Abstract
Purpose: :
Retinal microvascular changes, including reduced AVR, predict incident cardiovascular disease [1]. Such observations rely on accurate measurement of both arterial and venous diameters. Retinal vascular diameters are known to change through the cardiac cycle [2] but the relationship between arterial and venous diameters has not been examined within the same cohort.
Methods: :
We have developed a system to synchronise capture of red–free digital retinal images (Zeiss FF450+ fundus camera) with the cardiac cycle, at an operator selected delay following the ECG R–wave. A commercial 3–lead ECG monitor (Lifepulse 10) is used to detect QRS, and a dedicated micro–controller with purpose written firmware controls the fundus camera and CCD imager (Basler A101) to capture the image at the desired point with negligible latency. Images are downloaded to a PC for subsequent analysis. ECG synchronised retinal images were taken from 15 normotensive subjects (age 25–52 years; 8 female) at 100ms intervals through the cardiac cycle. Two images were captured at each interval using a disc centred 20º field of view. Resting blood pressure was measured by an automated device (Omron HEM–705–CP). Arterial and venous vessel segments of length 1/6th of optic disc diameter were selected for analysis, typically a similar distance distal to the disc boundary. Automatic image registration was used to identify identical vessel segments throughout each image sequence, and vessel diameters were measured using the Sliding Linear Regression Filter (SLRF) technique [3]. The mean diameter change at each point in the cycle was derived as a fraction of diameter at R–wave.
Results: :
Arterial diameters increased by 2.8±0.6 %, peaking at 300ms following R–wave, and venous diameters by 1.9±0.5 %, peaking 300ms later. No association was found between excursion in vessel diameter and blood pressure. Age correlated with venous (r=0.73, p=0.01) but not arterial diameter excursion.
Conclusions: :
ECG synchronisation of retinal image capture may reduce the scatter in measurement of vessel diameters. Arterial diameters appear to be most stable 400 to 600 ms following the R–wave, whereas arterial to venous diameter ratios appear more stable 600 to 700 ms after the R–wave.
References: :
[1] Wong TY, Klein R et al. Ophthalmology 2003; 110(5):933–940.
[2] Dumsky MJ, Kohner EM et al. Curr Eye Res. 1996; 15(6):625–32.
[3] Chapman N, Witt N et al. Br J Ophthalmol 2001; 85(1):74–9.
Keywords: clinical research methodology