Abstract
Purpose :
Optical coherence tomography angiography (OCTA) is a useful non-invasive imaging modality for visualizing the choriocapillaris (CC) and is increasingly being used in the diagnostic work up of uveitis patients. The purpose of this study is to quantify CC flow deficits (FDs) among posterior or panuveitis patients with choroidal involvement and to compare the flow deficits with hypocyanescent areas on indocyanine green angiography (ICGA).
Methods :
A retrospective review of OCTA images from patients with either posterior or panuveitis with choroidal involvement was conducted. Images were acquired on the spectral-domain Zeiss Cirrus 5000 platform and automated segmentation performed. CC slabs were analyzed using an investigator derived algorithm. Briefly, a binarized image of the superficial capillary plexus was subtracted from the CC image to reduce superficial retinal vessel projection artifacts. A phansalkar auto local thresholding technique was applied to generate binarized CC FD maps. Flow deficit number (FDN) and flow deficit density (FDD) were then calculated used the “Analyze Particles” function in ImageJ. Additionally, using the gray-level co-occurrence matrix (GLCM) function, we performed feature extraction for further characterization. Secondly, we compared the corresponding FDs on CC OCTA images with hypocyanescent areas on ICGA and calculated the degree of overlap of the two images using the Jaccard index.
Results :
Fifteen eyes of 9 patients were included in our analysis. Our cohort included patients with PIC (n=4), VKH (n=1), BCR (n=3), and granulomatous panuveitis (n=1). The mean age of patients was 46.33 ± 13.2 years. On 3 x 3 mm scans, the mean FDN was 2159 ± 257.9 and FDD (%) was 31.86 ± 3.4. FDD (%) had a moderate-high direct correlation with GLCM textural parameters, cluster shade (r = 0.70) and correlation (r = 0.60). Six eyes from our cohort were used for ICGA and OCTA comparison. The average Jaccard index was 0.548 ± 0.1. Of note, several ICGA images revealed larger FDs than that apparent on OCTA.
Conclusions :
Quantification of CC FDs may be useful in the management of uveitis patients. The direct correlation between FDD (%) and the GLCM textural parameters, cluster shade and correlation, suggest an increasing severity of texture disruption with increasing FDD (%) Further work is needed to better understand the role of texture analysis in uveitis.
This is a 2021 Imaging in the Eye Conference abstract.