A Computational Model of Peripheral Photocoagulation for the Prevention of Progressive Diabetic Capillary Occlusion

Figure 3

Evaluation of dot pattern and band pattern in terms of coagulated retinal area and predicted ischemic area. Replicate simulations were executed for dot and band patterns, divided into four groups with , 2, 4, and 6. Note that represents the control group without photocoagulation whereas other ’s indicate the total number of bands and therefor increasing areas of photocoagulation. For a given , the normalized burn areas are essentially equal for the dot and band patterns. is plotted on the -axis, while the fraction of retina which is ischemic, the normalized area at the end of three simulated years, is plotted on the -axis for the dot (a) and band (c) patterns. Each of these panels also includes standard PRP as a green square. Outcomes from replicate simulations are given as box and whisker plots. (b) and (d) plot equals number of bands or of dot burns of equal total area on the -axis and plot the normalized sum of treated and final ischemic areas on the -axis. Estimated PRP outcome is given based on a 0.28 normalized burn area for the total peripheral retinal [2] plus an estimated ischemic propagation for nontreated retina based on the control case.

(a)

(b)

(c)

(d)