Abstract
Purpose: :
To detect metabolic changes before morphologic alterations are visible. Such results can be used for patient-specific treatment and control of therapy.
Methods: :
FLIM was measured in 62 diabetics (mean 62.5±14years) having no signs of diabetic retinopathy. As control, 33 healthy subjects (mean 56±20years) were used. FLIM measurements were performed by the Jena Lifetime Laser Scanner Ophthalmoscope. Laser pulses (70ps duration, 80MHz repetition rate, 0.1mW) excited FLIM in a 30degree field. FLIM was detected by time-correlated single photon counting in 2 spectral channels (CH1=490-560nm, CH2=560-700nm). The decay of fluorescence intensity was 3-exponentially fitted. As result, histograms of amplitudes Ai, lifetimes Ti, and relative contribution Qi=Ai•Ti were calculated. Statistical comparison was done by Holm-Bonferroni method. Here, the range of the fitting parameters is divided in n intervals. A significant difference exists between distinctive values of fitting parameters of diabetics and healthy subjects, if the error probability, calculated by Wilcoxon test, is lower than a given error probability e.g. 5%, divided by the number n of intervals.
Results: :
Most sensitive for discrimination of diabetics and healthy subjects are measurements in CH1. In A1 interval (50-100%, width 1%) significant differences were found for 14 values (73 - 92%), for A2 (5-30%, width 1%) 4 values (5 - 14%), and for A3 (0-18%, 0.5% width) 21 values (1 - 17%). The lifetime T1 (30-120ps, width 10 ps) was significantly different only for 50 and 90 ps, for T2 (300-900ps, width 10 ps) at 640, 650, 660, and 670ps, and for T3 (2.3-7ns, 50 ps width) at 16 values (2,4 - 3.1ns). In Ch2, significant differences were found for A1 6 values (63 - 83%), for A2 at 25, 27, 29%, for A3 12 values (1- 8.5%). Significant differences were found for T1 at 100ps, no for T2, and for T3 at 3.1, 3.15, 3.2ns.
Conclusions: :
Differences in amplitudes and lifetimes of autofluorescence, predominantly in CH1, are probably caused by accumulation of advanced glycation end-products and by increased contribution of protein-bound NADH.
Keywords: imaging methods (CT, FA, ICG, MRI, OCT, RTA, SLO, ultrasound) • diabetic retinopathy • discrimination