Depth Swin Transformer Unet for Serial Section Biomedical Image Segmentation

Lin, Jun; Ge, Hongyu; Wang, Ying; Zhang, Ruobing

doi:10.1109/icsip55141.2022.9886148

Cited by 1 publication

(2 citation statements)

References 19 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…It has proven its effectiveness in handling complex image segmentation tasks. It is essential to highlight that our framework is highly flexible, enabling designers to freely choose different backbones, such as Res-Unet3D (Li et al 2022) and TransBTS (Lin et al 2022), among others. This adaptability further enhances the versatility and applicability of our approach.…”

Section: Backbone Networkmentioning

confidence: 99%

“…, HIVE-Net(Yuan et al 2021), Zhili(Li et al 2021), Peng and Yuan(Peng and Yuan 2019), Xiao(Xiao et al 2018), Res-Unet3D(Li et al 2022), HIVE-Net(Yuan et al 2021), and nnU-Net(Isensee et al 2021). For Transformer-based methods, we choose DSTUnet(Lin et al 2022), TransBTS(Wang et al 2021), UNETR(Hatamizadeh et al 2022b), nnFormer(Zhou et al 2021), SwinUNETR(Hatamizadeh et al 2022a), and 3D UX-Net(Lee et al 2023). The λ 1 and λ 2 in the loss function are set to be 1 and 0.5 following(Zou et al 2022).…”

mentioning

confidence: 99%

See 1 more Smart Citation

Evidential Uncertainty-Guided Mitochondria Segmentation for 3D EM Images

Shi,

Duan,

Huang

et al. 2024

AAAI

View full text Add to dashboard Cite

Recent advances in deep learning have greatly improved the segmentation of mitochondria from Electron Microscopy (EM) images. However, suffering from variations in mitochondrial morphology, imaging conditions, and image noise, existing methods still exhibit high uncertainty in their predictions. Moreover, in view of our findings, predictions with high levels of uncertainty are often accompanied by inaccuracies such as ambiguous boundaries and amount of false positive segments. To deal with the above problems, we propose a novel approach for mitochondria segmentation in 3D EM images that leverages evidential uncertainty estimation, which for the first time integrates evidential uncertainty to enhance the performance of segmentation. To be more specific, our proposed method not only provides accurate segmentation results, but also estimates associated uncertainty. Then, the estimated uncertainty is used to help improve the segmentation performance by an uncertainty rectification module, which leverages uncertainty maps and multi-scale information to refine the segmentation. Extensive experiments conducted on four challenging benchmarks demonstrate the superiority of our proposed method over existing approaches.

show abstract

Section: Backbone Networkmentioning

confidence: 99%