ShapeField-lung: continuous shape embedding for early lung cancer detection via pulmonary nodule segmentation - Takeaways - MDSpire

ShapeField-lung: continuous shape embedding for early lung cancer detection via pulmonary nodule segmentation

  • By

  • Xuyu Gu

  • Yifei Zhu

  • Chuangqi Li

  • Xinnan Xu

  • Kaiqi Jin

  • Li Xu

  • November 27, 2025

  • 0 min

Share

  • 1

    ShapeField-Nodule is a continuous shape embedding framework designed for accurate pulmonary nodule segmentation in low-dose CT scans.

  • 2

    The method utilizes a signed distance field to model nodule geometry, achieving sub-voxel precision and anatomically coherent contours.

  • 3

    A shape-aware refinement loss aligns SDF gradients with image edges, enhancing boundary sharpness and anatomical plausibility.

  • 4

    Extensive evaluations on LIDC-IDRI demonstrate that ShapeField-Nodule outperforms traditional voxel-based segmentation methods.

  • 5

    The approach integrates a lightweight MLP with a 3D U-Net backbone, combining dense feature extraction with continuous shape representation.

Original Source(s)

ShapeField-lung: continuous shape embedding for early lung cancer detection via pulmonary nodule segmentation

  • by Xuyu Gu, Yifei Zhu, Chuangqi Li, Xinnan Xu, Kaiqi Jin, Li Xu

  • November 27, 2025

Related Content

Predicting Invasiveness of Lung Adenocarcinoma from Chest CT with Few-shot Vision-Language Ternary Classification Model

Enhancing Access to Lung Cancer Screening with Emphasis on Health Disparities: Insights from Professionals in the UK NHS Lung Cancer Screening Initiative

Machine learning identifies TIME subtypes linking EGFR mutations and immune states in lung adenocarcinoma