Related Experiment Video
Updated: Sep 29, 2025
04:04
Measuring Retinal Vessel Diameter from Mouse Fluorescent Angiography Images
Published on: May 19, 2023
810
Comparing Measurements of Vascular Diameter Using Adaptative Optics Imaging and Conventional Fundus Imaging
Thibaud Mautuit1, Rachel Semecas1, Stephen Hogg2
1Department of Ophthalmology, University Hospital of Grenoble-Alpes, University Grenoble Alpes, Inserm, CHU Grenoble Alpes, HP2, 38000 Grenoble, France.
Diagnostics (Basel, Switzerland)
|March 25, 2022
Summary
This study compared retinal vascular diameter measurements from standard fundus photos and adaptive optics (AO) images. IVAN software showed stronger correlation with AO imaging for veins, but caution is advised for arterial measurements.
Area of Science:
- Ophthalmology
- Medical Imaging
- Retinal Imaging Analysis
Background:
- Accurate measurement of retinal vascular diameters is crucial for diagnosing and monitoring various ocular and systemic diseases.
- Standard fundus photography and advanced adaptive optics (AO) imaging offer different resolutions and capabilities for retinal visualization.
Purpose of the Study:
- To prospectively compare retinal vascular diameter measurements obtained from standard fundus images versus adaptive optics (AO) images.
- To evaluate the performance of different software tools (VAMPIRE, IVAN, AO_Detect_Artery™) in measuring these diameters across imaging modalities.
Main Methods:
- Retinal images from twenty healthy subjects were acquired using a Canon fundus camera and an Imagine Eyes rtx1 AO camera.
- Vascular diameters were measured using the VAMPIRE annotation tool, IVAN for funduscopic images, and AO_Detect_Artery™ for AO images.
Main Results:
- For arterial diameters, IVAN showed significant correlation (r=0.79) but a mean difference of 9.1 µm with AO_Detect_Artery™ (p=0.005). VAMPIRE had poor correlation (r=0.12) with AO measurements.
- For venous diameters, IVAN demonstrated high correlation (r=0.83) with a mean difference of 3.9 µm (p=0.35). VAMPIRE showed moderate correlation (r=0.41).
- VAMPIRE is accurate for manual measurement but shows poor correlation with AO. IVAN, while having slightly larger differences, shows stronger correlations, though it may underestimate arterial diameters.
Conclusions:
- Semi-automatic software like IVAN offers stronger correlations with adaptive optics imaging for retinal vein diameter measurements compared to manual tools.
- Arterial diameter measurements using IVAN should be interpreted with caution due to potential underestimation when compared to AO imaging.
- The choice of imaging modality and measurement software impacts the reliability and comparability of retinal vascular diameter assessments.
