# Benchmarking Retinal Blood Vessel Segmentation Models for Cross-Dataset and Cross-Disease Generalization

arXiv: https://arxiv.org/abs/2406.14994 · HTML: https://mlsanity.com/paper/2406.14994/

## Key facts

- Cited by 2 papers in the map (data through 14 Oct 2026).
- Builds on 17 of its 49 references that are in the map.
- First posted on arXiv 21 Jun 2024.

## Builds on (most cited first, 17 of 17)

- [U-Net: Convolutional Networks for Biomedical Image Segmentation](https://mlsanity.com/paper/1505.04597/) (2015)
- [V-Net: Fully Convolutional Neural Networks for Volumetric Medical Image Segmentation](https://mlsanity.com/paper/1606.04797/) (2016)
- [UNet++: A Nested U-Net Architecture for Medical Image Segmentation](https://mlsanity.com/paper/1807.10165/) (2018)
- [clDice -- A Novel Topology-Preserving Loss Function for Tubular Structure Segmentation](https://mlsanity.com/paper/2003.07311/) (2020)
- [CE-Net: Context Encoder Network for 2D Medical Image Segmentation](https://mlsanity.com/paper/1903.02740/) (2019)
- [SA-UNet: Spatial Attention U-Net for Retinal Vessel Segmentation](https://mlsanity.com/paper/2004.03696/) (2020)
- [IterNet: Retinal Image Segmentation Utilizing Structural Redundancy in Vessel Networks](https://mlsanity.com/paper/1912.05763/) (2019)
- [Generalist Vision Foundation Models for Medical Imaging: A Case Study of Segment Anything Model on Zero-Shot Medical Segmentation](https://mlsanity.com/paper/2304.12637/) (2023)
- [Deep Vessel Segmentation By Learning Graphical Connectivity](https://mlsanity.com/paper/1806.02279/) (2018)
- [LadderNet: Multi-path networks based on U-Net for medical image segmentation](https://mlsanity.com/paper/1810.07810/) (2018)
- [RV-GAN: Segmenting Retinal Vascular Structure in Fundus Photographs using a Novel Multi-scale Generative Adversarial Network](https://mlsanity.com/paper/2101.00535/) (2021)
- [From CNN to Transformer: A Review of Medical Image Segmentation Models](https://mlsanity.com/paper/2308.05305/) (2023)
- [Study Group Learning: Improving Retinal Vessel Segmentation Trained with Noisy Labels](https://mlsanity.com/paper/2103.03451/) (2021)
- [ROOD-MRI: Benchmarking the robustness of deep learning segmentation models to out-of-distribution and corrupted data in MRI](https://mlsanity.com/paper/2203.06060/) (2022)
- [On the Optimal Combination of Cross-Entropy and Soft Dice Losses for Lesion Segmentation with Out-of-Distribution Robustness](https://mlsanity.com/paper/2209.06078/) (2022)
- [The Little W-Net That Could: State-of-the-Art Retinal Vessel Segmentation with Minimalistic Models](https://mlsanity.com/paper/2009.01907/) (2020)
- [A Divide-and-Conquer Approach towards Understanding Deep Networks](https://mlsanity.com/paper/1907.06194/) (2019)

## Built on by (most cited first, 2 of 2)

- [Uncertainty-Aware Retinal Vessel Segmentation via Ensemble Distillation](https://mlsanity.com/paper/2509.11689/) (2025)
- [Effect of Input Resolution on Retinal Vessel Segmentation Performance: An Empirical Study Across Five Datasets](https://mlsanity.com/paper/2604.02977/) (2026)

## Abstract (arXiv)

Retinal blood vessel segmentation can extract clinically relevant information from fundus images. As manual tracing is cumbersome, algorithms based on Convolution Neural Networks have been developed. Such studies have used small publicly available datasets for training and measuring performance, running the risk of overfitting. Here, we provide a rigorous benchmark for various architectural and training choices commonly used in the literature on the largest dataset published to date. We train and evaluate five published models on the publicly available FIVES fundus image dataset, which exceeds previous ones in size and quality and which contains also images from common ophthalmological conditions (diabetic retinopathy, age-related macular degeneration, glaucoma). We compare the performance of different model architectures across different loss functions, levels of image qualitiy and ophthalmological conditions and assess their ability to perform well in the face of disease-induced domain shifts. Given sufficient training data, basic architectures such as U-Net perform just as well as more advanced ones, and transfer across disease-induced domain shifts typically works well for most architectures. However, we find that image quality is a key factor determining segmentation outcomes. When optimizing for segmentation performance, investing into a well curated dataset to train a standard architecture yields better results than tuning a sophisticated architecture on a smaller dataset or one with lower image quality. We distilled the utility of architectural advances in terms of their clinical relevance therefore providing practical guidance for model choices depending on the circumstances of the clinical setting