EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

FNeXter: A Multi-Scale Feature Fusion Network Based on ConvNeXt and Transformer for Retinal OCT Fluid Segmentation

Zhiyuan Niu, Zhuo Deng, Weihao Gao, Shurui Bai, Zheng Gong iD, Chucheng Chen, Fuju Rong iD, Fang Li, Lan Ma iD

Sensors · 10.04.2024 · 24 (8) · 2425

DOI: 10.3390/s24082425 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Sensors, online erschienen am 10.04.2024. Verfasst von Zhiyuan Niu und 8 weiteren Autor:innen. Beteiligte Einrichtung: Tsinghua Shenzhen International Graduate School, Tsinghua University, Shenzhen 518055, China. Hauptthemen laut MEDLINE: Tomography, Optical Coherence, Retina. Die Arbeit wurde bisher 22-mal zitiert (Crossref) und nennt 43 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Shenzhen Sustainable Development Project, Shenzhen Basic Research Project (Natural Science Fund), Shenzhen International Science and Technology Information Center and the Shenzhen Bay Laboratory.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

The accurate segmentation and quantification of retinal fluid in Optical Coherence Tomography (OCT) images are crucial for the diagnosis and treatment of ophthalmic diseases such as age-related macular degeneration. However, the accurate segmentation of retinal fluid is challenging due to significant variations in the size, position, and shape of fluid, as well as their complex, curved boundaries. To address these challenges, we propose a novel multi-scale feature fusion attention network (FNeXter), based on ConvNeXt and Transformer, for OCT fluid segmentation. In FNeXter, we introduce a novel global multi-scale hybrid encoder module that integrates ConvNeXt, Transformer, and region-aware spatial attention. This module can capture long-range dependencies and non-local similarities while also focusing on local features. Moreover, this module possesses the spatial region-aware capabilities, enabling it to adaptively focus on the lesions regions. Additionally, we propose a novel self-adaptive multi-scale feature fusion attention module to enhance the skip connections between the encoder and the decoder. The inclusion of this module elevates the model’s capacity to learn global features and multi-scale contextual information effectively. Finally, we conduct comprehensive experiments to evaluate the performance of the proposed FNeXter. Experimental results demonstrate that our proposed approach outperforms other state-of-the-art methods in the task of fluid segmentation.

Schlagwörter

MeSH-Schlagwörter (MEDLINE)

Tomography, Optical CoherenceRetinaHumansAlgorithmsNeural Networks, ComputerImage Processing, Computer-AssistedMacular Degeneration

Stichwörter der Autor:innen

Transformerattentionoptical coherence tomographyretinal fluid segmentation

Autor:innen und Institutionen

  • Tsinghua Shenzhen International Graduate School, Tsinghua University, Shenzhen 518055, China.

ORCID-Profile: Zheng Gong, Fuju Rong, Lan Ma

Förderung

  • Shenzhen Sustainable Development Project
  • Shenzhen Basic Research Project (Natural Science Fund)
  • Shenzhen International Science and Technology Information Center and the Shenzhen Bay Laboratory

Zitieren

APA

Niu, Z., Deng, Z., Gao, W., Bai, S., Gong, Z., Chen, C., Rong, F., Li, F., & Ma, L. (2024). FNeXter: A Multi-Scale Feature Fusion Network Based on ConvNeXt and Transformer for Retinal OCT Fluid Segmentation. Sensors, 24(8), 2425. https://doi.org/10.3390/s24082425

Vancouver

Niu Z, Deng Z, Gao W, Bai S, Gong Z, Chen C, et al. FNeXter: A Multi-Scale Feature Fusion Network Based on ConvNeXt and Transformer for Retinal OCT Fluid Segmentation. Sensors. 2024;24(8):2425. doi:10.3390/s24082425

BibTeX
@article{Niu2024,
  author = {Niu, Zhiyuan and Deng, Zhuo and Gao, Weihao and Bai, Shurui and Gong, Zheng and Chen, Chucheng and Rong, Fuju and Li, Fang and Ma, Lan},
  title = {FNeXter: A Multi-Scale Feature Fusion Network Based on ConvNeXt and Transformer for Retinal OCT Fluid Segmentation},
  journal = {Sensors},
  year = {2024},
  volume = {24},
  number = {8},
  pages = {2425},
  doi = {10.3390/s24082425}
}
Kostenlos starten →