[Paper Review] OCTolyzer: Fully automatic toolkit for segmentation and feature extracting in optical coherence tomography and scanning laser ophthalmoscopy data
OCTolyzer is an open-source, fully automatic toolkit for segmenting and extracting clinically meaningful features from optical coherence tomography (OCT) and scanning laser ophthalmoscopy (SLO) data. It enables high-reproducibility retinochoroidal measurements—such as choroid thickness, volume, and vascular index—using deep learning and metadata-driven segmentation, with OCT processing completed in under 2 seconds per B-scan on a standard CPU.
Optical coherence tomography (OCT) and scanning laser ophthalmoscopy (SLO) of the eye has become essential to ophthalmology and the emerging field of oculomics, thus requiring a need for transparent, reproducible, and rapid analysis of this data for clinical research and the wider research community. Here, we introduce OCTolyzer, the first open-source toolkit for retinochoroidal analysis in OCT/SLO data. It features two analysis suites for OCT and SLO data, facilitating deep learning-based anatomical segmentation and feature extraction of the cross-sectional retinal and choroidal layers and en face retinal vessels. We describe OCTolyzer and evaluate the reproducibility of its OCT choroid analysis. At the population level, metrics for choroid region thickness were highly reproducible, with a mean absolute error (MAE)/Pearson correlation for macular volume choroid thickness (CT) of 6.7$μ$m/0.99, macular B-scan CT of 11.6$μ$m/0.99, and peripapillary CT of 5.0$μ$m/0.99. Macular choroid vascular index (CVI) also showed strong reproducibility, with MAE/Pearson for volume CVI yielding 0.0271/0.97 and B-scan CVI 0.0130/0.91. At the eye level, measurement noise for regional and vessel metrics was below 5% and 20% of the population's variability, respectively. Outliers were caused by poor-quality B-scans with thick choroids and invisible choroid-sclera boundary. Processing times on a laptop CPU were under three seconds for macular/peripapillary B-scans and 85 seconds for volume scans. OCTolyzer can convert OCT/SLO data into reproducible and clinically meaningful retinochoroidal features and will improve the standardisation of ocular measurements in OCT/SLO image analysis, requiring no specialised training or proprietary software to be used. OCTolyzer is freely available here: https://github.com/jaburke166/OCTolyzer.
Motivation & Objective
- To address the lack of open-source, automated pipelines for processing OCT and SLO data into clinically meaningful retinochoroidal features.
- To improve standardization in retinal and choroidal image analysis across research and clinical settings.
- To enable high-throughput, reproducible measurement of choroid thickness, volume, and vascular index from OCT and SLO images.
- To support large-scale ophthalmic image analysis by eliminating manual segmentation and reducing dependency on proprietary software.
- To extend accessibility of advanced retinal and choroidal metrics to researchers without specialized training or software.
Proposed method
- Utilizes image metadata and deep learning to perform automatic retinal layer segmentation in OCT scans.
- Employs a U-Net-based deep learning model for choroid layer segmentation in OCT, trained on systemic health-related data.
- Integrates en face SLO images for anatomical segmentation and feature extraction of retinal vessels.
- Processes .vol RAW files from HEYEX viewer using EyePy for Python-based import, with support for SLO and OCT data.
- Applies physical space conversion from pixel coordinates to micrometers using OCT system calibration data.
- Computes metrics including macular and peripapillary choroid thickness, volume, and choroid vascular index (CVI) from segmented regions.
Experimental results
Research questions
- RQ1Can a fully automatic, open-source pipeline achieve high reproducibility in measuring choroid thickness and volume from OCT and SLO data?
- RQ2How does OCTolyzer perform in segmenting the choroid layer in OCT scans with varying image quality and anatomical variability?
- RQ3To what extent can OCTolyzer extract reliable en face retinal vessel features from SLO images without manual intervention?
- RQ4How does the pipeline’s performance compare across different OCT scanner types and acquisition protocols?
- RQ5Can OCTolyzer support large-scale, standardized retinochoroidal phenotyping in population-based studies?
Key findings
- Macular volume choroid thickness showed a mean absolute error of 6.7 µm, with Pearson and Spearman correlation coefficients of 0.9933 and 0.9969, respectively, indicating high reproducibility.
- Peripapillary choroid thickness had a mean absolute error of 5.0 µm and correlation coefficients of 0.9942 (Pearson) and 0.9940 (Spearman), demonstrating excellent reproducibility.
- Macular choroid vascular index (CVI) had a mean absolute error of 0.0271 for volume CVI and 0.0130 for B-scan CVI, with correlation coefficients of 0.9669 and 0.9090, respectively.
- At the eye level, measurement error in regional and vessel metrics was below 5% and 20% of population variability, respectively, indicating robustness.
- Major outliers were linked to poor-quality B-scans with thick choroids and indistinct choroid-sclera boundaries, highlighting image quality as a key factor.
- OCT processing on a standard laptop CPU took under 2 seconds for macular or peripapillary B-scans and 85 seconds for volume scans, demonstrating computational efficiency.
Better researchstarts right now
From reading papers to final review, dramatically reduce your research time.
No credit card · Free plan available
This review was created by AI and reviewed by human editors.