EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Data-Efficient Bone Segmentation Using Feature Pyramid- Based SegFormer

Naohiro Masuda iD, Keiko Ono iD, Daisuke Tawara iD, Yusuke Matsuura iD, Kentaro Sakabe

Sensors · 2024

DOI: 10.3390/s25010081 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Sensors, 2024. Verfasst von Naohiro Masuda und 4 weiteren Autor:innen. Beteiligte Einrichtung: Master's Program in Information and Computer Science, Doshisha University, Kyoto 610-0394, Japan. Hauptthemen laut MEDLINE: Image Processing, Computer-Assisted, Bone and Bones. Die Arbeit wurde bisher 9-mal zitiert (Crossref) und nennt 30 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: JSPS KAKENHI.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

The semantic segmentation of bone structures demands pixel-level classification accuracy to create reliable bone models for diagnosis. While Convolutional Neural Networks (CNNs) are commonly used for segmentation, they often struggle with complex shapes due to their focus on texture features and limited ability to incorporate positional information. As orthopedic surgery increasingly requires precise automatic diagnosis, we explored SegFormer, an enhanced Vision Transformer model that better handles spatial awareness in segmentation tasks. However, SegFormer’s effectiveness is typically limited by its need for extensive training data, which is particularly challenging in medical imaging, where obtaining labeled ground truths (GTs) is a costly and resource-intensive process. In this paper, we propose two models and their combination to enable accurate feature extraction from smaller datasets by improving SegFormer. Specifically, these include the data-efficient model, which deepens the hierarchical encoder by adding convolution layers to transformer blocks and increases feature map resolution within transformer blocks, and the FPN-based model, which enhances the decoder through a Feature Pyramid Network (FPN) and attention mechanisms. Testing our model on spine images from the Cancer Imaging Archive and our own hand and wrist dataset, ablation studies confirmed that our modifications outperform the original SegFormer, U-Net, and Mask2Former. These enhancements enable better image feature extraction and more precise object contour detection, which is particularly beneficial for medical imaging applications with limited training data.

Schlagwörter

MeSH-Schlagwörter (MEDLINE)

Image Processing, Computer-AssistedBone and BonesHumansNeural Networks, ComputerAlgorithmsSpine

Stichwörter der Autor:innen

Mask2FormerSegFormerfeature pyramid networksemantic segmentationtransformer block

Autor:innen und Institutionen

  • Master's Program in Information and Computer Science, Doshisha University, Kyoto 610-0394, Japan.
  • Department of Intelligent Information Engineering and Sciences, Doshisha University, Kyoto 610-0394, Japan.
  • Department of Advanced Science and Technology, Ryukoku University, Kyoto 520-2194, Japan.
  • Department of Orthopedic Surgery, Chiba University, Chiba 260-8677, Japan.

ORCID-Profile: Naohiro Masuda, Keiko Ono, Daisuke Tawara, Yusuke Matsuura

Förderung

  • JSPS KAKENHI

Zitieren

APA

Masuda, N., Ono, K., Tawara, D., Matsuura, Y., & Sakabe, K. (2024). Data-Efficient Bone Segmentation Using Feature Pyramid- Based SegFormer. Sensors. https://doi.org/10.3390/s25010081

Vancouver

Masuda N, Ono K, Tawara D, Matsuura Y, Sakabe K. Data-Efficient Bone Segmentation Using Feature Pyramid- Based SegFormer. Sensors. 2024. doi:10.3390/s25010081

BibTeX
@article{Masuda2024,
  author = {Masuda, Naohiro and Ono, Keiko and Tawara, Daisuke and Matsuura, Yusuke and Sakabe, Kentaro},
  title = {Data-Efficient Bone Segmentation Using Feature Pyramid- Based SegFormer},
  journal = {Sensors},
  year = {2024},
  doi = {10.3390/s25010081}
}
Kostenlos starten →