Beyond Natural-Image Foundation Models: Benchmarking Satellite Pretraining for Ophthalmic Image Analysis

Lovre Antonio Budimir; Mingya Alexa Gong; Alyssa Foong Quinney; Ivana Matovinović; Yukun Zhou; Pearse A. Keane; Sven Lončarić; Marinko V. Šarunić

cs.CVAug 15, 2026

Beyond Natural-Image Foundation Models: Benchmarking Satellite Pretraining for Ophthalmic Image Analysis

Authors: Lovre Antonio Budimir, Mingya Alexa Gong, Alyssa Foong Quinney, Ivana Matovinović, Yukun Zhou, Pearse A. Keane, Sven Lončarić, Marinko V. Šarunić

Organizations: Faculty of Electrical Engineering and Computing, University of Zagreb, HR · Medical Physics and Biomedical Engineering, University College London, UK · Institute of Ophthalmology, University College London, UK · Department of Computer Science, University College London, UK ·

Abstract

Vision Foundation Models (VFMs) have emerged as a promising approach in medical imaging, producing broadly applicable systems that can be efficiently adapted across diverse imaging modalities, anatomical regions, and clinical tasks. However, VFMs require extensive training data, and their progress in medical image analysis is constrained by limited data availability, privacy concerns, and high development costs. To alleviate these constraints, medical VFMs (MedVFMs) are often built upon weights from generalist models pretrained on vast amounts of publicly available natural images, introducing a substantial distribution shift for medical task adaptation. To address this, we propose satellite imagery as a novel pretraining domain for MedVFM development and benchmarking, motivated by its closer visual alignment with medical data and its freedom from the privacy constraints that limit medical datasets. Across multiple ophthalmic imaging modalities, we compare DINOv3-SAT493m pretrained on 493 million satellite images against DINOv3-LVD1689m pretrained on 1.7 billion natural images, together with two medical specialist baselines: DINOv3-RETFound and MAE-RETFound. Our experiments show that satellite imagery is a stronger pretraining source than natural images for ophthalmic tasks, particularly on en face vascular-rich modalities. On several tasks, satellite pretraining matches or exceeds the medical specialists on high-resolution en face inputs, despite using no medical data.

Figures & tables

Figure 1: Motivation. Satellite image of the Lena River Delta in Russia (left; source: NASA, public domain) and an OCT angiography image (right; OCTA-500 dataset [ 16 ] ). Zoomed-in regions in the middle highlight the structural similarities between branching river networks and retinal vascular patterns observed in OCTA.

Figure 2: Visualization of dense features extracted from DINOv3-LVD1689m (middle) and DINOv3-SAT493m (bottom) across five ophthalmic imaging modalities: OCTA, UWF, CFP, OCT, and AOSLO. The first three principal components of a PCA computed over the feature space are mapped to RGB for visualization.

Figure 3: Cosine similarity scores computed between the dense feature at the red query patch and all other patch features for DINOv3-LVD1689m and DINOv3-SAT493m. DINOv3-SAT493m results in more consistent similarity scores along OCTA vessels (top) and UWF hemorrhagic regions (bottom), better matching the corresponding ground truth masks.

Dataset	Modality	Task	# Images	Resolution
		Segmentation
OCTA-500 (3M) [ 16 ]	OCTA	Large Vessels, Artery, Vein, FAZ	200	$304\times 304$
OCTA-500 (6M) [ 16 ]	OCTA	Large Vessels, Artery, Vein, FAZ	300	$400\times 400$
IDRiD [ 25 ]	CFP	4 lesion types, optic disc	81	$4288\times 2848$
FIVES [ 14 ]	CFP	Vessels	800	$2048\times 2048$
UWF-RHS [ 33 ]	UWF	Hemorrhages	215	$3900\times 3072$

Table 1: Overview of retinal imaging datasets used in the benchmark setup.

	OCTA-500 (3M) [ 16 ]				OCTA-500 (6M) [ 16 ]
OCTA	Vessel	Artery	Vein	FAZ	Vessel	Artery	Vein	FAZ
DINOv3-LVD1689m [ 27 ]	68.2	58.3	54.7	94.7	61.6	47.8	51.8	86.1
DINOv3-SAT493m [ 27 ]	71.8	62.7	60.4	94.1	64.4	52.1	55.3	87.9

Table 2: Quantitative comparison on OCTA segmentation task using OCTA-500 [ 16 ] dataset. Results are reported as Dice scores (%). The best result is shown in bold.

	IDRiD [ 25 ]						FIVES [ 14 ]
CFP	HE	MA	EX	SE	OD	Overall	Vessel
MAE-RETFound [ 34 ]	32.4	10.2	41.9	37.0	94.5	43.2	57.0
DINOv3-RETFound [ 35 ]	32.6	8.7	40.4	36.0	94.2	42.4	54.6
DINOv3-LVD1689m [ 27 ]	38.1	8.4	40.4	46.9	94.7	45.7	60.6
DINOv3-SAT493m [ 27 ]	37.0	13.2	41.9	54.5	93.5	48.0	60.5

Table 3: Quantitative comparison on CFP segmentation task. Left: performance on five segmentation tasks from the IDRiD dataset [ 25 ] , covering diabetic retinopathy lesions and retinal structures: hemorrhages (HE), microaneurysms (MA), hard exudates (EX), soft exudates (SE), and optic disc (OD). Right: performance on retinal vessel segmentation from the FIVES dataset [ 14 ] . Results are reported as Dice scores (%). The best result is shown in bold and the second-best result is underlined.

	UWF-RHS [ 33 ]	PRIME-FP20 [ 6 ]
UWF	Hemorrhages	Vessel
DINOv3-LVD1689m [ 27 ]	37.3	46.4
DINOv3-SAT493m [ 27 ]	42.5	47.9

Table 4: Quantitative comparison on UWF segmentation task. Left: performance on retinal hemorrhage segmentation from the UWF-RHS dataset [ 33 ] . Right: performance on retinal vessel segmentation from the PRIME-FP20 dataset [ 6 ] . Results are reported as Dice scores (%) and the best is shown in bold.

OCT	HCMS [ 11 ]								AMD-SD [ 12 ]
	Retinal Layer								Lesions
	RNFL	GCL+IPL	INL	OPL	ONL	IS	OS	RPE	SRF	IRF	PED	SHRM	IS/OS
DINOv3-LVD1689m [ 27 ]	85.3	85.7	64.4	72.3	88.3	71.5	74.6	83.3	37.6	28.1	28.9	32.7	46.7
DINOv3-SAT493m [ 27 ]	87.6	88.8	71.2	77.2	90.4	77.2	80.3	86.0	40.3	24.7	28.2	32.5	48.9

Table 5: Quantitative comparison on OCT HCMS [ 11 ] retinal layer segmentation and AMD-SD [ 12 ] lesion segmentation tasks. Results are reported as Dice scores (%). The best is shown in bold.

AOSLO	Capillaries
DINOv3-LVD1689m [ 27 ]	56.1
DINOv3-SAT493m [ 27 ]	60.2

Table 6: Quantitative comparison on AOSLO capillary segmentation task. Results are reported as Dice scores (%) and the best is shown in bold.

Figure 4: Dice score (DS) over training epochs for vasculature segmentation (vessels and capillaries) across four en face modalities: OCTA, CFP, UWF, and AOSLO.

	CFP			UWF			OCT
Method	Acc	QWK	wF1	Acc	QWK	wF1	Acc	QWK	wF1
MAE-RETFound [ 34 ]	75.3	85.0	81.0	–	–	–	88.9	84.7	88.8
DINOv3-RETFound [ 35 ]	74.3	82.5	73.7	–	–	–	–	–	–
DINOv3-LVD1689m [ 27 ]	78.0	88.4	78.7	66.9	87.8	66.8	86.2	84.2	86.5
DINOv3-SAT493m [ 27 ]	80.6	89.5	81.0	70.0	89.0	69.5	83.9	81.8	85.0

Table 7: Classification performance across retinal imaging modalities on MMRDR [ 29 ] dataset. Results are reported as accuracy (Acc), quadratic weighted kappa (QWK), and weighted F1 score (wF1). Best results are highlighted in bold.

Figure 5: Qualitative results on segmentation tasks across four different modalities.

Figure 6: Visualization of dense CFP features extracted from DINOv3-LVD1689m, DINOv3-SAT493m , and DINOv3-RETFound. The first three principal components of a PCA computed over the feature space are mapped to RGB for visualization. Input image size to the DINOv3 model is $1024\times 1024$ .

Ophthalmic Image Analysis

See all 42 papers

LUMOS: Latent Universal Medical Priors for Segmentation
Mar 1, 2026Zhuonan Liang, Wei Guo, Jie Gan +4Medical Image Segmentation Transfer Learning
The University of Sydney, Sydney, NSW 2006, Australia. · Biological System & Engineering Division, Lawrence Berkeley National Laboratory, Berkeley, CA 94720, USA. · Berkeley Biomedical Data Science Center, Lawrence Berkeley National Laboratory, Berkeley, CA 94720, USA.
Representation Transfer of Foundation Models for Ultra-Widefield Retinal Imaging
Aug 1, 2026Mingya Alexa Gong, Da Ma, Lovre Antonio Budimir +7Medical Imaging Foundation Models Diabetic Retinopathy Grading
Institute of Ophthalmology, University College London, London, United Kingdom · Wake Forest University School of Medicine, Winston-Salem, NC, USA · Virginia Tech-Wake Forest University School of Biomedical Engineering and Sciences, Blacksburg, VA, USA +5
Clinician-Friendly Foundation Models for Ophthalmic Image Diagnostics without Fine-Tuning or Technical Barriers
Apr 22, 2025Meng Wang, Tian Lin, Qingshan Hou +69Medical Imaging Foundation Models Ophthalmic Image Analysis

Mar 1, 2026J/K move · Enter open · S save

LUMOS: Latent Universal Medical Priors for Segmentation

Zhuonan Liang, Wei Guo, Jie Gan, Yaxuan Song+3

The University of Sydney, Sydney, NSW 2006, Australia. · Biological System & Engineering Division, Lawrence Berkeley National Laboratory, Berkeley, CA 94720, USA. · Berkeley Biomedical Data Science Center, Lawrence Berkeley National Laboratory, Berkeley, CA 94720, USA.

Medical Image Segmentation Transfer Learning Vision Foundation Model Adaptation

Open paper AskPDF arXiv

General vision foundation models (VFMs) have been primarily developed on natural images, and their utility for medical image segmentation is therefore often considered to depend on costly adaptation or domain-specific fine-tuning. In this paper, we revisit this assumption from a different perspective: rather than requiring VFM segmentors to relearn visual regularities, we investigate whether the low-level visual priors necessary for anatomical delineation already lie dormant within general VFMs. We observe that frozen VFMs, despite lacking medical supervision, encode transferable visual regularities. These properties are not exclusive to natural images but are also fundamental to medical image understanding. Motivated by this observation, we propose Latent Universal Medical PriOrs for Segmentation (LUMOS), a novel framework that amplifies general VFM priors to conventional medical segmentors. LUMOS consists of two key components: (1) Pathfinder that distills visual cues from a frozen vision foundation model, and (2) Inspiror that sparks the conventional medical networks with spatial guidance from distilled visual regularities. In this way, the segmentor is relieved from learning complex visual regularities entirely from limited medical annotations and can instead focus on task-specific anatomical delineation. Across diverse medical datasets and token-based VFMs, LUMOS shows that general VFMs can serve as spatial prior generators when their frozen token spaces preserve patch-level pattern relevance. DINO provides stable matched-backbone gains, while SigLIP exposes VFM-specific sensitivity caused by its different token granularity and representation objective.

Abstract

Ophthalmic Image Analysis

Explore similar work

LUMOS: Latent Universal Medical Priors for Segmentation

Representation Transfer of Foundation Models for Ultra-Widefield Retinal Imaging

Clinician-Friendly Foundation Models for Ophthalmic Image Diagnostics without Fine-Tuning or Technical Barriers