Segmentation of the prostate in 3D CT images is a crucial step in treatment planning and procedure guidance such asbrachytherapy and radiotherapy. However, manual segmentation of the prostate is very time-consuming and depends onthe experience of the clinician. On the contrary, automated prostate segmentation is more helpful in practice, whereas thetask is very challenging due to low soft-tissue contrast in CT images. In this paper, we propose a 3D deeply supervisedfully-convolutional-network (FCN) with dilated convolution kernel to automatically segment prostate in CT images. Adeep supervision strategy could acquire more powerful discriminative capability and accelerate the optimizationconvergence in training stage, while concatenating the dilated convolution enlarges the receptive field to extract moreglobal contextual information for accurate prostate segmentation. The presented method was evaluated using 15 prostateCT images and obtained a mean Dice similarity coefficient (DSC) of 0.85±0.04 and mean surface distance (MSD) of1.92±0.46 mm. The experimental results show that our approach yields accurate CT prostate segmentation, which can beemployed for the prostate-cancer treatment planning of brachytherapy and external beam radiotherapy.
展开▼
