Search papers, labs, and topics across Lattice.
The ++nnU-Net introduces a novel data augmentation module that leverages image registration to enhance the performance of medical segmentation tasks in data-limited environments. By applying a two-stage registration process to generate new warped images and corresponding segmentation masks, the framework significantly improves the Dice Similarity Coefficient scores compared to the original nnU-Net. Notably, the ++nnU-Net achieves performance gains of up to 22%, underscoring its effectiveness in augmenting limited biomedical datasets while maintaining anatomical fidelity.
Registration-based data augmentation can boost medical segmentation performance by over 20% in data-scarce scenarios.
The nnU-Net has demonstrated continuous success in medical segmentation tasks, which heavily rely on the availability and diversity of annotated biomedical data. However, assembling medical imaging cohorts remains challenging due to numerous factors such as privacy regulations and annotation costs. As a result, data augmentation plays a crucial role in increasing data availability while maintaining anatomical feasibility. Hence, we propose the ++nnU-Net, a novel data augmentation module based on image registration that operates prior to preprocessing and training take place. Our framework was evaluated across five different 2D datasets. In this workflow, image data go through a two-stage registration process, generating new warped images. The transformations are then applied to the respective segmentation. In addition, the pipeline computes available disk space, generates supplementary binary synthetic masks and generates checkpoints. We demonstrate that the ++nnU-Net outperforms the nnU-Net baseline, yielding improvements in Dice Similarity Coefficient scores. In the most prominent cases, we observe performance gains of approximately 22\%. These findings highlight the effectiveness of registration-based data augmentation, particularly for 2D medical imaging datasets and suggest that the ++nnU-Net provides a practical and scalable approach for enhancing segmentation performance in data-limited settings. The source code for the ++nnU-Net is available at: https://github.com/sofia-adelie/plusplusnnunet.git
