Repeatability and Reproducibility of Retinal Blood Flow Measurement Using a Doppler Optical Coherence Tomography Flowmeter in Healthy Subjects

Transcription

1 Retina Repeatability and Reproducibility of Retinal Blood Flow Measurement Using a Doppler Optical Coherence Tomography Flowmeter in Healthy Subjects Tomofumi Tani, 1 Young-Seok Song, 1 Takafumi Yoshioka, 1 Tsuneaki Omae, 1 Akihiro Ishibazawa, 1 Masahiro Akiba, 2 and Akitoshi Yoshida 1 1 Department of Ophthalmology, Asahikawa Medical University, Asahikawa, Hokkaido, Japan 2 R&D Division, Topcon Corporation, Tokyo, Japan Correspondence: Young-Seok Song, Department of Ophthalmology, Asahikawa Medical University, Midorigaoka Higashi , Asahikawa, , Japan; ysong@asahikawa-med.ac.jp. Submitted: December 29, 2016 Accepted: April 10, 2017 Citation: Tani T, Song Y-S, Yoshioka T, et al. Repeatability and reproducibility of retinal blood flow measurement using a Doppler optical coherence tomography flowmeter in healthy subjects. Invest Ophthalmol Vis Sci. 2017;58: DOI: / iovs PURPOSE. To evaluate the repeatability and reproducibility of retinal blood flow (RBF) measurements in humans by using new auto-alignment and measurement software in a commercially available Doppler optical coherence tomography (DOCT) system. METHODS. The DOCT flowmeter assessed the intrasession repeatability and the intersession and interobserver reproducibility of the RBF measurements. For intrasession repeatability, the coefficients of variation (CVs) of five repeated RBF measurements were calculated at the retinal arteries and veins in 20 normal eyes of 20 healthy volunteers. For intersession reproducibility, two sets of three measurements obtained by one observer on 2 different days were compared. For interobserver reproducibility, two sets of three measurements obtained by two observers on the same day were compared. Intraclass correlation coefficients (ICCs) also were used to evaluate the repeatability and reproducibility. The relevance of the DOCT flowmeter and laser Doppler velocimetry (LDV) also was assessed. RESULTS. Regarding intrasession repeatability, the ICC of the RBF exceeded 0.90 in arterioles and venules (ICC: and 0.970, respectively). The CVs of the RBF in the arterioles and venules were 6.0% 6 3.4% and 8.8% 6 5.1%, respectively. The intersession and interobserver RBF values had high reproducibility in the arterioles (ICC: and 0.993, respectively) and venules (ICC: and 0.986, respectively). The RBF measured with the DOCT flowmeter was correlated strongly with LDV in the arterioles (r ¼ 0.76; P < 0.001). CONCLUSIONS. The DOCT flowmeter had good reproducibility in the arterioles and venules and precisely measured the RBF as compared to the LDV in the arterioles. Keywords: retinal blood flow, Doppler OCT flowmeter, laser Doppler velocimetry, retinal blood velocity profile Evaluation of the ocular circulation is highly important for gaining an understanding of the physiologic and pathologic features in a number of retinal diseases. 1 Although many devices are used to measure retinal ocular blood flow that is, the blue field entoptic technique, microsphere method, laser Doppler flowmeter, laser Doppler velocimetry (LDV), oxygen measurements, retinal vessel analyzer, color Doppler imaging, and laser speckle technique 2 we have shown that LDV is a reliable, noninvasive, and clinically useful tool for evaluating the retinal circulation in humans, based on the absolute vessel diameter and blood velocity. 3,4 Considering these advantages, we have measured the retinal blood flow (RBF) in several retinal diseases such as diabetic retinopathy (DR) in type 1 diabetes mellitus (DM), 5 DR in type 2 DM, 6 branch retinal vein occlusion, 7 and age-related macular degeneration. 8 However, the RBF measurements obtained with the LDV method have a number of disadvantages, that is, the required RBF should be considered as the parabolic flow distribution to calculate the absolute values of blood velocity, 9 fine alignment of the laser beam, and maintenance of strict eye fixation for measurements. Therefore, it is difficult to measure the RBF at the bifurcation or the margin of the optic nerve head because of uncertainty regarding the parabolic flow distribution. LDV also is limited to use in cooperative patients. A novel velocimetry technique using optical coherence tomography (OCT) technology, namely, Doppler OCT (DOCT), has been developed recently OCT can detect not only morphologic images but also a Doppler shift of reflected light, which provides information about flow and movement. 13,14 Although the technique facilitates measurements of the absolute RBF values in humans, with good repeatability, those DOCT methods have some shortcomings. In such methods, it is possible to measure the retinal venous blood flow velocity but not the arterial velocity. Another research group 15 has reported the results with the swept-source based OCT flowmeter in which both the arterial and venous total flow are assessed by using their prototype OCT. We have previously developed a DOCT instrument with novel software, referred to as a segmental-scanning method, 16 which enables simultaneous measurement of the RBF in the retinal arterioles and adjacent venules during one cardiac cycle. We have previously reported the accuracy of the measurements in in vitro glass capillaries. We then evaluated the reproducibility of the blood velocity measurements in the retinal arterioles and Copyright 2017 The Authors iovs.arvojournals.org j ISSN: This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.

2 Newly Developed Doppler OCT Flowmeter IOVS j June 2017 j Vol. 58 j No. 7 j 2892 venules in anesthetized cats. Recently, a new DOCT instrument has been developed that can be used in a clinical setting, namely, the DOCT flowmeter (Yoshida A, et al. IOVS 2016;57:ARVO E-Abstract 5922). The goal of the current study was to determine the repeatability and reproducibility of RBF measurements with the DOCT flowmeter in healthy subjects and the relevance of the arteriolar data when the DOCT flowmeter was compared with the LDV. MATERIALS AND METHODS Subjects and Ethics A total of 20 healthy volunteers (7 men, 13 women; age range, years) with no systemic or ophthalmic disease with mild refractive errors (1.00 and 3.75 diopters of spherical equivalent) participated in this study. This study adhered to the guidelines approved by the ethics committee of our institution and the tenets of the Declaration of Helsinki. Each subject provided written informed consent before enrollment after receiving a complete explanation of the study design and protocol. All volunteers underwent an ophthalmologic examination, including review of the medical history, slit-lamp biomicroscopy, intraocular pressure (IOP) measurement, and funduscopic examination using a 90-diopter lens. The axial length of each eye was measured with an A-mode ultrasound system (IOL- Master 500; Carl Zeiss Meditec, Jena, Germany). The mean arterial blood pressure (MABP) and heart rate (HR) were measured with an electronic sphygmomanometer (EP-88Si; Colin, Tokyo, Japan). The IOP was monitored by applanation tonometry (Haag Streit, Bern, Switzerland). The ocular perfusion pressure was calculated as 2/3MABP IOP. The pupil was dilated with one drop of 0.4% tropicamide (Mydrin M; Santen Pharmaceuticals, Osaka, Japan). The subjects were asked to abstain from drinking coffee and alcohol and smoking for 24 hours before the examination. Each subject rested for 10 minutes in a quiet room before the examination began. After the pupils were dilated, the subjects were required to rest in the sitting position for at least 5 minutes in a quiet dimly lit room in which the temperature was maintained at 258C, after which their BP and HR were measured in the left arm. The RBF measurements were assessed after the BP, HR, and IOP measurements. The relatively large arteries (>80 lm and sufficiently far from the bifurcations) chosen for measurement had relatively straight segments one disc diameter away from the optic disc. RBF Measurements An LDV system (Canon Laser Blood Flowmeter, model CLBF 100; Canon, Tokyo, Japan) was used to measure the blood flow in the superior branch of the first-order major temporal retinal vessels. The detailed system methodology has been described previously. 9 Briefly, the retinal LDV system allows noninvasive measurement of the absolute values of the red blood cells flowing in the centerline of the vessel, based on bidirectional LDV. 9 The mean retinal blood velocity (V mean ) was defined as the V of the averaged maximal speed during one cardiac cycle. Computer analysis of the signal produced by the arterial image on the array sensor automatically determined the retinal artery diameter (D). 9 The D measurements are corrected for the axial length (operator input) and refractive error of the eye, which is measured by the CLBF itself. The RBF was calculated as RBF ¼ V mean 3 Area, where V mean is calculated as V mean ¼ V of the averaged maximal speed/2, and area is the cross-sectional area of the retinal artery at the LDV measurement site. 9 Our DOCT flowmeter is based on a commercially available spectral-domain OCT system (3D OCT-1 Maestro; Topcon Corp., Tokyo, Japan) operated at an 800-nm wavelength range. 16 Only the software was modified for blood flow imaging, where extra image-processing software and quantification software were newly developed to measure blood flow. In DOCT, the flow velocity v(z) can be derived from the Doppler shift incurred by the moving blood: vz ¼ DUðz; sþ k 0 4p n s 1 cos h; ð1þ where z is the depth location, DU is the phase difference at the same depth location between the adjacent profiles after Fourier transform, k0 is the center wavelength, n is the refractive index of blood, s is the time interval between the adjacent profiles, and h is the Doppler angle between the flow vector and the incident probe beam. Besides measurement of the phase difference, the Doppler angle h must be known to calculate the velocity by using Equation 1. A detailed description has been published previously. 16 In the retina, most blood vessels run nearly parallel to the retinal surface except at the area around the optic nerve head. As h approaches 908, the measured velocity becomes very sensitive to the accuracy of h. To minimize this potential velocity measurement error, our Doppler blood flow measurements were performed where the Doppler angle h is considerably less than 908, that is, ~808. Automated alignment software was integrated to seek the proper Doppler angle. Such auto-alignment software traces the boundary of the internal limiting membrane in real time along the blood vessel to determine the Doppler angle. If the Doppler angle was out of appropriate range, the OCT beam was shifted farther away from the pupillary center. This alignment was repeated until the estimated Doppler angle was calculated in a certain range. The accurate Doppler angle estimation routine then was run where a pair of B-scan images was captured with a difference of 100 lm across the blood vessels. This measurement was repeated eight times to define the Doppler angle where it was chosen by using the median value among the eight repeated measurements. Since the time interval for capturing a pair of B- scan images is only 10 ms, most of the eye motion can be ignored. Obtaining the median value among eight repeated imaging makes our measurement more reliable and accurate for rejecting the unwanted value. DOCT imaging was performed for 2 seconds, followed by color fundus imaging. The blood vessels were detected automatically and identified from the OCT structured and phase images. The vessel diameter also was measured by using OCT phase images. Protocols To assess intrasession repeatability, 20 eyes of 20 subjects were recruited. Measurement of the right or left eye was determined by a random number table. One experienced observer (TY, observer 1) performed the DOCT flowmeter measurements for each subject. The RBF measurements of the arterioles and venules in either the superior or inferior temporal sites were repeated five times by using the follow-up mode in the software. The observer was masked to the results of the retinal circulatory parameters. The coefficients of variation (CVs) for the retinal circulatory parameters, retinal vessel D, maximal and averaged retinal blood velocity during one cardiac cycle (V maximum and V average ), and RBF were determined. The intersession reproducibility of the DOCT flowmeter measurements was evaluated. The same observer performed three measurements on 2 different days at each of the 20

3 Newly Developed Doppler OCT Flowmeter IOVS j June 2017 j Vol. 58 j No. 7 j 2893 arterioles and venules (days 1 and 2). Both measurements were performed in the morning under fasting conditions. The observer was masked to the results of the retinal circulatory parameters. The correlations for the different days of the D, V average, V maximum, and RBF were determined. The interobserver reproducibility of the DOCT flowmeter measurements was evaluated. Observer 1 and another inexperienced observer (TO, observer 2) measured the same sites of the same 20 eyes three times on the same day. A minimum of 10 minutes was allowed between measurements. The observers were masked to the results of the retinal circulatory parameters. The correlation between the two observers measurements was determined and the differences in the D, V average, V maximum, and RBF were determined. In the last part of this study, the correlations between the DOCT flowmeter and LDV were evaluated in the arterioles in the same 20 subjects. After an experienced observer (Y-SS, observer 3) performed the LDV measurements, observer 1 performed the DOCT flowmeter measurements at the same vessel sites of the LDV on the same day as those measured by experienced observer 4 (TT). The observers were masked to the results of the retinal circulatory parameters. A minimum of 10 minutes was allowed between the LDV and DOCT flowmeter measurements. The correlation of the measurements between the two different methods was determined in the RBF. Statistical Analysis All data are expressed as the mean 6 standard deviation (SD). Pearson s correlation analysis, determination of the CV, and Bland-Altman plotting were performed to assess the validity and reproducibility. 17 Intergroup differences were analyzed by using the paired t-test. P < 0.05 was considered significant. To evaluate the agreement in the repeatability of the intrasession measurements and the reproducibility of the intersession measurements and the interobserver measurements, the intraclass correlation coefficients (ICCs) one-way random model for the intrasession measurements and the two-way mixed model for the intersession and interobserver measurements with their 95% confidence intervals on absolute agreement were used to analyze measurement reliability. 18 The ICCs could range from 0 to 1, with a higher value indicating better reliability. An ICC above 0.9 indicated acceptable clinical repeatability. 19 IBM SPSS Statistics for Windows, version 24.0 (IBM Corp., Armonk, NY, USA) was used for statistical analyses. RESULTS Subjects and Images of DOCT Flowmeter Table 1 summarizes the systemic parameters of the population on the 2 study days. There were no significant differences in any systemic parameters between the 2 days. Figure 1 shows the imaging results from a healthy young subject, using our DOCT flowmeter. The OCT intensity and phase images are shown in Figures 1A and 1B, where the phase image was represented as a pseudocolor code. We confirmed that the blood velocity profiles of the superior temporal retinal arterioles and venules were obtained simultaneously by the DOCT flowmeter in a healthy young subject (Figs. 1C, 1D). Intrasession Repeatability Table 2 summarizes the results of intrasession repeatability of the arterioles and venules measured by DOCT flowmeter (n ¼ 20). In the arterioles, the Ds ranged from 80.9 to lm (mean, 94.5 lm). The V averages ranged from 9.2 to 21.3 mm/s (mean, 15.5 mm/s). The V maximums ranged from 18.1 to 44.4 TABLE 1. Systemic Parameters of the Population on the 2 Study Days Measurements Day 1, n ¼ 20 Day 2, n ¼ 20 P Value MABP, mm Hg IOP, mm Hg OPP, mm Hg HR, beats/min Data are expressed as the mean 6 SD.OPP,ocularperfusionpressure. mm/s (mean, 30.7 mm/s). The RBFs ranged from 3.2 to 14.2 ll/min (mean, 7.8 ll/min). The ICCs of the D, V average, V maximum, and RBF exceeded 0.9 (0.977, 0.957, 0.943, and 0.994, respectively). The CVs for the D, V average, V maximum, and RBF of the measured arterioles were 3.4% 6 2.1%, 8.8% 6 6.2%, 10.8% 6 3.9%, and 6.0% 6 3.4%, respectively. In the venules, the Ds ranged from 98.9 to lm (mean, lm). The V averages ranged from 5.6 to 21.2 mm/s (mean, 13.1 mm/s). The V maximums ranged from 7.2 to 26.3 mm/s (mean, 16.1 mm/s). The RBFs ranged from 4.9 to 19.4 ll/min (mean, 9.8 ll/min). The ICCs of the D, V average, V maximum, and RBF exceeded 0.9 (0.973, 0.977, 0.982, and 0.944, respectively). The CVs for the D, V average, V maximum, and RBF were 4.2% 6 2.7%, 8.8% 6 3.3%, 8.5% 6 3.6%, and 8.8% 6 5.1%, respectively. Intersession Reproducibility The DOCT flowmeter measurements obtained by the same examiner on 2 different days in the same 20 arterioles and venules were compared. The arteriolar correlations among the D, V average, V maximum, and RBF are shown in Figure 2A. Pearson s correlation analysis showed that the arteriolar circulatory parameters on day 1 were correlated strongly with those of day 2, that is, D (r ¼ 0.99; P < ), V average (r ¼ 0.90; P < ), V maximum (r ¼ 0.85; P < ), and RBF (r ¼ 0.92; P < ). In the Bland-Altman plots of intersession measurements, most values ranged within a mean 6 2 SDs of all retinal circulatory parameters (Fig. 2A). The ICCs of the D, V average, V maximum, and RBF exceeded 0.9 (0.986, 0.904, 0.919, and 0.980, respectively) (Table 3). The venular correlations among the D, V average, V maximum, and RBF measured by the DOCT flowmeter are shown in Figure 2B. Pearson s correlation analysis also showed a strong correlation in measurements between day 1 and day 2, that is, D(r ¼ 0.98; P < ), V average (r ¼ 0.98; P < ), V maximum (r ¼ 0.96; P < ), and RBF (r ¼ 0.98; P < ). In the Bland-Altman plots of intersession measurements, most values ranged within a mean 6 2 SDs of all retinal circulatory parameters (Fig. 2B). The ICCs of the D, V average, V maximum, and RBF exceeded 0.9 (0.975, 0.983, 0.962, and 0.982, respectively) (Table 3). Interobserver Reproducibility The DOCT flowmeter measurements performed by two observers on the same day at the same 20 arterioles and venules were compared. The arteriolar correlations among the D, V average, V maximum, and RBF are shown in Figure 3A. Pearson s correlation analysis showed that the arteriolar circulatory parameters of observer 1 were correlated strongly with those of observer 2, that is, D (r ¼ 0.97; P < ), V average (r ¼ 0.93; P < ), V maximum (r ¼ 0.96; P < ), and RBF (r ¼ 0.99; P < ). In the Bland-Altman plots of intersession measurements, most values ranged within a mean 6 2 SDs of all retinal circulatory parameters (Fig. 3A). The ICCs of the D, V average, V maximum, and RBF exceeded 0.9 (0.974, 0.929, 0.959, and 0.993, respectively) (Table 3).

4 Newly Developed Doppler OCT Flowmeter IOVS j June 2017 j Vol. 58 j No. 7 j 2894 FIGURE 1. Simultaneous measurement of the RBF in the retinal arterioles (A) and venules (V) in a healthy young subject. (A) A fundus image. The green bar indicates the measured location. (B) A phase image with color coding. (C) A velocity profile image of the red blood cells in the horizontal center of an artery at a given moment. (D) Blood velocity profiles of the retinal arterioles and venules measured simultaneously by DOCT flowmeter. TABLE 2. Results of Intrasession Repeatability of Arterioles and Venules Measured on DOCT Flowmeter Arterioles (n ¼ 20) Venules (n ¼ 20) Measurements Value ICC Value ICC CVs, % CVs, % Diameter, lm Averaged velocity, mm/s Maximum velocity, mm/s Blood flow, ll/min Data are expressed as the mean 6 SD.

5 Newly Developed Doppler OCT Flowmeter IOVS j June 2017 j Vol. 58 j No. 7 j 2895 FIGURE 2. (A) The left panels show the correlation of the D, V average, V maximum, and RBF between two occasionally different measurements by one observer. The right panels show the correlation of the D, V average, V maximum, and RBF in the arterioles in the Bland-Altman plots. (B) The left panels show the correlation of the D, V average, V maximum, and RBF between two occasionally different measurements by one observer; and the right panels show the correlation of the D, V average, V maximum, and RBF in the venules in the Bland-Altman plots. The venular correlations between the D, V average, V maximum, and RBF are shown in Figure 3B. Pearson s correlation analysis showed that the arteriolar circulatory parameters of observer 1 were correlated strongly with those of observer 2, that is, D (r ¼ 0.99; P < ), V average (r ¼ 0.98; P < ), V maximum (r ¼ 0.97; P < ), and RBF (r ¼ 0.99; P < ). In the Bland-Altman plots of intersession measurements, most values ranged within a mean 6 2 SDs of all retinal circulatory parameters (Fig. 3B). The ICCs of the D, V average, V maximum, and RBF exceeded 0.9 (0.986, 0.979, 0.973, and 0.986, respectively) (Table 3). Comparison of the DOCT Flowmeter and LDV Results in the Arterioles The RBF measured with the DOCT flowmeter ( ll/ min) was similar to that obtained by LDV ( ll/min) (paired t-test, P ¼ 0.35). The strong correlation (r ¼ 0.76; P <

6 Newly Developed Doppler OCT Flowmeter IOVS j June 2017 j Vol. 58 j No. 7 j 2896 TABLE 3. Reproducibility of Intersession and Interobserver of Arterioles and Venules Intersession Interobserver Measurements A (n ¼ 20) V (n ¼ 20) A (n ¼ 20) V (n ¼ 20) ICC ICC ICC ICC Diameter Averaged velocity Maximum velocity Blood flow A, arterioles; V, venules ) between the RBF measurements (n ¼ 20) is shown in Figure 4. DISCUSSION The current study showed the good repeatability of the intrasession measurements and the reproducibility of the intersession and interobserver measurements of the RBF, using the newly developed DOCT flowmeter in healthy subjects. This study also found that absolute RBF values measured with the DOCT flowmeter were similar to the RBF values of the same arterioles measured by LDV. The RBF measurements using the DOCT flowmeter and LDV also were correlated significantly, indicating that the performance of the DOCT flowmeter is equivalent to that of LDV (Fig. 4). In addition, in our previous LDV study, 4 the CVs of the RBF for intrasession repeatability reported are 13.8% in the arterioles and 12.7% in the venules. Regarding the current DOCT flowmeter, the CVs of the RBF were 6.0% in the arterioles and 8.8% in the venules (Table 2). The Bland-Altman plots showed that the deviations of the arterioles and venules in the RBF values were good in the plots of the intersession and interobserver measurements (Figs. 2, 3). These results also indicated that the DOCT flowmeter provides an accurate representation of the RBF. Until the advent of the DOCT technique, LDV was the only commercially available instrument that measured the absolute value of the RBF in the arterioles and venules. Werkmeister et al. 20 showed for the first time that the DOCT (dual-beam bidirectional Doppler Fourier-domain OCT) measurements of the human venules are well correlated with those obtained with LDV (LDV-5000; Oculix, Inc., Arbaz, Switzerland). No studies have investigated the correlation in the human retinal arterioles. Therefore, the current findings showed for the first time that the DOCT flowmeter measurements of the human arterioles are well correlated with those of the LDV (Fig. 4). No commercially available instrument has used the DOCT technique yet. Although the DOCT flowmeter is an experimental prototype, we are in the process of developing a commercially available product. The LDV method can measure only the maximal blood velocity. In this technique, the blood flow in the retinal vessels should be considered as laminar flow to calculate the absolute values of blood velocity. 9 It is difficult to measure the bifurcation or the margin of the optic nerve head where the flow is considered to be nonlaminar in those areas. In addition, since the LDV technique requires strict focusing on the target vessel, it must be performed by a well-trained examiner. Therefore, there are some difficult situations when measuring the velocity with LDV. However, since the DOCT technique can detect not only the maximal blood velocity but also the FIGURE 3. (A) The left panels show the correlation of the D, V average, V maximum, and RBF between two independent measurements by two observers; and the right panels show the correlation of the D, V average, V maximum, and RBF in the arterioles in the Bland-Altman plots. (B) The left panels show the correlation of the D, V average, V maximum, and RBF between two independent measurements by two observers; and the right panels show the correlation of the D, V average, V maximum, and RBF in the venules in the Bland-Altman plots. entire blood velocity, it does not require parabolic flow definition. There is no restriction on the measurement vessel site. Our DOCT flowmeter system required appropriate detection of the Doppler angle (<868 in target vessels). 16 It is difficult to manually find the proper Doppler angle. We overcame the angle problem by using an automatic setting of the proper Doppler angle using a combination of OCT intensity and phase images. 16 The DOCT flowmeter does not require any specialized technique for observation. After the simple preparations to set the head on the instrument (as with the standard OCT measurements) are completed, the observer selects the target vessels. The RBF is measured automatically. The simplicity of this method makes it suitable for screening large populations and for clinical use. These are the major advantages of the DOCT flowmeter. Another advantage of the DOCT flowmeter is the ability to analyze the blood flow pattern, that is, determination of the degree of laminar flow. Laminar flow is a type of blood flow in which the concentric layers of blood move in parallel through the vessel. The maximal velocity occurs in the center of the vessel and the minimal velocity is at the vessel wall. 21 The laminar flow is associated with shear stress, which is important for regulating blood flow. 22,23 The DOCT flowmeter can visualize the retinal blood velocity profile (RBVP), which is the velocity profile of the red blood cells in the horizontal center of a vessel at a given moment (Fig. 1C). Recently, we first showed an abnormal retinal arterial RBVP during the diastolic phase in a patient with Takayasu s arteritis and aortic insufficiency (Tani T, et al. IOVS 2015;56:ARVO E-Abstract 5892). These results indicate that the aortic valve function is related to the retinal arterial circulation. Therefore, not only absolute RBF but also RBVP analyses are expected to elucidate retinal circulatory disturbances. The DOCT flowmeter has some limitations. First, if the Doppler angle exceeds 868 at the measured vessels, even though the incident beam was shifted automatically, the RBF was not considered reliable in the DOCT flowmeter measurement. Second, in some clinical cases, that is, when a vessel is close to the optic disc or in patients with small pupils, it might be difficult to find the proper Doppler angle. Third, the number of vessels detected in a single scan is limited to a couple of vessels when using the 1-mm scan length. Even when the scan length was expanded to 4 mm, not all vessels

7 Newly Developed Doppler OCT Flowmeter IOVS j June 2017 j Vol. 58 j No. 7 j 2897 FIGURE 4. The correlation between the RBF as assessed with DOCT flowmeter and LDV, respectively, in the arterioles. could be covered in one scan. If the total blood flow measurement is required, multiple DOCT scans are necessary to cover all the vessels. Finally, scanning also might be difficult when there is an opacity in the eye, that is, a vitreous hemorrhage or hard cortical cataract. In conclusion, the current study confirmed the reproducibility of intraobserver, intersession, and interobserver measurements in a newly developed DOCT flowmeter and that the DOCT flowmeter can precisely measure the RBF, compared with the LDV. These data suggest that the DOCT flowmeter is acceptable for measuring the RBF measurements in humans. Thus, the DOCT flowmeter has potential for clinical use and screening RBF changes in a large population. Acknowledgments The authors thank Lynda Charters for manuscript review and also thank Jun Sakai and Shunsuke Nakamura for technical support. Supported by a Grant-in-Aid for Scientific Research from the Ministry of Education, Culture, Sports, Science, and Technology in Japan (15K20242 to TT and 16K20296 to TY). Disclosure: T. Tani, None; Y.-S. Song, None; T. Yoshioka, None; T. Omae, None; A. Ishibazawa, None; M. Akiba, None; A. Yoshida, None References 1. Pournaras CJ, Rungger-Brandle E, Riva CE, Hardarson SH, Stefansson E. Regulation of retinal blood flow in health and disease. Prog Retin Eye Res. 2008;27: Schmetterer L, Kiel J. Ocular Blood Flow. Berlin, Germany: Springer-Verlag Berlin Heidelberg; 2012: Nagaoka T, Ishii Y, Takeuchi T, Takahashi A, Sato E, Yoshida A. Relationship between the parameters of retinal circulation measured by laser Doppler velocimetry and a marker of early systemic atherosclerosis. Invest Ophthalmol Vis Sci. 2005;46: 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: Konno S, Feke GT, Yoshida A, Fujio N, Goger DG, Buzney SM. Retinal blood flow changes in type I diabetes: a long-term follow-up study. Invest Ophthalmol Vis Sci. 1996;37: Nagaoka T, Sato E, Takahashi A, Yokota H, Sogawa K, Yoshida A. Impaired retinal circulation in patients with type 2 diabetes mellitus: retinal laser Doppler velocimetry study. Invest Ophthalmol Vis Sci. 2010;51: Nagaoka T, Sogawa K, Yoshida A. Changes in retinal blood flow in patients with macular edema secondary to branch retinal vein occlusion before and after intravitreal injection of bevacizumab. Retina. 2014;34: Sato E, Feke GT, Menke MN, Wallace McMeel J. Retinal haemodynamics in patients with age-related macular degeneration. Eye (Lond). 2006;20: Nagaoka T, Yoshida A. Noninvasive evaluation of wall shear stress on retinal microcirculation in humans. Invest Ophthalmol Vis Sci. 2006;47: Doblhoff-Dier V, Schmetterer L, Vilser W, et al. Measurement of the total retinal blood flow using dual beam Fourierdomain Doppler optical coherence tomography with orthogonal detection planes. Biomed Opt Express. 2014;5: Leitgeb RA, Werkmeister RM, Blatter C, Schmetterer L. Doppler optical coherence tomography. Prog Retin Eye Res. 2014;41: Wang Y, Bower BA, Izatt JA, Tan O, Huang D. Retinal blood flow measurement by circumpapillary Fourier domain Doppler optical coherence tomography. J Biomed Opt. 2008;13:

8 Newly Developed Doppler OCT Flowmeter IOVS j June 2017 j Vol. 58 j No. 7 j Wang XJ, Milner TE, Nelson JS. Characterization of fluid flow velocity by optical Doppler tomography. Opt Lett. 1995;20: Zhao Y, Chen Z, Saxer C, Xiang S, de Boer JF, Nelson JS. Phaseresolved optical coherence tomography and optical Doppler tomography for imaging blood flow in human skin with fast scanning speed and high velocity sensitivity. Opt Lett. 2000; 25: Baumann B, Potsaid B, Kraus MF, et al. Total retinal blood flow measurement with ultrahigh speed swept source/fourier domain OCT. Biomed Opt Express. 2011;2: Nagaoka T, Tani T, Song YS, et al. Evaluation of retinal circulation using segmental-scanning doppler optical coherence tomography in anesthetized cats. Invest Ophthalmol Vis Sci. 2016;57: Bland JM, Altman DG. Statistical methods for assessing agreement between two methods of clinical measurement. Lancet. 1986;1: Shrout PE, Fleiss JL. Intraclass correlations: uses in assessing rater reliability. Psychol Bull. 1979;86: Kramer MS, Feinstein AR. Clinical biostatistics, LIV: the biostatistics of concordance. Clin Pharmacol Ther. 1981;29: Werkmeister RM, Palkovits S, Told R, et al. Response of retinal blood flow to systemic hyperoxia as measured with dualbeam bidirectional Doppler Fourier-domain optical coherence tomography. PLoS One. 2012;7:e Guyton AaJH. Textbook of Medical Physiology. Philadelphia, PA: W.B. Saunders Company; 2000: Ishibazawa A, Nagaoka T, Takahashi T, et al. Effects of shear stress on the gene expressions of endothelial nitric oxide synthase, endothelin-1, and thrombomodulin in human retinal microvascular endothelial cells. Invest Ophthalmol Vis Sci. 2011;52: Korenaga R, Ando J, Tsuboi H, et al. Laminar flow stimulates ATP- and shear stress-dependent nitric oxide production in cultured bovine endothelial cells. Biochem Biophys Res Commun. 1994;198: