radiotherapy. To this end we fit a series of 3D deformable templates to the contours of anatomical structures. The novelty of our approach resides in the ability to automatically adapt to different kinds of bladder images (homogenous, non-homogenous, contrasted or non-contrasted). The adaptation of the algorithm to inhomogeneities within the bladder improves the accuracy of the segmentation. We validate our approach on a database of tomodensitometric (CT) images of the lower abdomen of male patients.

Automatic Segmentation of Bladder and Prostate Using Coupled 3D Deformable Models

“…This process is illustrated in Figure 2. A more detailed explanation and validation of this method was presented in [24]. 2.…”

A geodesic deformable model for automatic segmentation of image sequences applied to radiation therapy

“…It may be applied for several purposes including segmentation, registration and tracking [1,2]. An essential part of the conformal treatment planning procedure is the segmentation of target volumes and organs at risk in CT images [3]. Because of the difficulty of accurately and reliably delineating structures in medical images, this work has been typically done manually by radiation oncologists [4,5].…”

“…Because of the difficulty of accurately and reliably delineating structures in medical images, this work has been typically done manually by radiation oncologists [4,5]. However, manual delineation is tedious, time-consuming and prone to error due to intra-and inter-user variability [3,6].…”

“…The method is applied to prostate cancer ROIs, that is, bladder and rectum. The bladder and rectum are considered the organs at risk that should be protected against high doses of radiation during the treatment of prostate cancer [3,5,14]. The application of a geodesic model for ROI tracking from a 2D CT image set of the pelvic area has not been considered before.…”

“…A good survey has been done by Shi et al [15]. These methods are based on mathematical morphology [16,17], region growing [14,17,18], shape deformation, including geometric [19,20] and parametric models [21] constrained by means of simple mesh [3] and atlas-based [22], initialization. All these approaches have their advantages when compared to manual delineation, but there are also some drawbacks due to the large variations in bladder geometry among patients and the limited contrast between the bladder and nearby organs.…”