Abstract
Abstract: :
Purpose: To analyze and compare laser Doppler flowmetry (LDF) signals obtained from light backscattered in the subfoveal choroid and collected either at the site of illumination (S) or apart from it (M). Methods: We used a simple laser Doppler fundus device detecting either only single (S) or multiple (M) backscattered light. For this purpose, we developed a new acquisition system based on Labview, which allows control of the analysis of the signal induced by light beating. 23 subjects (mean age 31±;10 years) were asked to look with their left eye at a faint light source for 30 seconds of measurement during which heart beat from the ear lobe or finger was recorded. After 1 minute, the same measurement was repeated. Both consecutives measurements were done first for M detection then for S detection. Zero and first moment of the power spectrum gives choroidal blood volume (ChBVol, arbitrary units), velocity (ChBVel, kHz) and flow ChBF = ChBVel x ChBVol. These parameters are usually used as simple descriptor for the power spectrum. The second moment (ChB2nd), usually not considered in the LDF technique, was taken into account. These parameters were average over the pulse phase of the heart and over the whole experiment. Power spectrum curves were compared between subjects, between detection methods and with the theory of Bonner & Nossal. Results: Based on Pearson's correlation coefficient, ChBVel, ChBVol are normally distributed but not ChB2nd. Sensitivities with N–detection were 7.5%, 6.1%, 9.6% for resp. ChBVel, ChBVol and ChB2nd and for S–detection 70%, 4% and 15%. None of the parameters has significant correlation with age, gender and smoker or non–smoker. M–ChBVel (4.9±;0.9kHz) is 2 times larger than S–ChBVel (2.6±;0.4kHz) (p<0.0001). M–signal is 4 times larger than S–signal (p<0.0001), but S–ChBVol is 40 times smaller than M–ChBVol (p<0.01). ChB2nd do not correlates neither with pulsatility of ChBVel nor with ChBVol. N– and S–Power spectrum curves fall much faster than an exponential as expected by Bonner & Nossal. Difference between parameters obtained, either from the mean spectrum or from a mean spectrum de–noised with wavelets, were smaller than 0.1%. Conclusions: In our normal population, age and gender are not discriminate by blood flow parameters. As expected, M–ChBvel is larger than S–ChBvel, due to multiple scattering. Detecting backscattered light around the illumination point (M) gives a higher signal and has a better sensitivity than detecting light at the site of illumination (S).
Keywords: choroid • retina • optical properties