Abstract: Abdominal multi-organ segmentation in computed tomography (CT) is crucial for many clinical applications including disease detection and treatment planning. Deep learning methods have shown unprecedented performance in this perspective. However, it is still quite challenging to accurately segment different organs utilizing a single network due to the vague boundaries of organs, the complex background, and the substantially different organ size scales. In this work, we investigated the feasibility of applying the famous nnU-Net to performing abdominal multi-organ segmentation in CT. By slightly modifying the configurations of nnU-Net, we obtained promising segmentation results. Specifically, quantitative evaluations on the FLARE2022 validation cases (20 cases) show that the method achieves an average Dice similarity coefficient (DSC) of 0.71 and average normalized surface distance (NSD) of 0.76. With further optimization, it is possible to obtain satisfactory segmentation results.
1 Reply
Loading