LesionMix: A Lesion-Level Data Augmentation Method for Medical Image Segmentation
Basaran BD, Zhang W, Qiao M, Kainz B, Matthews PM, Bai W (2024)
Publication Type: Conference contribution
Publication year: 2024
Journal
Publisher: Springer Science and Business Media Deutschland GmbH
Book Volume: 14379 LNCS
Pages Range: 73-83
Conference Proceedings Title: Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Event location: Vancouver, BC, CAN
ISBN: 9783031581700
DOI: 10.1007/978-3-031-58171-7_8
Abstract
Data augmentation has become a de facto component of deep learning-based medical image segmentation methods. Most data augmentation techniques used in medical imaging focus on spatial and intensity transformations to improve the diversity of training images. They are often designed at the image level, augmenting the full image, and do not pay attention to specific abnormalities within the image. Here, we present LesionMix, a novel and simple lesion-aware data augmentation method. It performs augmentation at the lesion level, increasing the diversity of lesion shape, location, intensity and load distribution, and allowing both lesion populating and inpainting. Experiments on different modalities and different lesion datasets, including four brain MR lesion datasets and one liver CT lesion dataset, demonstrate that LesionMix achieves promising performance in lesion image segmentation, outperforming several recent Mix-based data augmentation methods. The code will be released at https://github.com/dogabasaran/lesionmix.
Authors with CRIS profile
Involved external institutions
United Kingdom (GB)How to cite
APA:
Basaran, B.D., Zhang, W., Qiao, M., Kainz, B., Matthews, P.M., & Bai, W. (2024). LesionMix: A Lesion-Level Data Augmentation Method for Medical Image Segmentation. In Yuan Xue, Chen Chen, Chao Chen, Lianrui Zuo, Yihao Liu (Eds.), Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (pp. 73-83). Vancouver, BC, CAN: Springer Science and Business Media Deutschland GmbH.
MLA:
Basaran, Berke Doga, et al. "LesionMix: A Lesion-Level Data Augmentation Method for Medical Image Segmentation." Proceedings of the 3rd International Workshop on Data Augmentation, Labeling, and Imperfections, DALI 2023 in conjunction with the 26th International Conference on Medical Image Computing and Computer Assisted Intervention, MICCAI 2023, Vancouver, BC, CAN Ed. Yuan Xue, Chen Chen, Chao Chen, Lianrui Zuo, Yihao Liu, Springer Science and Business Media Deutschland GmbH, 2024. 73-83.
BibTeX: Download