EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Text-guided few-shot liver and tumor segmentation

Hongling Chen, Aibing Xu, Li Zhang, Miao Ouyang, Zhonghang Zhu, Jiacheng Wang

Frontiers in Digital Health · 29.05.2026 · 8

DOI: 10.3389/fdgth.2026.1823870 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Frontiers in Digital Health, online erschienen am 29.05.2026. Verfasst von Hongling Chen und 5 weiteren Autor:innen. Beteiligte Einrichtung: Nantong Tumor Hospital/Nantong University Affiliated Tumor Hospital, Nantong, Jiangsu, China. Die Arbeit nennt 52 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Introduction

High-precision liver and tumor segmentation is a cornerstone of digital oncology, yet its clinical deployment remains constrained by two persistent challenges: the scarcity of pixel-level annotations and severe performance degradation under cross-center domain shift. Although few shot learning offers a promising direction for data-efficient modeling, existing approaches relying solely on visual similarity often fail to generalize across heterogeneous clinical environments.

Method

In this study, we propose a text-guided few-shot segmentation framework that integrates clinical semantic information as an explicit inductive bias to bridge the gap between low-level pixels and high-level diagnostic reasoning. The framework consists of three synergistic modules: an Automated Semantic Generator that employs a large-scale vision-language model to encode clinically structured semantic descriptions into structured linguistic priors. A Text-Guided Gating (TGG) mechanism that adaptively modulates visual representations to filter scanner-dependent artifacts, and a Decoupled Prototype Learner that constructs unbiased class prototypes via per-image averaging and gradient detachment to address extreme class imbalance. Together, these components transform segmentation from a pure visual matching task into a knowledge guided reasoning process, ensuring both data efficiency and improved cross-dataset robustness across heterogeneous datasets.

Results

Experiments on LiTS and 3DIRCADb show that the method consistently outperforms state-of-the-art supervised, foundation-model-based, and few-shot baselines. On 3DIRCADb, compared to PANet, the strongest few-shot baseline, our method improves external liver Dice by 8.7 percentage points and external tumor Dice by 26.3 percentage points, effectively mitigating performance degradation in conventional supervised models.

Discussion

These results demonstrate that cross-modal semantic guidance enhances robust medical image segmentation under domain shift.

Schlagwörter

Stichwörter der Autor:innen

automated semantic guidancecross-dataset generalizationdigital oncologyfew-shot learningliver tumor segmentationvision-language models

Autor:innen und Institutionen

  • Nantong Tumor Hospital/Nantong University Affiliated Tumor Hospital, Nantong, Jiangsu, China.
  • School of Electronic Information, Wuhan University of Science and Technology, Wuhan, China.
  • Manteia Technology, Co., Ltd., Xiamen, Fujian, China.

Zitieren

APA

Chen, H., Xu, A., Zhang, L., Ouyang, M., Zhu, Z., & Wang, J. (2026). Text-guided few-shot liver and tumor segmentation. Frontiers in Digital Health, 8. https://doi.org/10.3389/fdgth.2026.1823870

Vancouver

Chen H, Xu A, Zhang L, Ouyang M, Zhu Z, Wang J. Text-guided few-shot liver and tumor segmentation. Frontiers in Digital Health. 2026;8. doi:10.3389/fdgth.2026.1823870

BibTeX
@article{Chen2026,
  author = {Chen, Hongling and Xu, Aibing and Zhang, Li and Ouyang, Miao and Zhu, Zhonghang and Wang, Jiacheng},
  title = {Text-guided few-shot liver and tumor segmentation},
  journal = {Frontiers in Digital Health},
  year = {2026},
  volume = {8},
  doi = {10.3389/fdgth.2026.1823870}
}
Kostenlos starten →