Fig 1.
Calculation of vessel tortuosity.
3x3 mm OCTA scans segmented to show the SCP (inner boundary at the ILM and the outer boundary 10 μm above the IPL). The actual and Euclidean lengths are measured for every individual branch and vessel tortuosity is calculated as the ratio between the sum of the actual lengths of all branches and the sum of their Euclidean distances.
Table 1.
Demographic and disease-related patient characteristics.
Fig 2.
Box plot representation of the perifoveal vessel tortuosity results in the five study groups.
No DR: No diabetic retinopathy, NPDR: Non-proliferative diabetic retinopathy, PDR: Proliferative diabetic retinopathy, PRP: Panretinal photocoagulation, PDR+PRP: Proliferative diabetic retinopathy receiving panretinal photocoagulation. (p<0.05,*significant vs Healthy controls, † significant vs PDR+PRP).
Table 2.
Post-hoc analysis of vessel tortuosity and fractal dimension results comparing the five groups.
Fig 3.
Box plot representation of the fractal dimension results in the 5 study groups.
No DR: No diabetic retinopathy, NPDR: Non-proliferative diabetic retinopathy, PDR: Proliferative diabetic retinopathy, PRP: Panretinal photocoagulation, PDR+PRP: Proliferative diabetic retinopathy receiving panretinal photocoagulation. (p<0.05,*significant vs Healthy controls,†significant vs no clinical DR, ● = significant vs NPDR, ɸ significant vs PDR).