Automated Cone Photoreceptor Detection in Adaptive Optics Flood Illumination Ophthalmoscopy
Ophthalmology Science · 2025-05 · 5 (3) · 100675
DOI: 10.1016/j.xops.2024.100675 ↗Kurz gefasst
Fachartikel in der Zeitschrift Ophthalmology Science, 2025-05. Verfasst von Sander Wooning und 19 weiteren Autor:innen. Beteiligte Einrichtung: Biomedical Imaging Group Rotterdam, Department of Radiology & Nuclear Medicine, Erasmus MC, Rotterdam, The Netherlands. Die Arbeit wurde bisher 8-mal zitiert (Crossref) und nennt 47 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY-NC-ND 4.0). Gefördert von: Health~Holland.
Abstract (Originalsprache)
Purpose
To develop and validate a deep learning-based model for detecting cone photoreceptor cells in adaptive optics flood illumination ophthalmoscopy (AO-FIO).
Design
Healthy volunteer study.
Participants
A total of 36 healthy participants were included.
Methods
The imaging protocol consisted of 21 AO-FIO images per eye acquired with the rtx1 adaptive optics retinal camera (Imagine Eyes), 4° × 4° each with 2° overlap, imaging a retinal patch 4° nasal (N) to 12° temporal (T) and -5° inferior to 5° superior relative to the fovea. Each image was divided into patches of 128 × 128 pixels, with a 20-pixel overlap. A training set (625 patches) from 18 subjects (32 ± 12 years, 6 males and 12 females) was annotated by a single center, whereas the test set (54 patches) from 18 subjects (40 ± 16 years, 11 males and 7 females) was annotated by graders from 3 different institutions. The deep learning model, based on the U-Net architecture, underwent a parameter search using the tree-structured Parzen estimator. Main Outcome Measures: The F1 score was used to determine both intragrader and intergrader agreements and to evaluate the model's performance compared with the automated detection by the manufacturer's software (AOdetect Mosaic).
Results
The average intragrader agreement was 0.85 ± 0.06 between 2°N and 2°T, followed by 0.83 ± 0.09 between 3 and 6°T, and 0.80 ± 0.10 between 7 and 10°T. The average intergrader agreement for the 3 centers was 0.84 ± 0.05, 0.79 ± 0.05, and 0.76 ± 0.06 at 2°N-2°T, 3-6°T, and 7-10°T, respectively. The best combination of hyperparameters based on the tree-structured Parzen estimator algorithm achieved an F1 score of 0.89 ± 0.04. The average agreement between the model and the graders was 0.87 ± 0.04, 0.85 ± 0.03, and 0.81 ± 0.03 at 2°N-2°T, 3°-6°T, and 7°-10°T, respectively. These values were higher than those between AOdetect's auto detection without manual correction and the graders (0.84 ± 0.05, 0.79 ± 0.03, and 0.68 ± 0.04, respectively). A reduction in cone density was noted at greater eccentricities, in line with previous research findings, and the model indicated variations in estimating cell density for individuals aged 18 to 30 compared with those aged ≥50 years.
Conclusions
The performance of the developed deep learning-based model, AO-FIO ConeDetect, was comparable to that of graders from 3 medical centers. It outperformed the manufacturers' software auto-detection, particularly at higher eccentricities (7°-10°T). Hence, the model could reduce the need for manual correction and enable faster cone mosaic analyses. Financial Disclosures: Proprietary or commercial disclosure may be found in the Footnotes and Disclosures at the end of this article.
Schlagwörter
Stichwörter der Autor:innen
Autor:innen und Institutionen
- Biomedical Imaging Group Rotterdam, Department of Radiology & Nuclear Medicine, Erasmus MC, Rotterdam, The Netherlands.
- Leiden Institute of Advanced Computer Science, Leiden University, Leiden, The Netherlands.
- Department of Ophthalmology, Erasmus MC, Rotterdam, The Netherlands.
- Department of Epidemiology, Erasmus MC, Rotterdam, The Netherlands.
- Department of Ophthalmology, Ghent University Hospital, Ghent, Belgium.
- Department of Head & Skin, Ghent University, Ghent, Belgium.
Förderung
- Health~Holland
Zitieren
APA
Wooning, S., Heutinck, P. A. T., Liman, K., Hennekam, S., van Haute, M., van den Broeck, F., Leroy, B., Sampson, D. M., Roshandel, D., Chen, F. K., Pelt, D. M., van den Born, L. I., Verhoeven, V. J. M., Klaver, C. C. W., Thiadens, A. A. H. J., Durand, M., Chateau, N., van Walsum, T., Andrade De Jesus, D., & Sanchez Brea, L. (2025). Automated Cone Photoreceptor Detection in Adaptive Optics Flood Illumination Ophthalmoscopy. Ophthalmology Science, 5(3), 100675. https://doi.org/10.1016/j.xops.2024.100675
Vancouver
Wooning S, Heutinck PAT, Liman K, Hennekam S, van Haute M, van den Broeck F, et al. Automated Cone Photoreceptor Detection in Adaptive Optics Flood Illumination Ophthalmoscopy. Ophthalmology Science. 2025;5(3):100675. doi:10.1016/j.xops.2024.100675
BibTeX
@article{Wooning2025,
author = {Wooning, Sander and Heutinck, Pam A T and Liman, Kubra and Hennekam, Sem and van Haute, Manon and van den Broeck, Filip and Leroy, Bart and Sampson, Danuta M and Roshandel, Danial and Chen, Fred K and Pelt, Daniel M and van den Born, L Ingeborgh and Verhoeven, Virginie J M and Klaver, Caroline C W and Thiadens, Alberta A H J and Durand, Marine and Chateau, Nicolas and van Walsum, Theo and Andrade De Jesus, Danilo and Sanchez Brea, Luisa},
title = {Automated Cone Photoreceptor Detection in Adaptive Optics Flood Illumination Ophthalmoscopy},
journal = {Ophthalmology Science},
year = {2025},
volume = {5},
number = {3},
pages = {100675},
doi = {10.1016/j.xops.2024.100675}
}
EUVIMED