Normal distribution of cardiac variation in retinal image measurement

Cardiac variation has a great impact on retinal vessel calibers, but it has been ignored in previous retinal caliber related studies. This is assuming that retinal vessel calibers follows normal distribution with random photography and measurement, while it didn't consider the cardiac variation and no evidences have proved it to date. This research aims to clarify this fundamental question. The results will improve descriptive statistics and statistical significance tests of retinal image analysis, as well as the current retinal photography method and related clinical applications. Retinal vessel caliber measurement has been used in screening or diagnosis of a disease by the comparison of images between two groups or two images from the same patient during a diagnostic duration. Cardiac cycle has impacts on retinal vessel regulation and there is cardiac variation in the retinal vessel caliber measurement. The current retinal photography method is to photograph retinal image randomly without the consideration of cardiac variation. In this research, we designed an advanced electrocardiogram synchronized retinal photography technique to capture nine retinal images in cardiac cycle. Our results provide graphical and statistical evidences that cardiac variations in the retinal vessel caliber measurements are normally distributed. However, a countable standard deviation exists in cardiac variation, which should be considered and verified in the future retinal caliber related studies and analysis.