Ophthalmic Surgery, Lasers and Imaging Retina

Clinical Science 

Noninvasive, High-Resolution Functional Macular Imaging in Subjects With Retinal Vein Occlusion

Thalmon R. Campagnoli, MD; Gábor Márk Somfai, MD; Jing Tian, MD; Delia Cabrera DeBuc, MD; William E. Smiddy, MD

Abstract

BACKGROUND AND OBJECTIVES:

Several imaging modalities have been developed to characterize ischemia inherent in retinal vascular diseases. This study aims to predict the impact and to better establish the mechanisms of visual deterioration. A high-resolution functional imaging device is used, yielding quantitative data for macular blood flow and capillary network features in healthy eyes and in eyes with central retinal vein occlusion (CRVO) or branch retinal vein occlusion (BRVO).

PATIENTS AND METHODS:

This prospective, cross-sectional, comparative case series measured blood flow velocities (BFVs) and noninvasive capillary perfusion maps (nCPMs) in macular vessels in patients with BRVO/CRVO and in healthy controls using the Retinal Function Imager (RFI; Optical Imaging, Rehovot, Israel).

RESULTS:

Twenty-two eyes of 21 subjects were studied (eight with CRVO, five with BRVO, and nine controls). A significant decrease was observed in the BFVs of both arterioles and venules in the affected macular region of patients with CRVO and BRVO (2.84 ± 1.21 mm/s and 2.67 ± 1.43 mm/s in CRVO/BRVO arterioles, respectively, vs. 4.23 ± 1.04 mm/s in healthy controls, P < .001; and 1.64 ± 0.51 mm/s and 1.60 ± 0.41 mm/s in CRVO/BRVO venules, respectively, vs. 2.88 ± 0.93 mm/s in healthy controls, P < .001). BFVs in non-affected macular regions of patients with BRVO were not statistically different from BFVs in healthy eyes (3.84 ± 1.04 mm/s and 3.17 ± 1.39 mm/s in BRVO patients vs. 4.23 ± 1.04 mm/s and 2.88 ± 0.93 mm/s in healthy controls' arterioles and venules, respectively; P ≥ .1). nCPMs allowed high-resolution imaging of the macular vasculature and successfully demonstrated ischemic areas in the RVO groups.

CONCLUSIONS:

The RFI provided high-resolution functional imaging of the retinal microvasculature and enabled quantitative measurement of BFVs in patients with RVO. Diminished flow velocity in arterioles and venules raises the possibility that RVO represents a panvascular compromise not confined to just venous stasis or its secondary arteriolar effects. The RFI offers potential to help with diagnosis and management of RVO cases. [Ophthalmic Surg Lasers Imaging Retina. 2017;48:799–809.]

Abstract

BACKGROUND AND OBJECTIVES:

Several imaging modalities have been developed to characterize ischemia inherent in retinal vascular diseases. This study aims to predict the impact and to better establish the mechanisms of visual deterioration. A high-resolution functional imaging device is used, yielding quantitative data for macular blood flow and capillary network features in healthy eyes and in eyes with central retinal vein occlusion (CRVO) or branch retinal vein occlusion (BRVO).

PATIENTS AND METHODS:

This prospective, cross-sectional, comparative case series measured blood flow velocities (BFVs) and noninvasive capillary perfusion maps (nCPMs) in macular vessels in patients with BRVO/CRVO and in healthy controls using the Retinal Function Imager (RFI; Optical Imaging, Rehovot, Israel).

RESULTS:

Twenty-two eyes of 21 subjects were studied (eight with CRVO, five with BRVO, and nine controls). A significant decrease was observed in the BFVs of both arterioles and venules in the affected macular region of patients with CRVO and BRVO (2.84 ± 1.21 mm/s and 2.67 ± 1.43 mm/s in CRVO/BRVO arterioles, respectively, vs. 4.23 ± 1.04 mm/s in healthy controls, P < .001; and 1.64 ± 0.51 mm/s and 1.60 ± 0.41 mm/s in CRVO/BRVO venules, respectively, vs. 2.88 ± 0.93 mm/s in healthy controls, P < .001). BFVs in non-affected macular regions of patients with BRVO were not statistically different from BFVs in healthy eyes (3.84 ± 1.04 mm/s and 3.17 ± 1.39 mm/s in BRVO patients vs. 4.23 ± 1.04 mm/s and 2.88 ± 0.93 mm/s in healthy controls' arterioles and venules, respectively; P ≥ .1). nCPMs allowed high-resolution imaging of the macular vasculature and successfully demonstrated ischemic areas in the RVO groups.

CONCLUSIONS:

The RFI provided high-resolution functional imaging of the retinal microvasculature and enabled quantitative measurement of BFVs in patients with RVO. Diminished flow velocity in arterioles and venules raises the possibility that RVO represents a panvascular compromise not confined to just venous stasis or its secondary arteriolar effects. The RFI offers potential to help with diagnosis and management of RVO cases. [Ophthalmic Surg Lasers Imaging Retina. 2017;48:799–809.]

Introduction

Retinal vein occlusions (RVOs) are the second leading cause of blindness and common retinal morbidity occasioned by retinal vascular diseases worldwide, carrying a 15-year cumulative incidence of 1.8% and 0.5% for branch retinal vein occlusion (BRVO) and central retinal vein occlusion (CRVO), respectively.1,2 BRVO occurs more frequently with systemic hypertension,3,4 but the pathogenesis of retinal vein occlusion is still incompletely understood.5

A glaring shortfall of existing modalities is quantitative blood flow measurements — a feature that may aid in understanding the pathophysiology and aid in predicting prognosis with or without treatment.

Several technologies have been used in attempting to quantify choroidal and retinal circulatory features, including video fluorescein angiography,6 ultrasound-7 and laser-based Doppler flowmetry techniques.8–13 Their requirements for precise eye fixation and alignment for relatively long periods of time, incapacity to noninvasively assess smaller diameter retinal vessels, provision of arbitrary units, and the use of dye limited the use of those technologies to provide clinically useful vascular information. Optical coherence tomography angiography (OCTA) has demonstrated a noninvasive generation of high-resolution maps of the retinal and choroidal microvasculature,14,15 but its capability to quantify blood flow features has yet to be established.

The Retinal Function Imager (RFI; Optical Imaging, Rehovot, Israel) is a device that provides fast, noninvasive, functional imaging of the retinal microvasculature, allowing the construction of high-resolution, noninvasive capillary perfusion maps (nCPMs) and the measurement of macular blood flow velocities (BFVs) at secondary and tertiary vascular segments. The RFI was originally used for functional optical imaging in the brain.16 The imaging acquisition process in the eye utilizes similar principles as in the brain and consists of a standard fundus camera coupled with a customized stroboscopic flash lamp and a fast (60 Hz), high-resolution (1,024 pixels × 1,024 pixels) digital camera system that captures the moving hemoglobin-filled erythrocytes utilizing green (“red-free”) light (wavelength centered at 548 nm at a bandwidth of 17 nm) and digital subtraction processing algorithms.17 BFV measurements are based on the generation of short movies (eight frames each) by the use of the machine's built-in software, and nCPMs are generated by the digital enhancement of alignment of those static frames.

The present study investigates the BFV imaging and blood velocity capabilities in eyes with CRVO/BRVO in comparison to a healthy control group.

Patients and Methods

This prospective, cross-sectional study was approved by the Human Subjects Committee of the University of Miami Miller School of Medicine and was conducted in accordance with the Declaration of Helsinki for research in human subjects. Study and healthy control patients were evaluated at the Bascom Palmer Eye Institute between September 1, 2014, and April 30, 2015, using the RFI and OCT (Spectralis; Heidelberg Engineering, Heidelberg, Germany) or Cirrus (Carl Zeiss Meditec, Dublin, CA). All study patients underwent a comprehensive ophthalmic exam including medical history, best-corrected visual acuity (BCVA), intraocular pressure (IOP) measurement, slit-lamp examination, and dilated fundus examination. The decision to perform fluorescein angiography (FA) was made at the physician's discretion on a case-by-case basis. Patients with substantial media opacities, previous ocular surgery other than uneventful cataract surgery at least 6 months prior to the imaging, and with refractive error more than 6 diopters were excluded from the study. All study patients and controls were 45 years old or older. Controls had normal cardiac rhythm, IOP, and ophthalmic exam, including BCVA 20/20.

The RFI scanning method consisted of a series of snapshots captured under the blood flow velocity operating mode and required about 10 minutes per eye. A probe was attached at the patient's fingertip or earlobe to synchronize the flash discharge to a selected phase of the cardiac cycle (100 milliseconds [ms] before systole) in order to control the effects of pulse variations on BFV measurements and to provide maximal flow velocities. A snapshot (one session) represents the delivery of eight consecutive flashes with an interflash interval of 17.5 ms, yielding a sequence of eight fundus images (eight frames) (See Supplemental Video). Proper alignment of these sequential images by a built-in software allowed construction of the nCPM. BFV measurements were obtained from software analysis of a short movie (eight frames) and allowed direct detection of the signals generated by the moving erythrocyte clumps. Each vessel in the region of interest was manually segmented with the supervision of a semi-automatic vessel-detecting algorithm, allowing the software to calculate the segment's blood flow velocity and also to identify the type of vessel (artery vs. vein) based on the direction of the flow in relation to the foveal center (centripetal = artery; centrifugal = vein).

The detailed methodology used to perform the RFI scanning and analysis has been described elsewhere.18 In summary, in order to allow reliable BFV measurements only eyes with three or more good-quality image acquisition sessions containing at least five good frames per session were included. All subjects were imaged by the same experienced photographer (TRC or GMS) centered at the fovea. Subjective analysis of optical resolution, level of light exposure, focus, and flow visibility by an experienced examiner determined the quality level of the images. All distal secondary and tertiary vessels within 3 mm from the foveal center and showing superior quality motion contrast were segmented in both long (> 100-pixel-long segments for 20° and > 80-pixel-long segments for 35° images) and short segments (< 100-pixel-long segments for 20° and < 80-pixel-long segments for 35° images). The process of vessel segmentation strictly within the macula was possible by precise central foveal detection and overlapping of the OCT thickness map with the RFI image as described elsewhere.19 Segmentation of vessel intersection regions and continued segmentation of vessels beyond its bifurcation or branching were avoided in order to prevent optical interference between arterial and venous flow and measurement bias in vessels with different diameters, as previously described18 (Figure 1). In the eyes with BRVO, a subanalysis was performed for the arterioles and venules in the macular regions not affected by the occlusion (ie, outside the region affected by the occlusion and not directly influenced by the circulation of a distorted retinal tissue).

This is an example of a normal subject's blood flow velocity segmentation (right eye). The individual vascular segments were identified and flow velocities were calculated by the machine's software. The arteries are color-coded red and have negative direction by convention, whereas the veins are purple with positive values. The arterial and venular flow velocities averaged 4.21 mm/s and 3.01 mm/s, respectively. Note in the temporal-inferior macula that continuous arteriole segmentation beyond its bifurcation was avoided through independent segmentation of two shorter distal vessels.

Figure 1.

This is an example of a normal subject's blood flow velocity segmentation (right eye). The individual vascular segments were identified and flow velocities were calculated by the machine's software. The arteries are color-coded red and have negative direction by convention, whereas the veins are purple with positive values. The arterial and venular flow velocities averaged 4.21 mm/s and 3.01 mm/s, respectively. Note in the temporal-inferior macula that continuous arteriole segmentation beyond its bifurcation was avoided through independent segmentation of two shorter distal vessels.

A path-constrained cross-correlation technique was used for BFV calculation, giving separate BFV results for arteries and veins.20 Combined nCPMs were generated through selection of three or more nCPM sessions.

In order to exclude poorly analyzed data, all vessel segments with 45% or greater coefficient of variance (standard deviation [SD]/mean) between sessions were excluded from the analysis. Eyes containing more than 33% of vessel segments with measurements over the coefficient of variance limit were also excluded from analysis.19

Data analysis was made using SPSS Statistics software version 22.0.0.0 (IBM, Armonk, NY). To assess the differences in arteriolar and venular BFVs between the RVO study groups and controls, we performed pairwise two-tailed t-tests (control vs. BRVO and control vs. CRVO), and set the level of significance at 5%.

Results

The current study evaluated 22 eyes of 21 study patients, including eight eyes diagnosed with CRVO, five with BRVO, and nine healthy control eyes. Detailed demographic and clinical information of the study patients with retinopathy enrolled are shown in Table 1. The healthy subjects group was composed of females presenting average age of 53 years old, ranging from 46 years old to 62 years old, and with no known systemic or ocular diseases as previously described.

Subjects' Demographic and Clinical Characteristics

Table 1:

Subjects' Demographic and Clinical Characteristics

Arteriolar BFV (mean ± SD) in control eyes was 4.23 mm/s ± 1.04 mm/s whereas venular BFV was 2.88 mm/s ± 0.93 mm/s. The number of arterioles and venules segmented per eye (mean ± SD) in the control group was 10 ± 4 and 12 ± 4 respectively a total of 93 arterioles and 106 venules was segmented in the control group (Table 2). Figure 1 shows the segmentation of a control eye and its respective nCPM.

Velocity Results Among the Patient Groups

Table 2:

Velocity Results Among the Patient Groups

An average of 11 ± 6 arterioles and 14 ± 4 venules were segmented per eye diagnosed with CRVO, totaling 90 arterioles and 109 venules segmented in the group. Both arteriolar and venular BFVs were significantly decreased in the CRVO group when compared to controls (P < .001), with mean ± SD values of 2.84 mm/s ± 1.21 mm/s for arterioles and 1.64 mm/s ± 0.51 mm/s for venules. All study patients in the CRVO group presented arteriolar or venular mean BFV above the mean BFV for the group, except for Patient 7 (arteriolar and venular BFV of 1.28 mm/s and 1.17 mm/s, respectively, versus mean 2.84 mm/s and 1.64 mm/s arterioles and venules BFV, respectively, in the CRVO group) (Table 3). The same subject presented BCVA of 20/400 at the time of RFI testing. Also, study Patient 7 had a total of four arterioles segmented; the lower number of segments available for analysis for this subject was the result of severe edema and intraretinal hemorrhage in the macular region leading to poor optical resolution, focus and flow visibility, and consequent coefficient of variance between sessions of 45% or greater in most of the arterioles segmented. Study Patient 8 presented to the clinic for evaluation of epiretinal membrane (ERM) having had four previous intravitreal bevacizumab (Avastin; Genentech, South San Francisco, CA) injections; fundus examination confirmed the ERM in the temporal-superior region of the macula, macular edema, and mild vascular tortuosity compared to the fellow eye. OCT demonstrated mild macular edema and paracentral ERM extending to the temporal-superior macula not judged to be causing BCVA of 20/400. There was suspicion of vessel occlusion; however, FA failed to demonstrate changes in the retinal blood flow. RFI scanning was performed following the study protocol and identified significantly reduced arteriolar and venular blood flow velocities in the four macular quadrants, supporting the otherwise tentative diagnosis of CRVO (Figure 2).

BFV Profiles of the Study Subjects

Table 3:

BFV Profiles of the Study Subjects

A 35° Retinal Functional Imager (RFI) image in Patient 8, who presented with visual loss in the left eye after having had a bevacizumab injection elsewhere, but the diagnosis was not clearly apparent. There was mild macular edema, but no intraretinal hemorrhage and only a suggestion of vascular dilation. He received two more cycles of bevacizumab, with slight visual improvement and resolution of the macular edema. (A) Demonstrates the late recirculation fluorescein angiogram phase that failed to evidence parafoveal ischemic changes. (B) Depicts the RFI vessel segmentation obtained on the same day that showed that mean arteriolar and venular flow velocities were about two-thirds and one-half of the healthy group, respectively.

Figure 2.

A 35° Retinal Functional Imager (RFI) image in Patient 8, who presented with visual loss in the left eye after having had a bevacizumab injection elsewhere, but the diagnosis was not clearly apparent. There was mild macular edema, but no intraretinal hemorrhage and only a suggestion of vascular dilation. He received two more cycles of bevacizumab, with slight visual improvement and resolution of the macular edema. (A) Demonstrates the late recirculation fluorescein angiogram phase that failed to evidence parafoveal ischemic changes. (B) Depicts the RFI vessel segmentation obtained on the same day that showed that mean arteriolar and venular flow velocities were about two-thirds and one-half of the healthy group, respectively.

The total number of arterioles and venules segmented in the BRVO group was 41 and 51, respectively, representing 8 ± 6 (mean ± SD) segmented arterioles and 10 ± 2 (mean ± SD) segmented venules per eye diagnosed with BRVO. A total of 17 arterioles and 17 venules were segmented in the macular regions affected by the venous occlusion, representing 3 ± 3 segmented arterioles (mean ± SD; range: zero segments to eight segments) and 3 ± 2 segmented venules (mean ± SD; range: one segment to six segments) in the macular region affected by the venous occlusion, per studied eye. In patients diagnosed with BRVO, a significant decrease in BFV was noticed when evaluating the total number of macular arterioles throughout all four macular quadrants (3.35 mm/s ± 1.33 mm/s in the BRVO versus 4.23 mm/s ± 1.04 mm/s in the healthy controls; P < .001), but no statistically significant BFV difference was seen when the total number of macular venules were evaluated (2.65 mm/s ± 1.37 mm/s in the BRVO group versus 2.88 mm/s ± 0.93 mm/s in the healthy controls; P = .2). When considering the vessels exclusively located in the macular regions affected by the vein occlusion (ie, arterioles and venules located distally to the occluded vessel), significant decrease in the BFV was detected for both arterioles and venules (2.67 mm/s ± 1.43 mm/s in the BRVO group arterioles versus 4.23 mm/s ± 1.04 mm/s in the healthy controls, P < .001; and 1.60 mm/s ± 0.41 mm/s in the BRVO group venules versus 2.88 mm/s ± 0.93 mm/s in the healthy controls, P < .001).

In the subanalysis of the macular regions not affected by the occlusion (ie, outside the region affected by the occlusion and not directly influenced by the circulation of a distorted retinal tissue), no changes were demonstrated in the arteriolar and venular BFVs when compared to the healthy group (3.84 mm/s ± 1.04 mm/s in the BRVO group arterioles versus 4.23 mm/s ± 1.04 mm/s in the healthy controls, P = .1; and 3.17 mm/s ± 1.39 mm/s in the BRVO group venules versus 2.88 mm/s ± 0.93 mm/s in the healthy controls, P = .17). Patient 13 had no arterioles meeting criteria to be included in the segmentation analysis due to severe macular edema; the patient's BCVA at the time of RFI testing was 20/100. All other patients in the BRVO group presented arterioles or venules mean BFV above the mean BFV for the group, except for study eye 9, which presented arteriolar and venular mean BFV below the mean BFV for the group (arterioles and venules BFV of 2.18 mm/s and 2.07 mm/s, respectively, versus a mean 3.35 mm/s and 2.65 mm/s arterioles and venules BFV in the BRVO group). The same study eye had a total of 13 arterioles and 12 venules segmented for the velocity analysis and presented BCVA of 20/400 at the time of RFI testing. Also of note, the other three study eyes with BRVO with arteriolar segments meeting inclusion criteria for the segmentation analysis presented arteriolar BFV over the mean arteriolar BFV for the group and BCVA equal or better than 20/40 at the time of RFI testing.

The nCPMs provided high-resolution images of the retinal microvasculature, in both healthy (Figure 1) and retinal vein occlusion eyes, enabling precise imaging of ischemic regions, comparable to FA images in cases of RVO (Figure 3).

This patient study (Patient 5) was obtained 7 months after an excellent response to intravitreal bevacizumab in the left eye. He was diagnosed with central retinal vein occlusion 62 months prior to the Retinal Functional Imager (RFI) scan and had a total of three intravitreal triamcinolone and four intravitreal bevacizumab injections. His vision at the time of the RFI scan was 20/20. The RFI image (A) closely parallels the appearance of the fluorescein angiogram (B) in terms of vascular resolution and depicting nonperfusion. The flow velocities, though diminished from the controls, were not as depressed as in other cases with poor visual acuity as demonstrated in Table 3.

Figure 3.

This patient study (Patient 5) was obtained 7 months after an excellent response to intravitreal bevacizumab in the left eye. He was diagnosed with central retinal vein occlusion 62 months prior to the Retinal Functional Imager (RFI) scan and had a total of three intravitreal triamcinolone and four intravitreal bevacizumab injections. His vision at the time of the RFI scan was 20/20. The RFI image (A) closely parallels the appearance of the fluorescein angiogram (B) in terms of vascular resolution and depicting nonperfusion. The flow velocities, though diminished from the controls, were not as depressed as in other cases with poor visual acuity as demonstrated in Table 3.

Discussion

This study demonstrated the capability of the RFI to provide noninvasive, functional, and quantitative information,21 as well as high-resolution mapping of the perfused macular vasculature in eyes with RVO.22 Although the RFI has been used to study retinal vein occlusions in preliminary studies to study response to anti-vascular endothelial growth factor treatment for RVO (partial normalization in oximetry), and to correlate blood flow velocities with OCT thickness (not a good correlation),23–25 the current study is the first to use it to aid in the diagnosis of RVO. The preliminary reports of use of the RFI suggest it might have a valuable use in retinal vascular disease since blood flow velocities (and oximetry) might shed different mechanistic and prognostic insight into the pathophysiology of retinal vascular disorders. The principal finding of the current study was that BFV was reduced in both venules (by 40%) and arterioles (by 30%) in eyes with BRVO or CRVO (Table 2). An overall BFV decrease was also noted in the arterioles of eyes with BRVO, but no statistically significant difference was seen in the overall BRVO venular BFV in comparison to control eyes. These findings suggest deficient retinal blood flow in the BRVO and CRVO cohorts, but the RFI does not directly measure blood volume. Autoregulation of blood flow supply prompted by damaged, metabolically hypoactive tissues might be mitigating the observed disparities in blood flow velocities.

Despite comparable age-matched study and control groups and standard imaging and segmentation methodologies between groups, the current study involves a limited number of eyes in different disease stages with a variable degree of fundus changes (eg, macular edema or intraretinal hemorrhage). Some of the eyes with BRVO also possess a very limited number of segmented vessels within the affected macular region, making such correlation speculative and not explanatory of good vision in eyes with lower than average arterial BFV. Some subjects from our cohort presented with primary open-angle glaucoma (POAG) at the time of the RFI testing. Although this could potentially influence the interpretation of BFV findings, a previous study comparing RFI BFV measurements between patients with POAG and healthy subjects have failed to demonstrate any significant difference between the groups as long as IOP was well-controlled.26

We did not assess the correlation between flow decrease and visual acuity, but there were individual cases that suggested this possibility (Patients 2, 4, and 20) (Table 3). However, it is important to note that the poorest BCVA in our study cohort (20/400) occurred in eyes with both arteriolar and venular mean BFV below the mean BFV for the group (Patients 7 and 9) (Table 3), and could be indicative of visual prognosis in a larger studied population. A larger sample size might allow a more precise insight on the correlation of BFV and visual function.

There are potential limitations in the RFI capabilities since anatomic disruption of retinal tissue and microcirculation not uncommonly precludes measurement of blood flow. For example, severe macular edema (ME) or intraretinal hemorrhages (IRH) precluded capturing valid BFV measurements in those regions. This occurred more frequently in affected BRVO regions and was one of the reasons why a lower-than-expected number of vessel segments in this setting was included in the study. The current study was not powered to assess the correlation between BFV and ME/IRH, and variable trends of BFVs were noticed in eyes with severe ME and/or IRHs (Table 3).

The vessel segmentation in the current study was confined to the distal secondary and tertiary vessels within 3 mm from the foveal center. The reason for this study design constraint was to optimize the quality of the imaging so as to maximize the BFV data20 and to allow reproducibility of measurements for correlation in anatomically distorted retinal tissue (eg, ME, ERM). Other RFI studies have evaluated retinal BFV,17,18,21,26–38 but their methodologies mostly differ in at least one of three key aspects: (1) the field of view (FOV) utilized for the images acquisition (20° versus 35°), (2) the length of vessel segmentation utilized in the calculation of BFVs (long segments: > 100 pixels long for the 20° images or > 80 pixels long for the 35° images; short segments: < 100 pixels long for the 20° images or < 80 pixels long for the 35° images; or mixed) and (3) the location of vessel segments involved in the analysis (exclusively in the macula versus also located peripherally to the macula). Our methodology was designed to provide more robust measurements based on a previous study.18 In our study we included eyes scanned with 20° FOV as well as 35°, and we used a mixed combination of vessel length segmentation that allowed good quality segmentation without approaching intersection regions or performing continued segmentation of vessels beyond its bifurcation or branching for the previously explained reasons.

Imaging acquisition with the RFI follows the same principles as the digital fundus camera, with higher magnification and increased focusing challenges for 20° imaging in comparison to 35°. From the BFV assessment standpoint, there is no statistical significant difference between BFV measurements taken from 20° to 35° imaging reported in the literature. We observed that the 20° imaged eyes were significantly more affected by mild media opacities (eg, trace nuclear sclerosis, vitreous opacities) or minimal retinal distortion commonly seen in diseased eyes (eg, intraretinal edema and/or hemorrhages, thickened posterior hyaloid, mild epiretinal membrane), consequently presenting higher rate of vessel segments rejection and exclusion from the study. In contrast, 35° FOV images allowed superior imaging in the presence of mildly higher levels of retinal distortion and/or media opacity, yielding better vessel segmentation while still presenting superior retinal detailing. nCPMs generated from 35° images were successful in showing capillary drop-out and clinically significant ischemic macular regions, comparable to FA imaging as evidenced in Figure 3, and perhaps superior to 20° imaging given better imaging quality in the setting of mild media opacities and/or retinal distortions commonly seen in diseased eyes. The use of a mixed combination of vessel length segmentation aimed to maximize the utilization of good motion contrast information within shorter segments with lower coefficient of variance in the analysis. Although exclusive long-segmentation pattern might provide more robust measurements, most RFI clinical reports utilize the mixed methodology for BFV analysis.18

To the best of our knowledge, this is the first report of imaging in patients with BRVO/CRVO with a technology enabling noninvasive and quantitative evaluation of retinal BFV at the capillary level. The nCPMs generated by the RFI device were demonstrated to be a useful tool to noninvasively assess macular ischemia, comparable to FA. In regard to BFV measurements, the study is limited mainly by the relatively low number of eyes in different disease stages and variable degree of fundus changes due to the baseline pathology (eg, ME, IRH). Also, there was a limited number of vessels segmented meeting inclusion criteria in some of the study eyes; however, the consistency of significant BFV decrease in venules and arterioles located in the affected macular regions of both BRVO and CRVO eyes (even in good BCVA and no comorbidities) seen in our study seemed reliable and potentially representative to stipulate a new quantifiable criterion for RVOs diagnosis. Larger, longitudinal studies are warranted to define the role of the RFI in studying eye diseases as the authors believe that the non-invasive technology used in the RFI can be a useful non-invasive tool for the differentiation between ophthalmoscopically mild BRVO and CRVO cases. Therefore, further studies evaluating BFV in BRVO/CRVO at different stages might significantly contribute to better disease understanding and prognosis.

References

  1. Klein R, Moss SE, Meuer SM, Klein BE. The 15-year cumulative incidence of retinal vein occlusion: The Beaver Dam Eye Study. Arch Ophthalmol. 2008;126(4):513–518. doi:10.1001/archopht.126.4.513 [CrossRef]
  2. No authors listed. Natural history and clinical management of central retinal vein occlusion. The Central Vein Occlusion Study Group. Arch Ophthalmol. 1997;115(4):486–491. doi:10.1001/archopht.1997.01100150488006 [CrossRef]
  3. Frangieh GT, Green WR, Barraquer-Somers E, Finkelstein D. Histopathologic study of nine branch retinal vein occlusions. Arch Ophthalmol. 1982;100(7):1132–1140. doi:10.1001/archopht.1982.01030040110020 [CrossRef]
  4. No authors listed. Risk factors for branch retinal vein occlusion. The Eye Disease Case-control Study Group. Am J Ophthalmol. 1993;116(3):286–296. doi:10.1016/S0002-9394(14)71345-5 [CrossRef]
  5. Green WR, Chan CC, Hutchins GM, Terry JM. Central retinal vein occlusion: A prospective histopathologic study of 29 eyes in 28 cases. Trans Am Ophthalmol Soc. 1981;79:371–422.
  6. Horio N, Horiguchi M. Retinal blood flow analysis using intraoperative video fluorescein angiography combined with optical fiber-free intravitreal surgery system. Am J Ophthalmol. 2004;138(6):1082–1083. doi:10.1016/j.ajo.2004.06.079 [CrossRef]
  7. Bohdanecka Z, Orgul S, Meyer AB, Prunte C, Flammer J. Relationship between blood flow velocities in retrobulbar vessels and laser Doppler flowmetry at the optic disk in glaucoma patients. Ophthalmologica. 1999;213(3):145–149. doi:10.1159/000027409 [CrossRef]
  8. Sakata K, Funatsu H, Harino S, Noma H, Hori S. Relationship between macular microcirculation and progression of diabetic macular edema. Ophthalmology. 2006;113(8):1385–1391. doi:10.1016/j.ophtha.2006.04.023 [CrossRef]
  9. Wang Y, Lu A, Gil-Flamer J, Tan O, Izatt JA, Huang D. Measurement of total blood flow in the normal human retina using Doppler Fourier-domain optical coherence tomography. Br J Ophthalmol. 2009;93(5):634–637. doi:10.1136/bjo.2008.150276 [CrossRef]
  10. Yoshida A, Feke GT, Mori F, et al. Reproducibility and clinical application of a newly developed stabilized retinal laser Doppler instrument. Am J Ophthalmol. 2003;135(3):356–361. doi:10.1016/S0002-9394(02)01949-9 [CrossRef]
  11. Kagemann L, Wollstein G, Ishikawa H, Townsend KA, Schuman JS. Validation of spectral domain optical coherence tomographic Doppler shifts using an in vitro flow model. Invest Ophthalmol Vis Sci. 2009;50(2):702–706. doi:10.1167/iovs.08-2305 [CrossRef]
  12. Jonescu-Cuypers CP, Harris A, Wilson R, et al. Reproducibility of the Heidelberg retinal flowmeter in determining low perfusion areas in peripapillary retina. Br J Ophthalmol. 2004;88(10):1266–1269. doi:10.1136/bjo.2003.039099 [CrossRef]
  13. Nagahara M, Tamaki Y, Tomidokoro A, Araie M. In vivo measurement of blood velocity in human major retinal vessels using the laser speckle method. Invest Ophthalmol Vis Sci. 2011;52(1):87–92. doi:10.1167/iovs.09-4422 [CrossRef]
  14. Puliafito CA. OCT angiography: The next era of OCT technology emerges. Ophthalmic Surg Lasers Imaging Retina. 2014;45(5):360. doi:10.3928/23258160-20140925-01 [CrossRef]
  15. Huang Y, Zhang Q, Thorell MR, et al. Swept-source OCT angiography of the retinal vasculature using intensity differentiation-based optical microangiography algorithms. Ophthalmic Surg Lasers Imaging Retina. 2014;45(5):382–389. doi:10.3928/23258160-20140909-08 [CrossRef]
  16. Grinvald A, Bonhoeffer T, Vanzetta I, et al. High-resolution functional optical imaging: From the neocortex to the eye. Ophthalmol Clin North Am. 2004;17(1):53–67. doi:10.1016/j.ohc.2003.12.003 [CrossRef]
  17. Izhaky D, Nelson DA, Burgansky-Eliash Z, Grinvald A. Functional imaging using the retinal function imager: Direct imaging of blood velocity, achieving fluorescein angiography-like images without any contrast agent, qualitative oximetry, and functional metabolic signals. Jpn J Ophthalmol. 2009;53(4):345–351. doi:10.1007/s10384-009-0689-0 [CrossRef]
  18. Somfai GM, Tian J, DeBuc DC. Assessment of potential vessel segmentation pitfalls in the analysis of blood flow velocity using the Retinal Function Imager. Graefes Arch Clin Exp Ophthalmol. 2016;254(6):1075–1081. doi:10.1007/s00417-015-3166-0 [CrossRef]
  19. Tian J, Somfai GM, Campagnoli TR, Smiddy WE, Debuc DC. Interactive retinal blood flow analysis of the macular region. Microvasc Res. 2016;104:1–10. doi:10.1016/j.mvr.2015.11.003 [CrossRef]
  20. Nelson DA, Krupsky S, Pollack A, et al. Special report: Noninvasive multi-parameter functional optical imaging of the eye. Ophthalmic Surg Lasers Imaging. 2005;36(1):57–66.
  21. Chhablani J, Bartsch DU, Cheng L, et al. Segmental reproducibility of retinal blood flow velocity measurements using retinal function imager. Graefes Arch Clin Exp Ophthalmol. 2013;251(12):2665–2670. doi:10.1007/s00417-013-2360-1 [CrossRef]
  22. Witkin AJ, Alshareef RA, Rezeq SS, et al. Comparative analysis of the retinal microvasculature visualized with fluorescein angiography and the retinal function imager. Am J Ophthalmol. 2012;154(5):901–907.e2. doi:10.1016/j.ajo.2012.03.052 [CrossRef]
  23. Jayadev C, Jain N, Sachdev S, Mohan A, Yadav NK. Utility of noninvasive imaging modalities in a retina practice. Indian J Ophthalmol. 2016;64(12):940–943. doi:10.4103/0301-4738.198850 [CrossRef]
  24. Traustason S, La Cour M, Larsen M. Retinal vascular oximetry during ranibizumab treatment of central retinal vein occlusion. Br J Ophthalmol. 2014;98(9):1208–1211. doi:10.1136/bjophthalmol-2013-304580 [CrossRef]
  25. Landa G, Garcia PMT., Rosen R.B.Correlation between retina blood flow velocity assessed by Retinal Function Imager and retina thickness estimated by scanning laser ophthalmoscopy/optical coherence tomograph. Ophthalmologica. 2009;223(3):155–161. doi:10.1159/000189819 [CrossRef]
  26. Burgansky-Eliash Z, Bartov E, Barak A, Grinvald A, Gaton D. Blood-flow velocity in glaucoma patients measured with the Retinal Function Imager. Curr Eye Res. 2016;41(7):965–970. doi:10.3109/02713683.2015.1080278 [CrossRef]
  27. Lopes de Faria JM, Andreazzi Duarte D, Larico Chavez RF, Arthur AM, Arthur R, Iano Y. Reliability and validity of digital assessment of perifoveal capillary network measurement using high-resolution imaging. Br J Ophthalmol. 2014;98(6):726–729. doi:10.1136/bjophthalmol-2013-304100 [CrossRef]
  28. Landa G, Garcia PM, Rosen RB. Correlation between retina blood flow velocity assessed by retinal function imager and retina thickness estimated by scanning laser ophthalmoscopy/optical coherence tomography. Ophthalmologica. 2009;223(3):155–161. doi:10.1159/000189819 [CrossRef]
  29. Burgansky-Eliash Z, Nelson DA, Bar-Tal OP, Lowenstein A, Grinvald A, Barak A. Reduced retinal blood flow velocity in diabetic retinopathy. Retina. 2010;30(5):765–773. doi:10.1097/IAE.0b013e3181c596c6 [CrossRef]
  30. Burgansky-Eliash Z, Lowenstein A, Neuderfer M, et al. The correlation between retinal blood flow velocity measured by the retinal function imager and various physiological parameters. Ophthalmic Surg Lasers Imaging Retina. 2013;44(1):51–58. doi:10.3928/23258160-20121221-13 [CrossRef]
  31. Barak A, Burgansky-Eliash Z, Barash H, Nelson DA, Grinvald A, Loewenstein A. The effect of intravitreal bevacizumab (Avastin) injection on retinal blood flow velocity in patients with choroidal neovascularization. Eur J Ophthalmol. 2012;22(3):423–430. doi:10.5301/ejo.5000074 [CrossRef]
  32. Burgansky-Eliash Z, Barash H, Nelson D, et al. Retinal blood flow velocity in patients with age-related macular degeneration. Curr Eye Res. 2014;39(3):304–311. doi:10.3109/02713683.2013.840384 [CrossRef]
  33. Beutelspacher SC, Serbecic N, Barash H, Burgansky-Eliash Z, Grinvald A, Jonas JB. Central serous chorioretinopathy shows reduced retinal flow circulation in retinal function imaging (RFI). Acta Ophthalmol. 2011;89(6):e479–482. doi:10.1111/j.1755-3768.2011.02136.x [CrossRef]
  34. Birger Y, Blumenfeld O, Bartov E, Burgansky-Eliash Z. Reduced retinal blood flow-velocity in severe hyperlipidemia measured by the retinal function imager. Graefes Arch Clin Exp Ophthalmol. 2011;249(10):1587–1590. doi:10.1007/s00417-011-1716-7 [CrossRef]
  35. Gutfreund S, Izkhakov E, Pokroy R, et al. Retinal blood flow velocity in metabolic syndrome. Graefes Arch Clin Exp Ophthalmol. 2013;251(6):1507–1513. doi:10.1007/s00417-013-2325-4 [CrossRef]
  36. Landa G, Amde W, Haileselassie Y, Rosen RB. Cilioretinal arteries in diabetic eyes are associated with increased retinal blood flow velocity and occurrence of diabetic macular edema. Retina. 2011;31(2):304–311. doi:10.1097/IAE.0b013e3181e91108 [CrossRef]
  37. Arbel Y, Sternfeld A, Barak A, et al. Inverse correlation between coronary and retinal blood flows in patients with normal coronary arteries and slow coronary blood flow. Atherosclerosis. 2014;232(1):149–154. doi:10.1016/j.atherosclerosis.2013.10.033 [CrossRef]
  38. Landa G, Jangi AA, Garcia PM, Rosen RB. Initial report of quantification of retinal blood flow velocity in normal human subjects using the Retinal Functional Imager (RFI). Int Ophthalmol. 2012;32(3):211–215. doi:10.1007/s10792-012-9547-z [CrossRef]

Subjects' Demographic and Clinical Characteristics


Subject Age (Years) Gender Studied Eye Systemic Hypertension Diabetes Mellitus Other Systemic Disease(s) Fellow Eye Disease(s) Other Disease(s) Affecting the Studied Eye Lens Status BCVA at Diagnosis Ellapsed Time From Diagnosis to RFI Testing (Months) Previous Tretament(s) Ellapsed Time From Last Treatment to RFI Testing (Months) BCVA at RFI Testing IOP at RFI Testing (mm Hg) BP at RFI Testing (mm Hg) HR at RFI Testing (bpm)
CRVO 1 60 F OS Y N Hypercholesterolemia, TIA None None Pseudophakic 20/50 73 IVTA × 9, IVA × 8, IVE × 16 2 20/100 (+2) 11 122/71 62
2 57 M OD N Y Hyperlipidemia None None Phakic 20/20 24 IVA × 9 2 20/40 (−1) 22 122/68 69
3 59 M OD N N Arthritis, lower-leg varices CIN CIN Pseudophakic 20/20 81 Laser F × 1; IVA × 5 2 20/30 (−2) 16 135/85 73
4 47 M OS N N None POAG POAG Phakic 20/150 0 None 0 20/150 16 144/92 79
5 53 M OS Y N Hypercholesterolemia POAG POAG Pseudophakic 20/300 62 IVTA × 3, IVA × 4, PPV+EL 7 20/20 15 144/94 59
6 50 M OS Y Y Hypercholesterolemia None NS +1 Phakic 20/300 (+1) 4 IVA × 3 2 20/200 20 150/80 84
7 79 F OD N Y Arthritis POAG, ERM POAG, trace PCO Pseudophakic CF'6 6 IVA × 3; IVTA × 1 1 20/400 16 130/67 87
8 55 M OS N N Low-grade prostate cancer, forehead osteoma None ERM Phakic Unknown 5 IVA × 4 2 20/200 19 133/75 60
BRVO 9 66 F OD Y N Hyperlipidemia, multiple clipped cerebral aneurysms POAG POAG Pseudophakic 20/200 (−1) 32 IVA × 9; IVTA × 5 2 20/400 13 131/79 61
10 63 M OD Y N Hypercholesterolemia None None Phakic 20/20 (−3) 10 IVA × 1 6 20/15 13 134/81 65
11 77 M OD Y N Hypercholesterolemia, hyperlipidemia, arthritis None Conjunctival malignancy Pseudophakic 20/100 116 Focal grid × 1, IVTA × 3, IVA × 21, IVE × 6 2 20/30 (−3) 22 155/81 55
12 73 F OD Y Y Cardiac insufficiency None None Pseudophakic 20/200 22 IVA × 1 17 20/40 15 116/55 59
13 56 F OS Y N Hyperlipidemia None NS +1 Phakic 20/800 60 Laser F × 2, IVA × 5 2 20/100 (−1) 10 115/65 68

Velocity Results Among the Patient Groups

Group Mean Arterioles Velocity (SD) Mean Venules Velocity (SD) Mean Arterioles Number (SD) Mean Venules Number (SD) Total Arterioles Segmented Total Venules Segmented P Value
Healthy 4.23 (1.04) 2.88 (0.93) 10.3 (3.7) 11.7 (4.1) 93 106
CRVO 2.84 (1.21) 1.64 (0.51) 11.25 (5.5) 12 (4.1) 90 109 < .001**
BRVO Overall 3.35 (1.33) 2.65 (1.37) 7.8 (6.1) 10 (1.5) 41 51 < .001***
BRVO Occluded 2.67 (1.43) 1.60 (0.41) 3.0 (2.97) 3.0 (1.82) 17 17 < .001**

BFV Profiles of the Study Subjects


Subject Mean Arterioles Velocity Total Segmented Arterioles Mean Venules Velocity Total Segmented Venules BCVA at RFI Testing
CRVO 1 2.31 13 1.43 15 20/100 (+2)
2 2.51 21 1.62 23 20/40 (−1)
3 5.07 8 1.99 14 20/30 (−2)
4 3 17 2.07 11 20/150
5 2.56 9 1.9 10 20/20
6 3.08 8 1.59 13 20/200
7 1.28 4 1.17 11 20/400
8 2.86 10 1.49 15 20/200
BRVO 9 2.18 13 2.07 12 20/400
10 3.76 15 4.13 9 20/15
11 3.82 5 2.63 8 20/30 (−3)
12 4.77 6 1.87 10 20/40
13 None None 2.86 11 20/100 (−1)
Healthy 14 4.21 12 3.01 10 20/20
15 5.08 12 3.89 14 20/20
16 4.72 10 3.37 9 20/20
17 3.64 17 2.17 22 20/20
18 4.33 12 2.97 10 20/20
19 5.11 11 2.93 10 20/20
20 2.86 7 2.23 11 20/20
21 3.41 8 1.88 9 20/20
22 4.38 4 3.85 11 20/20
Authors

From Bascom Palmer Eye Institute, Miller School of Medicine, University of Miami, Miami (TRC, GMS, JT, DCD, WES); and the Department of Ophthalmology, Semmelweis University, Faculty of Medicine, Budapest, Hungary (GMS).

Dr. DeBuc is a member of the scientific advisory board for Optical Imaging. The remaining authors report no relevant financial disclosures.

This study was supported in part by NIH R01EY020607, NIH Center Grant No. P30-EY014801, and by an unrestricted grant to the University of Miami from Research to Prevent Blindness.

Address correspondence to William E. Smiddy, MD, Bascom Palmer Eye Institute, University of Miami Miller School of Medicine, 900 NW 17th St, Office 272, Miami, FL 33136; email: wsmiddy@med.miami.edu.

Received: January 27, 2017
Accepted: April 28, 2017

10.3928/23258160-20170928-04

Sign up to receive

Journal E-contents