Accurate prostate segmentation on MR images plays an important role in the management of prostate diseases. Recently proposed deep learning architecture has been successfully applied for medical image segmentation to overcome the shortcomings of manual segmentation. Our study proposed a 3D UNet model for automatic and accurate prostate gland segmentation on both DWI and T2WI images. This model was tested in 3 different external cohorts and showed satisfactory results on T2WI images. The segmentation performance on DWI images was inferior but still inspiring in the external testing group. This study might benefit the management of prostate diseases in the future.
This abstract and the presentation materials are available to members only; a login is required.