EUVIMEDEUVIMED entdecken
Crossref

BiASAM: Bidirectional-Attention Guided Segment Anything Model for Very Few-Shot Medical Image Segmentation

Wei Zhou, Guilin Guan, Wei Cui, Yugen Yi

IEEE Signal Processing Letters · 2025 · 32 · 246-250

DOI: 10.1109/lsp.2024.3513240 ↗

Kurz gefasst

Fachartikel in der Zeitschrift IEEE Signal Processing Letters, 2025. Verfasst von Wei Zhou und 3 weiteren Autor:innen. Die Arbeit wurde bisher 12-mal zitiert (Crossref) und nennt 29 Literaturangaben. Gefördert von: National Natural Science Foundation of China, Natural Science Foundation of Liaoning Province, Outstanding Youth Project of Jiangxi Natural Science Foundation.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

The Segment Anything Model (SAM) excels in general segmentation but encounters difficulties in medical imaging due to few-shot learning challenges, particularly with extremely limited annotated data. Existing approaches often suffer from insufficient feature extraction and inadequate loss function balancing, resulting in decreased accuracy and poor generalization. To address these issues, we propose BiASAM, which uniquely incorporates two bidirectional attention mechanisms into SAM for medical image segmentation. Firstly, BiASAM integrates a spatial-frequency attention module to improve feature extraction, enhancing the model's ability to capture both fine and coarse details. Secondly, we employ an attention-based gradient update mechanism that dynamically adjusts loss weights, boosting the model's learning efficiency and adaptability in data-scarce scenarios. Additionally, BiASAM utilizes the point and box fusion prompt to enhance segmentation precision at both global and local levels. Experiments across various medical datasets show BiASAM achieves performance comparable to fully supervised methods with just two labeled samples.

Förderung

  • National Natural Science Foundation of China
  • Natural Science Foundation of Liaoning Province
  • Outstanding Youth Project of Jiangxi Natural Science Foundation
  • Jiangxi Province Key Subject Academic and Technical Leader Funding Project
  • Fundamental Research Funds for the Universities of Liaoning Province

Zitieren

APA

Zhou, W., Guan, G., Cui, W., & Yi, Y. (2025). BiASAM: Bidirectional-Attention Guided Segment Anything Model for Very Few-Shot Medical Image Segmentation. IEEE Signal Processing Letters, 32, 246-250. https://doi.org/10.1109/lsp.2024.3513240

Vancouver

Zhou W, Guan G, Cui W, Yi Y. BiASAM: Bidirectional-Attention Guided Segment Anything Model for Very Few-Shot Medical Image Segmentation. IEEE Signal Processing Letters. 2025;32:246-250. doi:10.1109/lsp.2024.3513240

BibTeX
@article{Zhou2025,
  author = {Zhou, Wei and Guan, Guilin and Cui, Wei and Yi, Yugen},
  title = {BiASAM: Bidirectional-Attention Guided Segment Anything Model for Very Few-Shot Medical Image Segmentation},
  journal = {IEEE Signal Processing Letters},
  year = {2025},
  volume = {32},
  pages = {246--250},
  doi = {10.1109/lsp.2024.3513240}
}
Kostenlos starten →