Scholarlink

Abstract <p>Accurate segmentation of lung adenocarcinoma nodules in computed tomography (CT) images is critical for clinical staging and diagnosis. However, irregular nodule shapes and ambiguous boundaries pose significant challenges for existing methods. This study introduces S<sup>3</sup>TU-Net, a hybrid CNN-Transformer architecture designed to enhance feature extraction, fusion, and global context modeling. The model integrates three key innovations: (1) structured convolution blocks (DWF-Conv/D<sup>2</sup>BR-Conv) for multi-scale feature extraction and overfitting mitigation; (2) S<sup>2</sup>-MLP Link, a spatial-shift-enhanced skip-connection module to improve multi-level feature fusion; and 3) residual-based superpixel vision transformer (RM-SViT) to capture long-range dependencies efficiently. Evaluated on the LIDC-IDRI dataset, S<sup>3</sup>TU-Net achieves a Dice score of 89.04%, precision of 90.73%, and IoU of 90.70%, outperforming recent methods by 4.52% in Dice. Validation on the EPDB dataset further confirms its generalizability (Dice, 86.40%). This work contributes to bridging the gap between local feature sensitivity and global context awareness by integrating structured convolutions and superpixel-based transformers, offering a robust tool for clinical decision support.</p> Graphical abstract <p></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

S\(^{3}\)TU-Net: Structured convolution and superpixel transformer for lung nodule segmentation

  • 期刊论文

  • 发表日期：

    2025年8月20日

查看全文
  • Yuke Wu,
  • Xiang Liu,
  • Yunyu Shi,
  • Xinyi Chen,
  • Zhenglei Wang,
  • YuQing Xu,
  • ShuoHong Wang

摘要

Abstract

Accurate segmentation of lung adenocarcinoma nodules in computed tomography (CT) images is critical for clinical staging and diagnosis. However, irregular nodule shapes and ambiguous boundaries pose significant challenges for existing methods. This study introduces S3TU-Net, a hybrid CNN-Transformer architecture designed to enhance feature extraction, fusion, and global context modeling. The model integrates three key innovations: (1) structured convolution blocks (DWF-Conv/D2BR-Conv) for multi-scale feature extraction and overfitting mitigation; (2) S2-MLP Link, a spatial-shift-enhanced skip-connection module to improve multi-level feature fusion; and 3) residual-based superpixel vision transformer (RM-SViT) to capture long-range dependencies efficiently. Evaluated on the LIDC-IDRI dataset, S3TU-Net achieves a Dice score of 89.04%, precision of 90.73%, and IoU of 90.70%, outperforming recent methods by 4.52% in Dice. Validation on the EPDB dataset further confirms its generalizability (Dice, 86.40%). This work contributes to bridging the gap between local feature sensitivity and global context awareness by integrating structured convolutions and superpixel-based transformers, offering a robust tool for clinical decision support.

Graphical abstract