QUARTZ: Cross-Dataset Artificial Intelligence-Enabled Retinal Vasculometry
Examination of retinal images provides a non-invasive means to assess the body's microvascular system, with vessel morphology linked to ocular and systemic diseases. QUARTZ (QUantitative Analysis of Retinal vessel Topology and siZe) is an automated, artificial intelligence-enabled retinal vasculometry system designed for processing large-scale retinal image datasets, including over 4 million images from within the National Health Service, to facilitate epidemiological studies. The system is uniquely capable of extracting useful information from lower quality images, including those which are partially illuminated, enabling analysis of a greater proportion of images. To enhance cross-dataset generalisation of the system from an earlier dataset-tailored approach, this work leverages 26 annotated retinal image datasets spanning diverse demographics, diseases, camera types, and imaging protocols. Evaluation on the external EPIC-Norfolk dataset demonstrates robust performance with F1 scores of 0.8305, 0.8477, 0.8741, 0.9532, and 0.8677 for arteriole, venule, vessel, optic disc, and optic cup segmentation, respectively, and 0.9197 for image quality classification. The cross-dataset generalisation approach demonstrated increased performance for arteriole/venule segmentation and optic disc segmentation. However, improvements in performance were not observed for optic cup segmentation, or for image quality classification. An online tool is provided to support wider adoption and reproducible biomarker research across diverse populations and imaging settings. In addition, the new public CHASE-AV database is introduced for arteriole and venule segmentation.
Authors
- Paolo Remagnino (ORCID: https://orcid.org/0000-0002-9168-7746)
- Royce Shakespeare (ORCID: https://orcid.org/0000-0001-9955-4377)
- John Anderson (ORCID: https://orcid.org/0000-0002-2355-9742)
- Ellen E. Freeman (ORCID: https://orcid.org/0000-0002-1403-8427)
- Adnan Tufail (ORCID: https://orcid.org/0000-0001-6131-7640)
- Jiri Fajtl (ORCID: https://orcid.org/0000-0003-2943-5342)
- Abraham Olvera‐Barrios (ORCID: https://orcid.org/0000-0002-3305-4465)
- Louis Bolter (ORCID: https://orcid.org/0000-0002-8710-8916)
- Roshan A. Welikala (ORCID: https://orcid.org/0000-0001-8503-6536)
- Sarah Ann Barman (ORCID: https://orcid.org/0000-0001-5302-0169)
- Catherine A. Egan (ORCID: https://orcid.org/0000-0001-5593-1169)
- Christopher G. Owen (ORCID: https://orcid.org/0000-0003-1135-5977)
- Ryan Chambers (ORCID: https://orcid.org/0000-0001-5322-4209)
- Miriam Kolko (ORCID: https://orcid.org/0000-0001-8697-0734)
- Anthony P. Khawaja (ORCID: https://orcid.org/0000-0001-6802-8585)
- Farzana Rahman (ORCID: https://orcid.org/0000-0002-4785-5845)
- Paul J. Foster (ORCID: https://orcid.org/0000-0002-4755-177X)
- Alicja R. Rudnicka (ORCID: https://orcid.org/0000-0003-0369-8574)
- Alasdair Warwick (ORCID: https://orcid.org/0000-0002-0800-2890)
- Gordon Johnson (ORCID: https://orcid.org/0000-0002-0588-9174)
- M. V. Ramesh (ORCID: https://orcid.org/0009-0007-7964-2741)
- EPIC-Norfolk Investigators
Institutions
- University of Copenhagen (DK)
- Moorfields Eye Hospital NHS Foundation Trust (GB)
- University of Ottawa (CA)
- St George's, University of London (GB)
- Durham University (GB)
- Kingston University London (GB)
- Homerton University Hospital NHS Foundation Trust (GB)
- Moorfields Eye Hospital (GB)
- University College London (GB)
Publication Details
- Journal
- medRxiv
- Published
- 2026-10-29
- DOI
- https://doi.org/10.64898/2026.10.28.26363792
- Primary Topic
- Retinal Imaging and Analysis
- Type
- preprint
Funders
- Wellcome Trust
- Government of Canada
- Cancer Research UK
- National Institute for Health and Care Research
- Moorfields Eye Hospital NHS Foundation Trust
- Canadian Institutes of Health Research
- Medical Research Council