Abstract
Purpose: :
The purpose of the study was to test the reliability of vessel diameter measurements with a newly developed retinal oximeter.
Methods: :
Twelve healthy individuals participated in the study. Retinal images were taken with the Oxymap Retinal Oximeter (Oxymap ehf. Reykjavik, Iceland). Diameters of retinal vessels were measured automatically with the Oxymap Analyzer software. Repeated measurements on the same vessel segments were compared. The automatic measurements were also compared with semi-automatic measurements with an add-on to the ImageJ software http://rsbweb.nih.gov/ij/
Results: :
Variance coefficient (standard deviation / mean) from the repeatability test was 2.8% and 4.0% for first and second degree venules and 3.5% and 5.4% for first and second degree arterioles, respectively. ImageJ measured larger mean diameters than Oxymap Analyzer in all cases and the differences were 5.1±2.2, 2.9±1.3 and 2.7±1.6 pixels for first, second and third degree venules and 3.1±1.2, 2.7±0.9 and 2.9±1.4 pixels for first, second and third degree arterioles (mean ±standard deviation, n=12). The difference between the programs was significant (p<0.05 except for third degree venules).
Conclusions: :
Vessel diameter measurements with the oximeter are repeatable and comparison with an established method demonstrates a relatively stable offset where the standard deviation of the difference is rather small. Different definitions of vessel borders may be the reason for this difference.
Keywords: blood supply • imaging/image analysis: non-clinical • ischemia