June 2023
Volume 12, Issue 6
Open Access
Retina  |   June 2023
The Impact of Cataracts on the Measurement of Macular Choriocapillaris Flow Deficits Using Swept-Source OCT Angiography
Author Affiliations & Notes
  • Jianqing Li
    Department of Ophthalmology, Bascom Palmer Eye Institute, University of Miami Miller School of Medicine, Miami, FL, USA
    Department of Ophthalmology, First Affiliated Hospital of Soochow University, Suzhou, Jiangsu, China
  • Mengxi Shen
    Department of Ophthalmology, Bascom Palmer Eye Institute, University of Miami Miller School of Medicine, Miami, FL, USA
  • Yuxuan Cheng
    Department of Bioengineering, University of Washington, Seattle, WA, USA
  • Qinqin Zhang
    Department of Bioengineering, University of Washington, Seattle, WA, USA
    Research and Development, Carl Zeiss Meditec, Inc., Dublin, CA, USA
  • Jeremy Liu
    Department of Ophthalmology, Bascom Palmer Eye Institute, University of Miami Miller School of Medicine, Miami, FL, USA
  • Luis de Sisternes
    Research and Development, Carl Zeiss Meditec, Inc., Dublin, CA, USA
  • Warren H. Lewis
    Bayside Photonics, Inc., Yellow Springs, OH, USA
  • Ruikang K. Wang
    Department of Bioengineering, University of Washington, Seattle, WA, USA
    Department of Ophthalmology, University of Washington, Seattle, WA, USA
  • Giovanni Gregori
    Department of Ophthalmology, Bascom Palmer Eye Institute, University of Miami Miller School of Medicine, Miami, FL, USA
  • Philip J. Rosenfeld
    Department of Ophthalmology, Bascom Palmer Eye Institute, University of Miami Miller School of Medicine, Miami, FL, USA
  • Correspondence: Philip J. Rosenfeld, Department of Ophthalmology, Bascom Palmer Eye Institute, University of Miami Miller School of Medicine, 900 Northwest 17th Street, Miami, FL 33136, USA. e-mail: prosenfeld@miami.edu 
Translational Vision Science & Technology June 2023, Vol.12, 7. doi:https://doi.org/10.1167/tvst.12.6.7
  • Views
  • PDF
  • Share
  • Tools
    • Alerts
      ×
      This feature is available to authenticated users only.
      Sign In or Create an Account ×
    • Get Citation

      Jianqing Li, Mengxi Shen, Yuxuan Cheng, Qinqin Zhang, Jeremy Liu, Luis de Sisternes, Warren H. Lewis, Ruikang K. Wang, Giovanni Gregori, Philip J. Rosenfeld; The Impact of Cataracts on the Measurement of Macular Choriocapillaris Flow Deficits Using Swept-Source OCT Angiography. Trans. Vis. Sci. Tech. 2023;12(6):7. https://doi.org/10.1167/tvst.12.6.7.

      Download citation file:


      © ARVO (1962-2015); The Authors (2016-present)

      ×
  • Supplements
Abstract

Purpose: The impact of cataracts on the measurement of macular choriocapillaris flow deficits (CC FDs) was assessed by comparing the quantitative results before and after cataract surgery using an image quality algorithm developed for swept-source optical coherence tomography angiography (SS-OCTA) scans and a validated strategy for quantifying the CC FDs.

Methods: SS-OCTA image quality scores and CC FDs measurements within the fovea-centered 1-mm, 3-mm, and 5-mm diameter circles were compared before and after cataract surgery. CC FDs changes in a modified Early Treatment Diabetic Retinopathy Study (ETDRS) grid were further investigated.

Results: Twenty-four eyes were studied. Overall image quality in all three circles was observed to improve significantly following the removal of cataracts (all P < 0.05). Although there was good repeatability in the measurements of CC FDs at both visits (intraclass correlation coefficients were over 0.95), significant decreases in CC FD measurements were observed after surgery within the 1-mm circle (P < 0.001) and the 3-mm circle (P = 0.011), but no changes were observed within the 5-mm circle (P = 0.509) or any of the quadrant sectors of the modified ETDRS grid (all P > 0.05).

Conclusions: The presence of cataracts resulted in worse image quality and increased CC FD measurements within the fovea-centered 1-mm and 3-mm circles, with the 1-mm circle being impacted the most.

Translational Relevance: The impaired detection of CC perfusion deficits within the central macula of cataract eyes needs to be appreciated when imaging the CC in phakic eyes, especially in clinical trials.

Introduction
The choriocapillaris (CC) is the innermost layer of the choroid, consisting of a capillary monolayer that represents the terminal and only capillaries of the choroid and is responsible for nourishing the retinal pigment epithelium and photoreceptors.1 Swept-source optical coherence tomography angiography (SS-OCTA) has been used to image the CC, and algorithms to quantify choriocapillaris flow deficits (CC FDs) that exceed the average normal intercapillary distance of 24 µm have been validated.2,3 
The quantitation of CC FDs should be interpreted with caution, as the detection of these FDs can be affected by factors such as the choice of thresholding methods,4,5 image artifacts,6 scan tilt,7 and image quality.8 In the study by Lee et al.,8 neutral density filters were positioned between the instrument and the eyes to simulate the impact of media opacities, which caused uniform signal reductions across the scan field. They found that the decrease in the OCTA signal strength led to increased CC FDs. Although neutral density filters approximated naturally occurring media opacities such as cataracts, the inhomogeneous nature of cataracts and the different types of cataracts would have varying degrees of signal reduction in different regions of the scan area and might cause a variable impact on the quantitation of CC FDs. In addition, Lee et al.8 chose a Phansalkar local thresholding method using a 15-pixel radius, which has been shown to be inappropriate.4,5 The selection of the best thresholding strategy is even more important in the presence of neutral density filters and cataracts due to the expected degradation of signal strength and the increase in noise within a scan. 
In our study, we prospectively enrolled patients with cataracts and investigated the impact of cataracts on image quality and our ability to quantitate CC FDs by comparing the quantitative results before and after cataract surgery using a novel OCTA quality map algorithm and a validated strategy for quantifying the CC. 
Methods
The Comparison of Low Luminance Visual Acuity Testing Before and After Cataract Surgery (COMPARE) study was a prospective study approved by the institutional review board of the University of Miami Miller School of Medicine, and all participants signed an informed consent for this prospective study. The study was performed in accordance with the tenets of the Declaration of Helsinki and complied with the Health Insurance Portability and Accountability Act of 1996. 
Part 1 of the COMPARE study enrolled patients with cataracts who were scheduled to undergo cataract surgery. Previously, we reported on the impact of cataracts on photopic luminance best-corrected visual acuity (BCVA), low luminance BCVA, and low luminance visual acuity deficit measurements in this study.9 Exclusion criteria included confounding ocular conditions such as an axial length ≤ 23 mm or ≥ 26 mm, glaucoma, and any evidence of macular atrophy or hyperpigmentation that occupied more than 20% of the fovea-centered 1-mm, 3-mm, or 5-mm diameter circles. Macular atrophy was defined by the presence of any persistent choroidal hypertransmission defects with a greatest linear dimension ≥ 250 µm, which was determined using a 6 × 6-mm SS-OCTA scan (PLEX Elite 9000; Carl Zeiss Meditec, Dublin, CA).1012 Hyperpigmentation was defined by the presence of any choroidal hypotransmission defects as previously described using the same 6 × 6-mm SS-OCTA scan pattern.13 
Cataract patients were identified and underwent SS-OCTA imaging by the same photographer before and after cataract surgery. All subjects were instructed to blink prior to image acquisition, and the use of artificial tears was left to the discretion of the operator. Four repeated 6 × 6-mm scans were acquired for each study eye at each visit. All scans were centered on the fovea. The angiographic flow information was generated using the previously published and validated complex optical microangiographic (OMAGC) algorithm.1416 Scans with signal strength less than 7 based on the output of the instrument and scans with significant motion artifacts were excluded. 
The proprietary OCTA quality map algorithm used in this study was provided by the manufacturer of the device (Carl Zeiss Meditec; https://arinetworkhub.com/). In brief, given an input OCTA volume, this algorithm provides an en face map measuring the quality of the OCTA signal at each A-scan location in a quantitative manner from 1 (lowest quality) to 5 (highest quality) in value (as illustrated in Fig. 1). Two-dimensional images are generated describing different textural features for each A-scan location, such as intensity, contrast, and sharpness. Quality is measured based on the analysis of these textural features extracted from the flow and structural slabs obtained from the OCTA volume. Four different en face slabs (one for maximum flow information and three for average, maximum, and minimum intensity properties, respectively) were generated from each scan volume, restricting the en face projection of the retina from the internal limiting membrane to the retinal pigment epithelium and processed to generate 88 Haralick feature maps (22 maps from each independent slab). Each of these feature maps provides a pixel-by-pixel representation of the statistical and textural properties of the analyzed image within a small neighborhood, and they are commonly used in the machine learning literature.17 Because the extracted features describe intrinsic characteristics of the images, no resizing was done for any of the slabs or feature maps, keeping the original size of 500 × 500 pixels for a 6 × 6-mm en face field of view. A machine learning algorithm was then trained to use these features to match quality maps drawn by expert annotators in a pixel-by-pixel manner. Training of the algorithm was conducted in a set of images outlined by graders in terms of quality (5-scale of increasing quality gradings, which were drawn directly pixel-by-pixel in the en face images) where five independent graders evaluated the same given images. All graders received the necessary training with specific instructions to judge the quality of given OCTA regions using the 5-scale. 
Figure 1.
 
Example of increased image quality and decreased CC FD measurements after cataract surgery. (A–E) The top row shows image quality and quantitation of CC FDs from a scan before cataract surgery. (F–J) The bottom row shows image quality and CC FDs from a scan of the same subject from after surgery. (A, F) Retinal en face angiographic images. (B, G) Retinal en face structural images. (C, H) Quality maps produced by the algorithm with different colors indicating different levels of image quality. Different quality levels are shown in the scale beneath panel H, with 5 (red) being the highest image quality (scaled 1–5). The 1-mm, 3-mm, and 5-mm circles centered on the fovea are shown. Image quality scores before surgery (C) were 2.33, 2.06, and 2.16 in the 1-mm, 3-mm, and 5-mm circles respectively. After surgery (H), image quality scores were 4.82, 4.86, and 4.64 respectively in the 1-mm, 3-mm, and 5-mm circles, respectively. (D, I) CC en face structural images produced by the CC algorithm. (E, J) Binarized CC flow images with white regions denoting the CC FDs. Circles depict the 1-mm, 3-mm, and 5-mm circles centered on the fovea. (E) Before cataract surgery, the CC FD% measurements were 20.12%, 12.27%, and 9.50% in the central 1-mm, 3-mm, and 5-mm circles, respectively. (J) After cataract surgery, the CC FD% measurements decreased to 16.62%, 11.66%, and 8.95% in the 1-mm, 3-mm, and 5-mm circles, respectively.
Figure 1.
 
Example of increased image quality and decreased CC FD measurements after cataract surgery. (A–E) The top row shows image quality and quantitation of CC FDs from a scan before cataract surgery. (F–J) The bottom row shows image quality and CC FDs from a scan of the same subject from after surgery. (A, F) Retinal en face angiographic images. (B, G) Retinal en face structural images. (C, H) Quality maps produced by the algorithm with different colors indicating different levels of image quality. Different quality levels are shown in the scale beneath panel H, with 5 (red) being the highest image quality (scaled 1–5). The 1-mm, 3-mm, and 5-mm circles centered on the fovea are shown. Image quality scores before surgery (C) were 2.33, 2.06, and 2.16 in the 1-mm, 3-mm, and 5-mm circles respectively. After surgery (H), image quality scores were 4.82, 4.86, and 4.64 respectively in the 1-mm, 3-mm, and 5-mm circles, respectively. (D, I) CC en face structural images produced by the CC algorithm. (E, J) Binarized CC flow images with white regions denoting the CC FDs. Circles depict the 1-mm, 3-mm, and 5-mm circles centered on the fovea. (E) Before cataract surgery, the CC FD% measurements were 20.12%, 12.27%, and 9.50% in the central 1-mm, 3-mm, and 5-mm circles, respectively. (J) After cataract surgery, the CC FD% measurements decreased to 16.62%, 11.66%, and 8.95% in the 1-mm, 3-mm, and 5-mm circles, respectively.
In the expert quality annotations, poor image quality was generally due to a combination of four factors: (1) poor signal level, (2) high noise, (3) poor resolution, and (4) significant motion artifacts. The ground truth training target for a supervised machine learning model was established as the average quality map resulting from these five graders. Least absolute shrinkage and selection operator (LASSO) regression (a linear regression model where each feature map receives a weight to account for computed quality) was then used to generate a linear model matching the feature map values to the ground truth drawn by the graders in a pixel-by-pixel manner.18 Although further details about this method remain undisclosed because this OCTA quality map algorithm is part of the software present in the latest commercial release of the instrument, its performance has passed the quality standards established by the manufacturer of the device when tested in a set of independent images from the training set against pixel-by-pixel outlined quality annotations by a different set of graders. This quality map is used to judge the quality of an individual scan across the field of view, to quantify the difference in quality among several acquisitions of the same subject, and to provide an overall quality metric of a single acquisition by averaging the map values in a region of interest. 
Quantitation of the CC FDs was performed using the four repeated 6 × 6-mm scans. The CC measurements were obtained from a 16-µm-thick slab with its anterior boundary located 4 µm beneath Bruch's membrane (BM).3 Areas of hyperpigmentation were excluded using the mask that was manually outlined based on en face structural images from a slab located 64 to 400 µm beneath BM.13 The retina and CC en face structural slabs were used to compensate for any signal loss due to the overlying anatomy, and retinal vessel projection artifacts were removed from the CC flow slab.19 The percentage of CC FDs (FD%) within a given area was measured using the compensated CC en face flow images after excluding the areas of hyperpigmentation and quantified using the fuzzy C-means thresholding algorithm as previously described.4,5 The CC FDs were measured and compared within fovea-centered 1-mm, 3-mm, and 5-mm circles superimposed on the scans before and after cataract surgery, and this avoided the need to register the before and after scans. The fovea in each scan was identified by manually examining the A-scans and B-scans to identify the foveal center. 
Statistical analyses were performed using SPSS Statistics 28 (IBM Corporation, Chicago, IL), with P < 0.05 considered to be statistically significant. Continuous data were described as mean ± standard deviation (SD). Generalized estimating equation models with exchangeable correlation structure were used to account for the inclusion of two eyes of some participants when assessing change in image quality and CC FDs values before and after cataract surgery. Intraclass correlation coefficients (ICCs) were analyzed to study the repeatability of the CC FDs quantification. 
Results
A total of 25 eyes from 20 participants were included in the COMPARE study from January 2018 to August 2020. One eye was excluded because the area of hyperpigmentation occupied more than 20% of the central 1-mm circle. The average age of the remaining 19 participants was 71.47 ± 8.69 years, and 10 of them were female (52.6%). Out of the enrolled 24 eyes, nine eyes were considered normal without any history of ocular disease, 13 eyes were diagnosed with intermediate age-related macular degeneration, one patient had a history of Plaquenil use for 15 years without retinopathy, and one patient was diagnosed with diabetes mellitus without any retinopathy. There were no complications of cataract surgery in any of the patients. The first visit was 21.92 ± 32.04 days before cataract surgery, and the second visit was 57.63 ± 40.90 days after surgery. 
Comparisons of image quality values before and after cataract surgery are shown in Table 1. In the central 1-mm circle, the mean image quality score in cataract eyes before surgery was 2.75, and this number improved to 3.14 after surgery. In the central 3-mm circle, the mean score was 2.96 before surgery, and this value improved to 3.50 after surgery. In the central 5-mm circle, the mean image quality increased from 3.14 to 3.70 after surgery. The differences in the image quality scores before and after surgery were statistically significant in all the three circles, with 95% confidence intervals (CIs) excluding zero and P < 0.05. Among the 24 eyes, 17 cases presented with improved image quality (mean image quality score improved from 2.82 to 3.65 in the 1-mm circle, from 3.07 to 4.07 in the 3-mm circle, and from 3.23 to 4.21 in the 5-mm circle), and seven of them had worse quality (mean image quality score decreased from 2.58 to 1.90 in the 1-mm circle, from 2.71 to 2.15 in the 3-mm circle, and from 2.93 to 2.48 in the 5-mm circle) after surgery compared with preoperative scans. Figure 1 shows a case with better image quality postoperatively where the scores improved from 2.33, 2.06, and 2.16 to 4.82, 4.86, and 4.64 in the 1-mm, 3-mm, and 5-mm circles, respectively. Figure 2 shows a case with worse image quality after surgery. The numbers decreased from 3.52, 4.24, and 4.41 in the 1-mm, 3-mm, and 5-mm circles to 3.44, 4.10, and 4.11, respectively. This decrease appeared to be due to an increase in vitreous opacities in the lower half of the scan. 
Table 1.
 
Comparison of Image Quality in Different Macular Regions Before and After Cataract Surgery
Table 1.
 
Comparison of Image Quality in Different Macular Regions Before and After Cataract Surgery
Figure 2.
 
Example of decreased image quality and increased choriocapillaris flow deficit measurements after cataract surgery. (A–E) The top row shows image quality and quantitation of CC FDs from a scan before cataract surgery. (F–J) The bottom row shows image quality and CC FDs from a scan of the same subject after surgery. (A, F) Retinal en face angiographic images. (B, G) Retinal en face structural images. (C, H) Quality maps produced by the algorithm with different colors indicating different levels of image quality. Different quality levels are shown in the scale beneath panel H, with 5 (red) being the highest image quality (scaled 1–5). The 1-mm, 3-mm, and 5-mm circles centered on the fovea are shown. Image quality scores before surgery (C) were 3.52, 4.24, and 4.41 in the 1-mm, 3-mm, and 5-mm circles, respectively. After surgery (H), image quality scores were 3.44, 4.10, and 4.11 in the 1-mm, 3-mm, and 5-mm circles, respectively. (D, I) CC en face structural images produced by the CC algorithm. (E, J) Binarized CC flow images, with white regions denoting the CC FDs. Circles depict the 1-mm, 3-mm, and 5-mm circles centered on the fovea. (E) Before cataract surgery, CC FD% measurements were 9.99%, 12.76%, and 10.33%, respectively, in the central 1-mm, 3-mm, and 5-mm circles. (J) After cataract surgery, CC, FD% measurements decreased to 10.28%, 13.78%, and 11.26% in the 1-mm, 3-mm, and 5-mm circles, respectively.
Figure 2.
 
Example of decreased image quality and increased choriocapillaris flow deficit measurements after cataract surgery. (A–E) The top row shows image quality and quantitation of CC FDs from a scan before cataract surgery. (F–J) The bottom row shows image quality and CC FDs from a scan of the same subject after surgery. (A, F) Retinal en face angiographic images. (B, G) Retinal en face structural images. (C, H) Quality maps produced by the algorithm with different colors indicating different levels of image quality. Different quality levels are shown in the scale beneath panel H, with 5 (red) being the highest image quality (scaled 1–5). The 1-mm, 3-mm, and 5-mm circles centered on the fovea are shown. Image quality scores before surgery (C) were 3.52, 4.24, and 4.41 in the 1-mm, 3-mm, and 5-mm circles, respectively. After surgery (H), image quality scores were 3.44, 4.10, and 4.11 in the 1-mm, 3-mm, and 5-mm circles, respectively. (D, I) CC en face structural images produced by the CC algorithm. (E, J) Binarized CC flow images, with white regions denoting the CC FDs. Circles depict the 1-mm, 3-mm, and 5-mm circles centered on the fovea. (E) Before cataract surgery, CC FD% measurements were 9.99%, 12.76%, and 10.33%, respectively, in the central 1-mm, 3-mm, and 5-mm circles. (J) After cataract surgery, CC, FD% measurements decreased to 10.28%, 13.78%, and 11.26% in the 1-mm, 3-mm, and 5-mm circles, respectively.
CC FDs were quantified from the four repeated scans obtained before and after cataract surgery, and an analysis of repeatability was performed (Table 2). In eyes with and without cataracts, the CC FD measurements were shown to have high repeatability within the central 1-mm, 3-mm, and 5-mm circles (all values over 0.95). 
Table 2.
 
Repeatability of CC FD Measurements in Different Macular Regions Before and After Cataract Surgery
Table 2.
 
Repeatability of CC FD Measurements in Different Macular Regions Before and After Cataract Surgery
Figure 3 depicts the modified Early Treatment Diabetic Retinopathy Study (ETDRS) grid used for the comparisons of CC FD% measurements. Table 3 shows comparisons of CC FD% measurements before and after cataract surgery within different regions. In the central 1-mm circle, the mean CC FD% decreased from 15.17% before surgery to 12.87% after surgery (P < 0.001). The difference was 2.31%, with a 95% CI between 0.99% and 3.62%. In the 3-mm circle, the mean CC FD% decreased significantly from 12.03% to 11.36% after cataract removal (P = 0.011; mean difference, 0.67%; 95% CI, 0.15–1.20). However, in the 5-mm circle, the difference in CC FD% measurements before and after surgery was not statistically significant (P = 0.509; mean difference, 0.11%; 95% CI, −0.22 to 0.45). In the 17 cases with better image quality after surgery, the mean CC FD% decreased from 14.70% to 11.82% in the 1-mm circle, from 12.25% to 11.17% in the 3-mm circle, and from 9.78% to 9.25% in the 5-mm circle. In the seven cases with worse image quality after surgery, the mean CC FD% increased from 15.08% to 16.09% in the 1-mm circle, from 11.18% to 11.48% in the 3-mm circle, and from 8.97% to 9.66% in the 5-mm circle. In the case shown in Figure 1 with better image quality after surgery, the CC FD% measurements decreased from 20.12%, 12.27%, and 9.50% in the central 1-mm, 3-mm, and 5-mm circles to 16.62%, 11.66%, and 8.95%, respectively. In the case shown in Figure 2 with worse image quality after surgery, there were increases in CC FD% measurements from 9.99%, 12.76%, and 10.33% to 10.28%, 13.78%, and 11.26% in the 1-mm, 3-mm, and 5-mm circles, respectively. Table 3 also shows all of the CC FD% measurements from all of the grid regions, and there were no significant changes in CC FD% measurements in any of the inner and outer quadrants. Not surprisingly, we found statistically significant negative correlations when we correlated the change in image quality with the change in the CC FD% within the 1-mm circle (P = 0.004), 3-mm circle (P < 0.001), and the 5-mm circle (P < 0.001). 
Figure 3.
 
ETDRS grid used to assess macular choriocapillaris perfusion. The modified ETDRS grid in this study consisted of 1-mm, 3-mm, and 5-mm circles centered on the fovea, with the region between the 1-mm and 5-mm circles being divided into different quadrants (inner nasal, inner superior, inner temporal, inner inferior, outer nasal, outer superior, outer temporal, and outer inferior).
Figure 3.
 
ETDRS grid used to assess macular choriocapillaris perfusion. The modified ETDRS grid in this study consisted of 1-mm, 3-mm, and 5-mm circles centered on the fovea, with the region between the 1-mm and 5-mm circles being divided into different quadrants (inner nasal, inner superior, inner temporal, inner inferior, outer nasal, outer superior, outer temporal, and outer inferior).
Table 3.
 
Comparison of CC FD Measurements Before and After Cataract Surgery Within Different Macular Regions
Table 3.
 
Comparison of CC FD Measurements Before and After Cataract Surgery Within Different Macular Regions
Discussion
The COMPARE Study showed that, although overall image quality improved after cataract surgery, this improvement primarily impacted the measurements of CC FDs in the 1-mm fovea-centered circle, and there were no significant changes in the quadrants within the 3-mm and 5 mm circles. In previous studies, neutral density filters were used to simulate the impact of media opacities.8,2022 However, our study specifically analyzed the impact of cataracts, which were real-world media opacities that we found impacted both image quality and OCTA-based quantitative measurements. Although the overall image quality of our scans did improve after cataract surgery, there were seven cases in which the image quality decreased, and these decreases were primarily due to an increase in postsurgical vitreous opacities and a variability in the corneal tear film, as the use of artificial tears was at the discretion of the OCT operator. When studying the reliability of the CC FD% measurements using our quantitative algorithm, the repeatability of the measurements on the four repeated scans of the same patients obtained by the same technician before and after surgery were compared. Although the ICCs were slightly higher after surgery, all of the ICC values were over 0.95, which represented an overall high level of reliability for the CC FD% measurements both before and after cataract surgery. 
Križanović et al.23 used a Heidelberg OCTA device to obtain scans centered on the fovea and assessed the images by using a quality index, Q, which was computed by the instrument's software. In their study, they reported an increase in retinal macular perfusion in all layers after phacoemulsification, but they failed to find any significant changes of OCTA image quality or vascular parameters within the CC. Haddad et al.24 did find a significant increase in the signal strength index, which was calculated by the instrument's software, but they only found significant perfusion changes in the retina, not in the CC. In contrast, Lee et al.8 reported that neutral density filters resulted in both a decrease in image quality, which was reported by the instrument using the signal strength, and a significant increase in CC FDs. In our study, we applied a novel algorithm to assess the image quality that was provided by the manufacturer of the instrument. Although the overall image quality improved after cataract surgery, statistically significant decreases in CC FDs% were only detected in the 1-mm and 3-mm circles. We further removed the 1-mm region from the 3-mm circle and divided the rim into quadrants and found that there were no significant changes in the CC FD% measurements within the modified ETDRS quadrants. This indicates that the fovea-centered 1-mm circle, being located along the visual axis, is influenced the most by cataracts, which is consistent with the loss of vision being the primary indicator for cataract surgery.25 This also suggests that the use of uniform neutral density filters may not adequately simulate the effects of cataracts, as the crystalline lens is a biconvex structure26 that is thicker along the visual axis. The analogous use of neutral density filters to simulate a cataract may require a gradient neutral density filter with a higher attenuation along the visual axis and gradually reduced attenuation toward the radial direction. Of note, in this current report, as the image quality improved after surgery, the CC FDs in the 1-mm circles decreased, whereas in the seven eyes that developed worse quality after surgery the CC FD% measurements tended to increase (Fig. 2). 
The main limitations of our study were the relatively small sample size and the absence of a formal grading system to assess cataract severity, but we wanted this study to reflect a real-world population of patients undergoing cataract surgery. However, despite this, we did observe significant improvements in image quality after surgery and the decreased CC FDs in the 1-mm circle, which would be expected to persist with a larger sample size. These results highlight the importance of considering phakic status and image quality when measuring CC FDs within central regions of the macula. However, the results in other areas may have to be interpreted cautiously because statistically significant changes in CC FD% might become evident with a larger number of patients. Another possible limitation of this study is the applicability of the OCTA image quality algorithm used in this work to images from different manufacturers. The algorithm is specifically available for 6 × 6-mm scans as provided by the manufacturer of the instrument. Although the described technique should work for OCTA scans with other fields of view, with different resolutions, and from different manufacturers, it has not been validated previously for such use and, due to the machine learning nature of the algorithm, it should require additional training, including these other images with the proper manual annotations. 
In summary, the presence of cataracts corresponded with worse image quality and increased CC FD measurements within the central 1-mm circle, which would suggest a decrease in the ability to accurately assess CC perfusion in the presence of cataracts. When measuring CC FDs, especially CC FDs in the central 1-mm circle, the phakic status as well as image quality need to be considered. In phakic eyes, the impact of cataracts and decreased image quality may result in inaccurate CC flow measurements, particularly within regions along the visual axis. 
Acknowledgments
The authors appreciate the extraordinary technical expertise of Linda O'Koren and Mark Lazcano in obtaining all of the SS-OCT scans. 
Supported by a grant from the Salah Foundation; a grant from the National Eye Institute, National Institutes of Health (R01EY028753); by Carl Zeiss Meditec; by an unrestricted grant from Research to Prevent Blindness; and by a National Eye Institute Center Core Grant (P30EY014801) to the Department of Ophthalmology, University of Miami Miller School of Medicine. The funding organizations had no role in the design or conduct of this research. 
Disclosure: J. Li, None; M. Shen, None; Y. Cheng, None; Q. Zhang, None; J. Liu, None; L. de Sisternes, Carl Zeiss Meditec (E); W.H. Lewis, Carl Zeiss Meditec (C); R.K. Wang, Carl Zeiss Meditec (C), Insight Photonic Solutions (C), Colgate Palmolive (R), Estee Lauder Inc. (R), Oregon Health and Science University (I), University of Washington (I); G. Gregori, Carl Zeiss Meditec (P, R), University of Miami (P); P.J. Rosenfeld, Alexion (R), Annexon (C), Apellis (C, F), Bayer (C), Boehringer-Ingelheim (C), Carl Zeiss Meditec (C, R), Chengdu Kanghong Biotech (C), Gyroscope Therapeutics (R), InflammX Therapeutics (C), Ocudyne (C, F), Regeneron (C), Stealth Bio Therapeutics (R), Unity Biotechnology (C), Valitor (F), Verana Health (F) 
References
Lejoyeux R, Benillouche J, Ong J, et al. Choriocapillaris: fundamentals and advancements. Prog Retin Eye Res. 2022; 87: 100997. [CrossRef] [PubMed]
Zhang Q, Shi Y, Zhou H, et al. Accurate estimation of choriocapillaris flow deficits beyond normal intercapillary spacing with swept source OCT angiography. Quant Imaging Med Surg. 2018; 8(7): 658–666. [CrossRef] [PubMed]
Chu Z, Zhang Q, Gregori G, Rosenfeld PJ, Wang RK. Guidelines for imaging the choriocapillaris using OCT angiography. Am J Ophthalmol. 2021; 222: 92–101. [CrossRef] [PubMed]
Chu Z, Zhou H, Cheng Y, Zhang Q, Wang RK. Improving visualization and quantitative assessment of choriocapillaris with swept source OCTA through registration and averaging applicable to clinical systems. Sci Rep. 2018; 8(1): 16826. [CrossRef] [PubMed]
Chu Z, Gregori G, Rosenfeld PJ, Wang RK. Quantification of choriocapillaris with optical coherence tomography angiography: a comparison study. Am J Ophthalmol. 2019; 208: 111–123. [CrossRef] [PubMed]
Spaide RF, Fujimoto JG, Waheed NK. Image artifacts in optical coherence tomography angiography. Retina. 2015; 35(11): 2163–2180. [CrossRef] [PubMed]
Ji YS, Alagorie AR, Byon I, Sadda SR. Impact of scan tilt on quantitative assessments using optical coherence tomography angiography. Transl Vis Sci Technol. 2020; 9(7): 46. [CrossRef] [PubMed]
Lee JJ, Lee JE, Sadda SR, Park SW, Byon I. Impact of signal strength on quantitative retinal and choriocapillaris flow measurement from optical coherence tomography angiography. Sci Rep. 2022; 12(1): 4692. [CrossRef] [PubMed]
Shen M, Shi Y, Wang L, et al. Impact of cataract surgery on low luminance visual acuity deficit measurements. Ophthalmol Sci. 2022; 2(3): 100170. [CrossRef] [PubMed]
Shi Y, Zhang Q, Zhou H, et al. Correlations between choriocapillaris and choroidal measurements and the growth of geographic atrophy using swept source OCT imaging. Am J Ophthalmol. 2021; 224: 321–331. [CrossRef] [PubMed]
Laiginhas R, Shi Y, Shen M, et al. Persistent hypertransmission defects detected on en face swept source optical computed tomography images predict the formation of geographic atrophy in age-related macular degeneration. Am J Ophthalmol. 2022; 237: 58–70. [CrossRef] [PubMed]
Liu J, Laiginhas R, Corvi F, et al. Diagnosing persistent hypertransmission defects on en face OCT imaging of age-related macular degeneration. Ophthalmol Retina. 2022; 6(5): 387–397. [CrossRef] [PubMed]
Laiginhas R, Liu J, Shen M, et al. Multimodal imaging, OCT B-scan localization, and en face OCT detection of macular hyperpigmentation in eyes with intermediate age-related macular degeneration. Ophthalmol Sci. 2022; 2(2): 100116. [CrossRef] [PubMed]
Zhang Q, Zhang A, Lee CS, et al. Projection artifact removal improves visualization and quantitation of macular neovascularization imaged by optical coherence tomography angiography. Ophthalmol Retina. 2017; 1(2): 124–136. [CrossRef] [PubMed]
Chu Z, Chen CL, Zhang Q, et al. Complex signal-based optical coherence tomography angiography enables in vivo visualization of choriocapillaris in human choroid. J Biomed Opt. 2017; 22(12): 1–10. [CrossRef] [PubMed]
Wang RK, An L, Francis P, Wilson DJ. Depth-resolved imaging of capillary networks in retina and choroid using ultrahigh sensitive optical microangiography. Opt Lett. 2010; 35(9): 1467–1469. [CrossRef] [PubMed]
Haralick RM. Statistical and structural approaches to texture. Proc IEEE. 1979; 67(5): 786–804. [CrossRef]
Tibshirani R. Regression shrinkage and selection via the LASSO. J R Stat Soc Series B Stat Methodol. 1996; 58(1): 267–288.
Zhang Q, Zheng F, Motulsky EH, et al. A novel strategy for quantifying choriocapillaris flow voids using swept-source OCT angiography. Invest Ophthalmol Vis Sci. 2018; 59(1): 203–211. [CrossRef] [PubMed]
Yu JJ, Camino A, Liu L, et al. Signal strength reduction effects in OCT angiography. Ophthalmol Retina. 2019; 3(10): 835–842. [CrossRef] [PubMed]
Zhang J, Tang FY, Cheung CY, Chen H. Different effect of media opacity on vessel density measured by different optical coherence tomography angiography algorithms. Transl Vis Sci Technol. 2020; 9(8): 19. [CrossRef] [PubMed]
Zhang J, Tang FY, Cheung C, Chen X, Chen H. Different effect of media opacity on automated and manual measurement of foveal avascular zone of optical coherence tomography angiographies. Br J Ophthalmol. 2021; 105(6): 812–818. [CrossRef] [PubMed]
Križanović A, Bjelos M, Busic M, Elabjer BK, Rak B, Vukojevic N. Macular perfusion analysed by optical coherence tomography angiography after uncomplicated phacoemulsification: benefits beyond restoring vision. BMC Ophthalmol. 2021; 21(1): 71. [CrossRef] [PubMed]
Haddad F, Mimouni M, Nemet A, et al. Changes in chorioretinal flow index after cataract surgery: an optical coherence tomography angiography study. Int Ophthalmol. 2023; 43(1): 35–41. [CrossRef] [PubMed]
Michael R, Bron AJ. The ageing lens and cataract: a model of normal and pathological ageing. Philos Trans R Soc Lond B Biol Sci. 2011; 366(1568): 1278–1292. [CrossRef] [PubMed]
Tripathi RC, Tripathi BJ. Lens morphology, aging, and cataract. J Gerontol. 1983; 38(3): 258–270. [CrossRef] [PubMed]
Figure 1.
 
Example of increased image quality and decreased CC FD measurements after cataract surgery. (A–E) The top row shows image quality and quantitation of CC FDs from a scan before cataract surgery. (F–J) The bottom row shows image quality and CC FDs from a scan of the same subject from after surgery. (A, F) Retinal en face angiographic images. (B, G) Retinal en face structural images. (C, H) Quality maps produced by the algorithm with different colors indicating different levels of image quality. Different quality levels are shown in the scale beneath panel H, with 5 (red) being the highest image quality (scaled 1–5). The 1-mm, 3-mm, and 5-mm circles centered on the fovea are shown. Image quality scores before surgery (C) were 2.33, 2.06, and 2.16 in the 1-mm, 3-mm, and 5-mm circles respectively. After surgery (H), image quality scores were 4.82, 4.86, and 4.64 respectively in the 1-mm, 3-mm, and 5-mm circles, respectively. (D, I) CC en face structural images produced by the CC algorithm. (E, J) Binarized CC flow images with white regions denoting the CC FDs. Circles depict the 1-mm, 3-mm, and 5-mm circles centered on the fovea. (E) Before cataract surgery, the CC FD% measurements were 20.12%, 12.27%, and 9.50% in the central 1-mm, 3-mm, and 5-mm circles, respectively. (J) After cataract surgery, the CC FD% measurements decreased to 16.62%, 11.66%, and 8.95% in the 1-mm, 3-mm, and 5-mm circles, respectively.
Figure 1.
 
Example of increased image quality and decreased CC FD measurements after cataract surgery. (A–E) The top row shows image quality and quantitation of CC FDs from a scan before cataract surgery. (F–J) The bottom row shows image quality and CC FDs from a scan of the same subject from after surgery. (A, F) Retinal en face angiographic images. (B, G) Retinal en face structural images. (C, H) Quality maps produced by the algorithm with different colors indicating different levels of image quality. Different quality levels are shown in the scale beneath panel H, with 5 (red) being the highest image quality (scaled 1–5). The 1-mm, 3-mm, and 5-mm circles centered on the fovea are shown. Image quality scores before surgery (C) were 2.33, 2.06, and 2.16 in the 1-mm, 3-mm, and 5-mm circles respectively. After surgery (H), image quality scores were 4.82, 4.86, and 4.64 respectively in the 1-mm, 3-mm, and 5-mm circles, respectively. (D, I) CC en face structural images produced by the CC algorithm. (E, J) Binarized CC flow images with white regions denoting the CC FDs. Circles depict the 1-mm, 3-mm, and 5-mm circles centered on the fovea. (E) Before cataract surgery, the CC FD% measurements were 20.12%, 12.27%, and 9.50% in the central 1-mm, 3-mm, and 5-mm circles, respectively. (J) After cataract surgery, the CC FD% measurements decreased to 16.62%, 11.66%, and 8.95% in the 1-mm, 3-mm, and 5-mm circles, respectively.
Figure 2.
 
Example of decreased image quality and increased choriocapillaris flow deficit measurements after cataract surgery. (A–E) The top row shows image quality and quantitation of CC FDs from a scan before cataract surgery. (F–J) The bottom row shows image quality and CC FDs from a scan of the same subject after surgery. (A, F) Retinal en face angiographic images. (B, G) Retinal en face structural images. (C, H) Quality maps produced by the algorithm with different colors indicating different levels of image quality. Different quality levels are shown in the scale beneath panel H, with 5 (red) being the highest image quality (scaled 1–5). The 1-mm, 3-mm, and 5-mm circles centered on the fovea are shown. Image quality scores before surgery (C) were 3.52, 4.24, and 4.41 in the 1-mm, 3-mm, and 5-mm circles, respectively. After surgery (H), image quality scores were 3.44, 4.10, and 4.11 in the 1-mm, 3-mm, and 5-mm circles, respectively. (D, I) CC en face structural images produced by the CC algorithm. (E, J) Binarized CC flow images, with white regions denoting the CC FDs. Circles depict the 1-mm, 3-mm, and 5-mm circles centered on the fovea. (E) Before cataract surgery, CC FD% measurements were 9.99%, 12.76%, and 10.33%, respectively, in the central 1-mm, 3-mm, and 5-mm circles. (J) After cataract surgery, CC, FD% measurements decreased to 10.28%, 13.78%, and 11.26% in the 1-mm, 3-mm, and 5-mm circles, respectively.
Figure 2.
 
Example of decreased image quality and increased choriocapillaris flow deficit measurements after cataract surgery. (A–E) The top row shows image quality and quantitation of CC FDs from a scan before cataract surgery. (F–J) The bottom row shows image quality and CC FDs from a scan of the same subject after surgery. (A, F) Retinal en face angiographic images. (B, G) Retinal en face structural images. (C, H) Quality maps produced by the algorithm with different colors indicating different levels of image quality. Different quality levels are shown in the scale beneath panel H, with 5 (red) being the highest image quality (scaled 1–5). The 1-mm, 3-mm, and 5-mm circles centered on the fovea are shown. Image quality scores before surgery (C) were 3.52, 4.24, and 4.41 in the 1-mm, 3-mm, and 5-mm circles, respectively. After surgery (H), image quality scores were 3.44, 4.10, and 4.11 in the 1-mm, 3-mm, and 5-mm circles, respectively. (D, I) CC en face structural images produced by the CC algorithm. (E, J) Binarized CC flow images, with white regions denoting the CC FDs. Circles depict the 1-mm, 3-mm, and 5-mm circles centered on the fovea. (E) Before cataract surgery, CC FD% measurements were 9.99%, 12.76%, and 10.33%, respectively, in the central 1-mm, 3-mm, and 5-mm circles. (J) After cataract surgery, CC, FD% measurements decreased to 10.28%, 13.78%, and 11.26% in the 1-mm, 3-mm, and 5-mm circles, respectively.
Figure 3.
 
ETDRS grid used to assess macular choriocapillaris perfusion. The modified ETDRS grid in this study consisted of 1-mm, 3-mm, and 5-mm circles centered on the fovea, with the region between the 1-mm and 5-mm circles being divided into different quadrants (inner nasal, inner superior, inner temporal, inner inferior, outer nasal, outer superior, outer temporal, and outer inferior).
Figure 3.
 
ETDRS grid used to assess macular choriocapillaris perfusion. The modified ETDRS grid in this study consisted of 1-mm, 3-mm, and 5-mm circles centered on the fovea, with the region between the 1-mm and 5-mm circles being divided into different quadrants (inner nasal, inner superior, inner temporal, inner inferior, outer nasal, outer superior, outer temporal, and outer inferior).
Table 1.
 
Comparison of Image Quality in Different Macular Regions Before and After Cataract Surgery
Table 1.
 
Comparison of Image Quality in Different Macular Regions Before and After Cataract Surgery
Table 2.
 
Repeatability of CC FD Measurements in Different Macular Regions Before and After Cataract Surgery
Table 2.
 
Repeatability of CC FD Measurements in Different Macular Regions Before and After Cataract Surgery
Table 3.
 
Comparison of CC FD Measurements Before and After Cataract Surgery Within Different Macular Regions
Table 3.
 
Comparison of CC FD Measurements Before and After Cataract Surgery Within Different Macular Regions
×
×

This PDF is available to Subscribers Only

Sign in or purchase a subscription to access this content. ×

You must be signed into an individual account to use this feature.

×