Projection Artifact Removal Improves Visualization and Quantitation of Macular Neovascularization Imaged by Optical Coherence Tomography Angiography

Purpose: This study assesses the frequency of projection artifacts in optical coherence tomography angiography (OCTA) en face images and compares images before and after applying a 3-dimensional projection artifact removal (3D-PAR) algorithm. Methods: This is a single-center, retrospective study that included consecutive patients with any underlying diagnosis who had OCTA obtained from January to March 2017. Patients with various retinal diseases and also healthy eyes were included. All participants underwent imaging with a scan area of 3 mm × 3 mm. The 4 default en face slabs were analyzed: superficial capillary plexus (SCP), deep capillary plexus (DCP), outer retina (OR), and choriocapillaris (CC). Images were qualitatively analyzed before and after 3D-PAR by 2 independent graders. Results: None of the SCP images had projection artifact before or after 3D-PAR. Scans of the DCP presented projection artifact in 96.5% of the cases. After 3D-PAR, 14.7% had a complete improvement of projection artifact, 56.5% had a partial improvement, 14.1% were worse, and 14.7% presented no change. In the OR, 2.9% had projection artifact, with a complete improvement after 3D-PAR in 40%, partial improvement in 20%, and no change in 40%. Projection artifact was initially present in 97.6% of the images in the CC. After 3D-PAR, there was a complete improvement in 72.9%, partial improvement in 26.5%, and no change in 0.6%. Choroidal neovascularization (CNV) was detected in 29 eyes (17.1%), and 3D-PAR improved detection of CNV in 12 cases (41.4%). Conclusions: OCTA with 3D-PAR technology minimizes the appearance of projection artifacts in the DCP and CC slabs.

Download Full-text

Motion artifact removal of optical coherence tomography angiography based on tensor voting

Optical Coherence Tomography and Coherence Domain Optical Methods in Biomedicine XXIII ◽

10.1117/12.2510559 ◽

2019 ◽

Author(s):

Ang Li ◽

Congwu Du ◽

Yingtian Pan

Keyword(s):

Optical Coherence Tomography ◽

Optical Coherence Tomography Angiography ◽

Motion Artifact ◽

Optical Coherence ◽

Tensor Voting ◽

Artifact Removal

Download Full-text

Analysis of Foveal Microvascular Abnormalities in Diabetic Retinopathy Using Optical Coherence Tomography Angiography with Projection Artifact Removal

Journal of Ophthalmology ◽

10.1155/2018/3926745 ◽

2018 ◽

Vol 2018 ◽

pp. 1-9 ◽

Cited By ~ 8

Author(s):

Lun Liu ◽

Jian Gao ◽

Weili Bao ◽

Chengyang Hu ◽

Yajing Xu ◽

...

Keyword(s):

Optical Coherence Tomography ◽

Diabetic Retinopathy ◽

Optical Coherence Tomography Angiography ◽

Vessel Density ◽

Foveal Avascular Zone ◽

Optical Coherence ◽

Artifact Removal ◽

Significant Difference ◽

Patients With Diabetes ◽

Capillary Plexus

Purpose. To analyze foveal microvascular abnormalities in different stages of diabetic retinopathy (DR) using optical coherence tomography angiography (OCTA) with projection artifact removal (PAR). Methods. We analyzed 93 eyes of 59 patients with diabetes—31 with no DR (no DR), 34 with mild to moderate nonproliferative DR (mild DR), and 28 with severe nonproliferative DR to proliferative DR (severe DR)—and 31 age-matched healthy controls. Sections measuring 3 × 3 mm2 centered on the fovea were obtained using OCTA. The area, perimeter, and acircularity index (AI) of the foveal avascular zone (FAZ), vessel density within a 300 μm wide region of the FAZ (FD-300), and parafoveal vessel density in the superficial capillary plexus (SCP) and deep capillary plexus (DCP) were calculated using novel built-in software with PAR. Results. There was no statistically significant difference in the FAZ area (p=0.162). There was a statistically significant difference in the FAZ perimeter (p=0.010) and the AI (p<0.001) between the four groups. There was a correlation between the AI and the increasing severity of DR (p=0.010). Statistically significant decreases of vessel density in the FD-300, SCP, and DCP were observed (all p<0.001). There was a difference in parafoveal vessel density in the DCP between the healthy control eyes and the eyes with diabetes without DR (p=0.027). There was a significant correlation between vessel density and increasing severity of DR (p<0.001). Conclusion. Compared with the FAZ area, AI allows a more helpful quantitative assessment of the changes in the FAZ. Vessel density determined using OCTA with PAR might be a useful parameter indicating the progression of DR. Parafoveal vessel density in the DCP after PAR might be a potential early biomarker of DR before appearance of clinically evident retinopathy and needs further investigation.

Download Full-text