EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Retinal Vessel Segmentation Based on a Lightweight U-Net and Reverse Attention

Fernando Daniel Hernandez-Gutierrez, Eli Gabriel Avina-Bravo, Mario Alberto Ibarra-Manzano, Jose Ruiz-Pinales, Emmanuel Ovalle-Magallanes, Juan Gabriel Avina-Cervantes

Mathematics · 05.07.2025 · 13 (13) · 2203

DOI: 10.3390/math13132203 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Mathematics, online erschienen am 05.07.2025. Verfasst von Fernando Daniel Hernandez-Gutierrez und 5 weiteren Autor:innen. Die Arbeit wurde bisher 11-mal zitiert (Crossref) und nennt 52 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: University of Guanajuato, Secretary of Science, Humanities, Technology and Innovation (SECIHTI).

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

U-shaped architectures have achieved exceptional performance in medical image segmentation. Their aim is to extract features by two symmetrical paths: an encoder and a decoder. We propose a lightweight U-Net incorporating reverse attention and a preprocessing framework for accurate retinal vessel segmentation. This concept could be of benefit to portable or embedded recognition systems with limited resources for real-time operation. Compared to the baseline model (7.7 M parameters), the proposed U-Net model has only 1.9 M parameters and was tested on the DRIVE (Digital Retinal Images for Vesselness Extraction), CHASE (Child Heart and Health Study in England), and HRF (High-Resolution Fundus) datasets for vesselness analysis. The proposed model achieved Dice coefficients and IoU scores of 0.7871 and 0.6318 on the DRIVE dataset, 0.8036 and 0.6910 on the CHASE-DB1 Retinal Vessel Reference dataset, as well as 0.6902 and 0.5270 on the HRF dataset, respectively. Notably, the integration of the reverse attention mechanism contributed to a more accurate delineation of thin and peripheral vessels, which are often undetected by conventional models. The model comprised 1.94 million parameters and 12.21 GFLOPs. Furthermore, during inference, the model achieved a frame rate average of 208 FPS and a latency of 4.81 ms. These findings support the applicability of the proposed model in real-world clinical and mobile healthcare environments where efficiency and Accuracy are essential.

Förderung

  • University of Guanajuato
  • Secretary of Science, Humanities, Technology and Innovation (SECIHTI)

Zitieren

APA

Hernandez-Gutierrez, F. D., Avina-Bravo, E. G., Ibarra-Manzano, M. A., Ruiz-Pinales, J., Ovalle-Magallanes, E., & Avina-Cervantes, J. G. (2025). Retinal Vessel Segmentation Based on a Lightweight U-Net and Reverse Attention. Mathematics, 13(13), 2203. https://doi.org/10.3390/math13132203

Vancouver

Hernandez-Gutierrez FD, Avina-Bravo EG, Ibarra-Manzano MA, Ruiz-Pinales J, Ovalle-Magallanes E, Avina-Cervantes JG. Retinal Vessel Segmentation Based on a Lightweight U-Net and Reverse Attention. Mathematics. 2025;13(13):2203. doi:10.3390/math13132203

BibTeX
@article{HernandezGutierrez2025,
  author = {Hernandez-Gutierrez, Fernando Daniel and Avina-Bravo, Eli Gabriel and Ibarra-Manzano, Mario Alberto and Ruiz-Pinales, Jose and Ovalle-Magallanes, Emmanuel and Avina-Cervantes, Juan Gabriel},
  title = {Retinal Vessel Segmentation Based on a Lightweight U-Net and Reverse Attention},
  journal = {Mathematics},
  year = {2025},
  volume = {13},
  number = {13},
  pages = {2203},
  doi = {10.3390/math13132203}
}
Kostenlos starten →