![[2511.14654] Improving segmentation of retinal arteries and veins using cardiac signal in doppler holograms](https://arxiv.org/static/browse/0.3.4/images/arxiv-logo-fb.png){kind=logo}
[2511.14654] Improving segmentation of retinal arteries and veins using cardiac signal in doppler holograms
Summary
This article presents a novel approach to segmenting retinal arteries and veins using cardiac signals in Doppler holograms, enhancing traditional methods by incorporating temporal dynamics for improved accuracy.
Why It Matters
Accurate segmentation of retinal blood vessels is crucial for assessing retinal health and diagnosing diseases. This research highlights the potential of integrating temporal data into existing models, which could lead to better diagnostic tools and insights into retinal hemodynamics.
Key Takeaways
- Introduces a new method for retinal artery-vein segmentation using Doppler holography.
- Incorporates temporal dynamics into standard segmentation architectures like U-Nets.
- Demonstrates that time-resolved preprocessing can enhance deep learning performance.
- Provides a publicly available dataset for further research and validation.
- Highlights the importance of quantitative assessment in retinal hemodynamics.
Computer Science > Computer Vision and Pattern Recognition arXiv:2511.14654 (cs) [Submitted on 18 Nov 2025 (v1), last revised 19 Feb 2026 (this version, v2)] Title:Improving segmentation of retinal arteries and veins using cardiac signal in doppler holograms Authors:Marius Dubosc, Yann Fischer, Zacharie Auray, Nicolas Boutry, Edwin Carlinet, Michael Atlan, Thierry Geraud View a PDF of the paper titled Improving segmentation of retinal arteries and veins using cardiac signal in doppler holograms, by Marius Dubosc and 6 other authors View PDF HTML (experimental) Abstract:Doppler holography is an emerging retinal imaging technique that captures the dynamic behavior of blood flow with high temporal resolution, enabling quantitative assessment of retinal hemodynamics. This requires accurate segmentation of retinal arteries and veins, but traditional segmentation methods focus solely on spatial information and overlook the temporal richness of holographic data. In this work, we propose a simple yet effective approach for artery-vein segmentation in temporal Doppler holograms using standard segmentation architectures. By incorporating features derived from a dedicated pulse analysis pipeline, our method allows conventional U-Nets to exploit temporal dynamics and achieve performance comparable to more complex attention- or iteration-based models. These findings demonstrate that time-resolved preprocessing can unlock the full potential of deep learning for Doppler holography, openi...