Adaptive composite loss for volumetric whole heart segmentation
Issued Date
2025-12-01
Resource Type
eISSN
20452322
Scopus ID
2-s2.0-105022851261
Journal Title
Scientific Reports
Volume
15
Issue
1
Rights Holder(s)
SCOPUS
Bibliographic Citation
Scientific Reports Vol.15 No.1 (2025)
Suggested Citation
Sutassananon K., Kusakunniran W., Orgun M., Siriapisith T. Adaptive composite loss for volumetric whole heart segmentation. Scientific Reports Vol.15 No.1 (2025). doi:10.1038/s41598-025-25785-9 Retrieved from: https://repository.li.mahidol.ac.th/handle/123456789/113355
Title
Adaptive composite loss for volumetric whole heart segmentation
Author(s)
Author's Affiliation
Corresponding Author(s)
Other Contributor(s)
Abstract
Accurate segmentation in medical imaging requires loss functions that capture both regional overlap and boundary alignment. This study evaluates composite losses combining binary cross-entropy (BCE) and a boundary-based term under fixed and adaptive weighting schemes, using U-Net and SwinUNETR on the MM-WHS dataset. For U-Net, a small boundary contribution with adaptive weighting yielded the best results: Standard SoftAdapt (90/10 BCE + BoundaryDoU) achieved the highest Dice score (), surpassing both the baseline () and fixed ratios. In contrast, SwinUNETR achieved its strongest performance with a fixed 70% BCE + 10% boundary ratio (0.919 ± 0.02). The result showed that combining a boundary-based loss term helps improve the segmentation accuracy. However, the performance gain is dependent on the architecture of the segmentation model; convolution-based U-Net benefited from the adaptive loss weighting scheme, whereas Transformer-based SwinUNETR without strong inductive bias did not benefit from increased influence of the boundary loss term.