“…Other authors, such as Díaz et al [ 30 ], also studied the automated segmentation of the FAZ in healthy and diabetic patients and their comparison with manual segmentation, obtaining poor results in the FAZ that were irregular and did not follow the acircularity, especially in the 3 × 3 OCTA protocol. Although many studies are available regarding different image processing techniques for automatic segmentation of the FAZ in various retinal imaging modalities [ 30 , 31 , 32 , 33 , 34 , 35 , 36 ], studies focusing on automatic FAZ segmentation in OCTA were usually conducted on healthy subjects. In addition, a few studies assessing the accuracy of FAZ delineation in OCTA images of diabetic eyes have obtained poor results due to the high incidence of signal noise and artefacts in OCTA imaging of diabetic patients [ 7 , 8 , 9 , 10 , 11 , 12 , 13 , 14 , 15 , 16 , 17 , 18 , 19 , 20 , 21 , 22 , 23 , 24 , 25 , 26 , 27 , 28 , 29 , 30 , 31 , 32 , 33 , 34 , 35 , 36 , 37 , 38 , 39 ].…”