No abstract

Purpose: To propose a new clinical evaluation index, foveal avascular zone (FAZ) volume, and analyze its statistical significance. Methods: A semiautomatic method is proposed to measure the FAZ volume in optical coherence tomography angiography images as follows: The region of interest was flattened and annotated axially. The labeled pixels in the restored region of interest were counted as the FAZ volume. Linear regression and the independent samples t-test were performed for the statistical analysis. Results: Sixty-one normal, 64 high myopia, and 42 diabetic retinopathy eyes were imaged using optical coherence tomography angiography. For normal eyes, the FAZ volume correlates inversely with central macular thickness (superficial: P = 0.004; deep: P < 0.001) and positively with area (P < 0.001). For high myopia eyes, the deep plexus FAZ (P = 0.34) and total FAZ (P = 0.38) volumes show no significant difference, whereas the superficial plexus FAZ volume is significantly larger than control (P < 0.001). For diabetic retinopathy eyes, the superficial plexus FAZ (P = 0.001), deep plexus FAZ (P = 0.014), and total volumes (P = 0.002) are significantly larger than control. Conclusion: The FAZ volume is proposed for depicting the 3D structure of the FAZ. It shows greater sensitivity for vascular alteration that makes it meaningful for clinical analysis.

The foveal avascular zone (FAZ) is sensitive to retinal pathological process in the macular fovea area. For the purpose of efficient FAZ 3D quantification, we firstly propose a priors-guided convolutional neural network (CNN) to provide a tailor-made solution for 3D FAZ segmentation for optical coherence tomography angiography (OCTA) images. Location and topology priors are taken into account. The random central crop module is utilized to restrict the region to be processed, while the non-local attention gates are contained in the network to capture long-range dependency. The topological consistency constraint is calculated on maximum and mean projection maps through persistent homology to keep topological correctness of the model’s prediction. Our method was evaluated on two OCTA datasets with 478 eyes and the experimental results demonstrate that our method can not only alleviate the over-segmentation prominently but also fit better on the contour of FAZ region.