Scholarlink

<p>Colposcopy is an essential tool for cervical precancer evaluation and biopsy guidance. Most existing artificial intelligence (AI) tools for cervicogram analysis are lesion-centric or limited to global transformation zone (TZ) classification and therefore do not explicitly delineate TZ-related anatomical landmarks. Because the TZ is a major site of cervical carcinogenesis and is defined by the original squamocolumnar junction (SCJ), the new SCJ, and the external cervical os, landmark-level TZ segmentation may provide clinically meaningful support for colposcopic interpretation. To address this need, we reformulated TZ-related anatomical segmentation as a landmark-driven, four-class semantic segmentation task. We propose the Clinically Modulated Boundary-aware Network (CMB-Net), which integrates patient-specific clinical variables to account for heterogeneous SCJ visibility and incorporates multi-scale boundary supervision to improve the delineation of ambiguous anatomical interfaces. Applied to 889 internal cervicograms, CMB-Net achieved an mDice of <InlineEquation ID="IEq1"> <EquationSource Format="TEX">\(87.37 \pm 0.17\)</EquationSource> <EquationSource Format="MATHML"><math> <mrow> <mn>87.37</mn> <mo>±</mo> <mn>0.17</mn> </mrow> </math></EquationSource> </InlineEquation>% and an mIoU of <InlineEquation ID="IEq2"> <EquationSource Format="TEX">\(78.39 \pm 0.24\)</EquationSource> <EquationSource Format="MATHML"><math> <mrow> <mn>78.39</mn> <mo>±</mo> <mn>0.24</mn> </mrow> </math></EquationSource> </InlineEquation>%. In two independent external cohorts comprising 310 cases from two additional centers, it achieved an mDice/mIoU of 80.95%/69.28%, outperforming CNN- and transformer-based baselines. In the same 310-case external test cohort used for observer comparison, CMB-Net exceeded junior colposcopist-reference agreement (69.47%) and was comparable to senior colposcopist-reference agreement (80.85%). These results suggest that combining patient-conditioned priors with boundary supervision can support robust and interpretable TZ-related anatomical segmentation.</p>

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

CMB-Net: A Clinically Modulated Boundary-Aware Network for Anatomical Segmentation of the Cervical Transformation Zone in Colposcopy

  • 期刊论文

  • 发表日期：

    2026年7月29日

查看全文
  • Ling Yan,
  • Jiali Wu,
  • Yi Guo,
  • Peng Ren,
  • Jingjing Yang,
  • Xingfa Shen,
  • Ying Li,
  • Li Ding,
  • Xudong Ma,
  • Shan Jiang

摘要

Colposcopy is an essential tool for cervical precancer evaluation and biopsy guidance. Most existing artificial intelligence (AI) tools for cervicogram analysis are lesion-centric or limited to global transformation zone (TZ) classification and therefore do not explicitly delineate TZ-related anatomical landmarks. Because the TZ is a major site of cervical carcinogenesis and is defined by the original squamocolumnar junction (SCJ), the new SCJ, and the external cervical os, landmark-level TZ segmentation may provide clinically meaningful support for colposcopic interpretation. To address this need, we reformulated TZ-related anatomical segmentation as a landmark-driven, four-class semantic segmentation task. We propose the Clinically Modulated Boundary-aware Network (CMB-Net), which integrates patient-specific clinical variables to account for heterogeneous SCJ visibility and incorporates multi-scale boundary supervision to improve the delineation of ambiguous anatomical interfaces. Applied to 889 internal cervicograms, CMB-Net achieved an mDice of \(87.37 \pm 0.17\) 87.37 ± 0.17 % and an mIoU of \(78.39 \pm 0.24\) 78.39 ± 0.24 %. In two independent external cohorts comprising 310 cases from two additional centers, it achieved an mDice/mIoU of 80.95%/69.28%, outperforming CNN- and transformer-based baselines. In the same 310-case external test cohort used for observer comparison, CMB-Net exceeded junior colposcopist-reference agreement (69.47%) and was comparable to senior colposcopist-reference agreement (80.85%). These results suggest that combining patient-conditioned priors with boundary supervision can support robust and interpretable TZ-related anatomical segmentation.