Retinal Images Dataset for Model Eyes Distortion Analysis

This dataset contains retinal images and their corresponding vessel tracings collected for analyzing optical distortions in model eyes. The data was collected as part of our research on Quantifying Retinal Size and Shape Distortion in Different Ultra-Widefield Imaging Systems.

Dataset Contents

This dataset includes two compressed files:
- img.zip: Contains original retinal images
- trace.zip: Contains corresponding tracing retinal images

Image Specifications

01Daytona

- Resolution: 3900 × 3072 pixels
- File format: JPG
- Number of images: 24

02Mirante

- Resolution: 2048 × 2048 pixels  
- File format: JPG
- Number of images: 24

03CLARUS 500

- Resolution: 4064 × 4064 pixels
- File format: JPG
- Number of images: 24

04VG2001

- Resolution: 1280 × 1034 pixels
- File format: PNG
- Number of images: 24

05BMizar

- Resolution: 1536 × 1280 pixels
- File format: JPG
- Number of images: 24

06TRC-NW8

- Resolution: 2809 × 2108 pixels
- File format: PNG
- Number of images: 24

Data Collection Methods

Each model eye was imaged using six devices, including Daytona (Optos, Dunfermline, UK) with a FOV of 200°, Mirante (Nidek, Gamagori, Japan) with a FOV of 163°, CLARUS 500 (Carl Zeiss, Dublin, USA) with a FOV of 133°, VG2001 SS-OCT (SVision Imaging, Henan, China) with a FOV of 120°, BMizar SS-OCT (TowardPi, Beijing, China) with a FOV of 130°, and a traditional fundus camera, TRC-NW8 (Topcon, Tokyo, Japan) with a FOV of 45°.

Potential Applications

This dataset is particularly valuable for:
- Developing retinal image analysis tools
- Measuring retinal calibre in the periphery
- Validating image processing algorithms
- Investigating optical distortions in imaging devices

Usage Notes

Each image filename contains information about:
- Device type
- Model eye type
- Model eye axial length
- Image ID

Filename format: [device]_[type]_[axial length]_[ID].[ext]

Technical Validation

All images were reviewed for quality and clarity before inclusion in the dataset. 

Acknowledgements

We thank Joint Shantou International Eye Center Special Examination Department for their support in data collection and processing.

Citation

If you use this dataset, please cite: Huang B, Zheng C, Chen S, et al. Quantifying retinal size and shape distortion in different ultrawidefield imaging systems. BMJ Open Ophthalmology 2025;10:e001965. doi:10.1136/ bmjophth-2024-001965

If you have any problems, please contact drchenhaoyu@gmail.com.