Semi-automated algorithm using directional filter for the precise quantification of non-perfusion area on widefield swept-source optical coherence tomography angiograms.

ABSTRACT

Background: The clinical application of optical coherence tomography angiography (OCTA) has been well documented in literature with its promising potential in dye-less evaluation of various retinal vascular pathologies. Recent advances in OCTA help us gather wider field of view with 12 mm × 12 mm and montage compared to the standard dye-based scans, which has a higher accuracy and sensitivity in detection of peripheral pathologies. The aim of this study is to build a semi-automated algorithm to precisely quantify the non-perfusion areas (NPAs) on widefield swept-source optical coherence tomography angiography (WF SS-OCTA). Methods: All subjects underwent imaging on 100 kHz SS-OCTA device acquiring 12 mm × 12 mm angiograms centered on fovea and optic disc. After a comprehensive literature review, a novel algorithm using FIJI (ImageJ) was designed to calculate the NPAs (mm2) after excluding the threshold and segmentation artifact areas from the total field of view. Segmentation and threshold artifacts were first removed from enface structure images using the spatial variance and mean filter respectively. Vessel enhancement was achieved by using 'Subtract Background' followed by directional filter. The cut off for Huang's fuzzy black and white thresholding was defined from the pixel values based of the foveal avascular zone. Then, the NPAs were calculated using the 'Analyze Particles' command with a minimum size of ~0.15 mm2. Finally, the artifact area was subtracted from to give the corrected NPAs. Results: Our cohort had 44 eyes of 30 control patients and 107 eyes of 73 patients with diabetes mellitus (both median age 55 years, P=0.89). Of 107 eyes, 21 eyes had no evidence of diabetic retinopathy (DR), 50 eyes had non-proliferative DR and 36 eyes had proliferative DR. The median NPA was 0.20 (0.07-0.40) in controls, 0.28 (0.12-0.72) in no DR, 5.54 (3.12-9.10) in non-proliferative DR and 13.38 (8.73-26.32) in proliferative DR eyes. Using mixed effects-multiple linear regression analysis adjusting for age, there was significant progressive increase in NPA with increasing DR severity. Conclusions: This is one of the first study to use the directional filter for WFSS-OCTA image processing which is known to be superior to other Hessian based multiscale, linear, and non-linear filters especially for vascular analysis. Our method could greatly refine and streamline the calculation of signal void area proportion, while being much quicker and accurate than manual delineation of NPAs and subsequent estimation. This combined with the wide field of view can have a great prognostic and diagnostic clinical impact for future applications in DR and other ischemic retinal pathologies.