Please use this identifier to cite or link to this item:
Title: Integrating Residual, Dense, and Inception Blocks into the nnUNet
Authors: McConnell, N
Li, Y
Keywords: nnUnet;Biomedical image segmentation;Residual networks;Dense networks;Inception networks
Issue Date: 1-Jun-2022
Publisher: IEEE
Citation: McConnell, N. and Li, Y. (2022) 'Integrating Residual, Dense, and Inception Blocks into the nnUNet'.CBMS 2022: IEEE 35th International Symposium on Computer Based Medical Systems.
Abstract: The nnUNet is a fully automated and generalisable framework which automatically configures the full training pipeline for the segmentation task it is applied on, while taking into account dataset properties and hardware constraints. It utilises a basic UNet type architecture which is self-configuring in terms of topology. In this work, we propose to extend the nnUNet by integrating mechanisms from more advanced UNet variations such as the residual, dense, and inception blocks, resulting in three new nnUNet variations, namely the Residual nnUNet, Dense-nnUNet, and Inception-nnUNet. We have evaluated the segmentation performance on eight datasets consisting of 20 target anatomical structures. Our results demonstrate that altering network architecture may lead to performance gains, but the extent of gains and the optimally chosen nnUNet variation is dataset dependent.
Appears in Collections:Dept of Computer Science Research Papers

Files in This Item:
File Description SizeFormat 
FullText.pdf2.62 MBAdobe PDFView/Open

Items in BURA are protected by copyright, with all rights reserved, unless otherwise indicated.