Scholarlink

Accurate segmentation of anatomical structures from cone-beam computed tomography (CBCT) is essential for clinical applications in dentistry, maxillofacial surgery, and orthodontics. The ToothFairy3 Challenge has a comprehensive 77-class segmentation task, emphasizing both accuracy and computational efficiency. In this work, we present a method based on the nnU-Net framework, enhanced with a Structure Aware Post-processing (SAP) strategy. nnU-Net serves as a backbone for multi-class segmentation, while SAP refines predictions by introducing individualized thresholds for each anatomical structure, thereby mitigating noise and preserving clinically important fine structures. To further improve efficiency, we disabled mirroring augmentation during training and employed inference acceleration strategies, including the removal of test-time augmentation and optimized interpolation on floating-point tensors. Experimental results validate the effectiveness of our approach in balancing segmentation accuracy with computational efficiency. To further ensure robustness in challenging clinical scenarios, we also utilize an interactive refinement module based on nnInteractive. This strategy allows clinicians to correct local segmentation errors with minimal user guidance, providing a safety net for complex anatomical variations.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Efficient CBCT Segmentation via nnU-Net with Structure-Aware Post-processing and Interactive Refinement

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • Changkai Ji,
  • Yusheng Liu,
  • Yuxian Jiang,
  • Lisheng Wang

摘要

Accurate segmentation of anatomical structures from cone-beam computed tomography (CBCT) is essential for clinical applications in dentistry, maxillofacial surgery, and orthodontics. The ToothFairy3 Challenge has a comprehensive 77-class segmentation task, emphasizing both accuracy and computational efficiency. In this work, we present a method based on the nnU-Net framework, enhanced with a Structure Aware Post-processing (SAP) strategy. nnU-Net serves as a backbone for multi-class segmentation, while SAP refines predictions by introducing individualized thresholds for each anatomical structure, thereby mitigating noise and preserving clinically important fine structures. To further improve efficiency, we disabled mirroring augmentation during training and employed inference acceleration strategies, including the removal of test-time augmentation and optimized interpolation on floating-point tensors. Experimental results validate the effectiveness of our approach in balancing segmentation accuracy with computational efficiency. To further ensure robustness in challenging clinical scenarios, we also utilize an interactive refinement module based on nnInteractive. This strategy allows clinicians to correct local segmentation errors with minimal user guidance, providing a safety net for complex anatomical variations.