Review Article. Use of laser speckle flowgraphy in ocular blood flow research. Introduction. Basic technology

Transcription

1 Review Article Use of laser speckle flowgraphy in ocular blood flow research Tetsuya Sugiyama, 1 Makoto Araie, 2 Charles E. Riva, 3 Leopold Schmetterer 4,5 and Selim Orgul 6 for the Ocular Blood Flow Research Association 1 Department of Ophthalmology, Osaka Medical College, Osaka, Japan 2 Department of Ophthalmology, University of Tokyo Graduate School of Medicine, Tokyo, Japan 3 Department of Ophthalmology, University of Bologna, Bologna, Italy 4 Centre for Biomedical Engineering and Physics, Medical University of Vienna, Vienna, Austria 5 Division of Ophthalmo-Pharmacology, Department of Clinical Pharmacology, Vienna, Austria 6 University Eye Clinic, University of Basel, Basel, Switzerland ABSTRACT. Laser speckle flowgraphy (LSFG) allows for the quantitative estimation of blood flow in the optic nerve head, choroid, retina and iris in vivo. It was developed to facilitate the non-contact analysis of ocular blood flow in living eyes, utilizing the laser speckle phenomenon. The technique uses a fundus camera, a diode laser, an image sensor, an infrared charge-coupled device (CCD) camera and a high-resolution digital CCD camera. Normalized blur (NB), an approximate reciprocal of speckle contrast, represents an index of blood velocity, and shows a good correlation with tissue blood flow rates determined with the microsphere method in the retina, choroid or iris, as well as blood flow rates determined with the hydrogen gas clearance method in the optic nerve head. The square blur ratio (SBR), another index for quantitative estimation of blood velocity, is proportional to the square of the NB. The SBR is theoretically a more exact measurement which is proportional to velocity, whereas the NB is an approximation. Normalized blur was calculated in earlier versions of LSFG because of technical limitations; the SBR is used in current versions of the LSFG instrument. As these values are in arbitrary units, they should not be used to make comparisons between different eyes or different sites in an eye. Clinical protocols, calibration, evaluation procedures and possible limitations of the LSFG technique are described and the results of ocular blood flow studies using LSFG are briefly summarized. The LSFG method is suitable for monitoring the time course of change in the tissue circulation at the same site in the same eye at various intervals, ranging from seconds to months. Unresolved issues concern the effect of pupil size on measurement results, the effects of various stimulations, and how to measure choroidal and retinal blood flow velocity separately without using the blue-component of argon laser. Key words: iris laser speckle method ocular blood flow optic nerve head physiology retina Acta Ophthalmol. 2010: 88: ª 2009 The Authors Journal compilation ª 2009 Acta Ophthalmol doi: /j x Introduction The present article describes laser speckle flowgraphy (LSFG), one of the methods for quantitative in vivo estimation of circulation in the optic nerve head (ONH), choroid and retina. The importance of the ocular microcirculation, for example in the pathogenesis of glaucoma, has been described in detail elsewhere (Flammer et al. 2002; Okuno et al. 2004; Harris et al. 2005). Here, we describe the basic technology and application of LSFG, including its possible limitations and unresolved questions. Basic technology Underlying physical principles and devices available In principle, LSFG assesses circulation in ocular tissues using the laser speckle phenomenon. The laser speckle phenomenon is an interference event observed when coherent light sources, such as lasers, are scattered by a diffusing surface. The speckle pattern which appears under the illumination of laser irradiation can only be described statistically. In accordance with the movement of blood cells (i.e. blood flow) in the tissue, the structure of the pattern varies rapidly, depending on blood flow velocity. 723

2 Fercher & Briers (1981) first presented pictures of the velocity distribution of red blood cells in the retina by means of laser speckle photography. However, their method only allowed for semi-quantitative estimation of the retinal circulation and did not enable analysis of changes over time. Recently Tamaki et al. (1994, 1995, 1997a) developed a new apparatus for non-contact, two-dimensional and quantitative analysis of ocular blood flow in living eyes. One of the most useful statistics is the standard deviation of the intensity distribution of the speckle pattern (d), which is equal to the mean intensity (<I>) under an ideal condition (where a perfectly coherent light source and a perfect diffuser are involved), but is less than <I> under non-ideal conditions. The fundamental statistical properties of these timevarying speckles can be studied by analysing the space time correlation function of the speckle intensity fluctuation. When the laser exposure time is constant, the reciprocal of speckle contrast, d <I>, correlates with the ratio of the correlation time of the temporal fluctuations in intensity (s c ) to exposure time (T) and indicates the velocity of moving substances (Fercher & Briers 1981; Briers & Fercher 1982) (blood velocity when the laser is projected onto living tissue). Speckle contrast is calculated as follows: sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi d c <I> 1 s c 2T 2T 1 exp s c ð1þ where c is the constant given by the laser exposure time (T) (Ohtsubo & Asakura 1976). The blood velocity, v, is determined thus: on which the laser beam is focused is observed by an infrared CCD camera and a high-resolution digital CCD camera is used to measure the retinal vessel diameter and record fundus photographs (Figs 1 and 2). The scattered laser light is imaged on an image sensor corresponding to a field area of mm or mm, according to the magnification of the fundus camera (30- or 45-degree whole visual angle), on the human fundus, on which a speckle pattern appears. The scanning speed of the image sensor is 98 scans second in an earlier model (Tamaki et al. 1994, 1995) and 512 scans second in the current model. In accordance with the movement of blood cells in the tissue, the structure of Fig. 1. The laser speckle flowgraphy system. the pattern varies rapidly: the greater the blood cell velocity, the greater the rate of variation. Each successive scan of the image sensor results in a different profile of output signal intensity. The diode laser is used for measurements of circulation in the ONH, choroid or iris. To measure the iridal circulation, the maximum laser output is increased from 2 mw on the corneal surface, which is suitable for measurements of the fundus, to approximately 3 mw, and the optical system (the focusing distance) of the camera is modified so that the iris image is optimally imaged on an image sensor (Tomidokoro et al. 1998). V1 <I>2 d 2 ð2þ A fundus camera (TRC-NW5S; Topcon Corp., Tokyo, Japan) is equipped with a diode laser (wavelength 808 nm), image sensor ( pixels, BASIS; Canon Corp., Tokyo, Japan), an infrared charge-coupled device (CCD) camera and a high-resolution digital CCD (LMD-1000; Victor Company of Japan, Tokyo, Japan) camera. A diode laser and an image sensor are used for the laser speckle measurements. The area of the fundus Fig. 2. Schematic view of the laser speckle flowgraphy system. 724

3 To measure the retinal microcirculation, a blue-component argon laser (wavelength 488 nm, maximum power 3 mw, SLL; JDS Uniphase Corp., San Jose, CA, USA) is substituted for the diode laser (Tamaki et al. 1994). Blue-component argon laser is mostly scattered in the superficial retina and absorbed in the retinal pigment epithelium. In Dutch rabbit eyes, normalized blur (NB) values obtained by using a blue-component of argon laser mostly reflected the retinal circulation and the signal was only minimally affected by the choroidal circulation (Tamaki et al. 1994). The laser power is adjusted to the minimum level at which the image sensor can receive a usable signal. The greatest power actually used for measurement is 0.2 mw. In an earlier model, blood velocity was calculated as follows. The pulse height of the output signal obtained at the horizontally x-th and vertically y-th pixel point (x, y; x =1,2,3,, 100, y =1,2,3,, 100) in the k-th scan (k; k =1,2,3,, 98) is defined as I (x, y, k). Normalized blur (x, y) is calculated as follows: NBðx; yþ I mean ðx; yþ ¼ ð1=98þ P98 jði mean ðx; yþ Iðx; y; kþþj where: k¼1 I mean ðx; yþ ¼ P 98 k¼1 ð3þ Iðx; y; kþ 98 : ð4þ The denominator of eqn (3) is used as an approximation of standard deviation (SD), d in eqn (2), and the NB value is used as an approximation of <I> 2 d 2 for the following reasons. Firstly, the high-performance arithmetical and control units were not readily available at the time the earlier model was developed and, thus, there were practical difficulties in calculating <I> 2 d 2 at each pixel every seconds, as the current model does. Secondly, although NB is an approximation of <I> 2 d 2, it shows a good linear correlation (r = , n =1215) with the mean velocity of blood flowing through a glass capillary tube (calculated from the blood flow rate generated by a calibrated peristaltic pump) (Nagahara et al. 1999) and with the speed of rotation of a ground-glass disc (calculated from the velocity of diffusing particles, which are models of blood cells, on the glass) (r = 0.99, n = 18), when the blood velocity or the velocity of diffusing substances is relatively low (10 60 mm second or mm second) (Tamaki et al. 1994, 1995; Nagahara et al. 1999). Thirdly, the NB value shows good correlation with tissue blood flow rate determined with the microsphere method in the retina, choroid and iris (r = 0.59, p < 0.001; r = 0.60, p < 0.001; r = 0.61, p = , respectively), and with blood flow rate determined with the hydrogen gas clearance method in the ONH (r = 0.92, p < 0.01) in rabbit eyes (Tamaki et al. 1994, 1995, 1996a, 2003; Sugiyama et al. 1996; Tomidokoro et al. 1998; Tomita et al. 1999; Takayama et al. 2003). Agreement between the NB value and tissue blood flow determined by the invasive methods described above is thought to rely on the fact that the tissue blood velocity is relatively low. In the current model, <I> 2 d 2 is calculated every second by incorporating a high-performance arithmetical and control unit into the apparatus (Tamaki et al. 1997a). <I> 2 d 2 is calculated for each pixel as follows: I mean ðx; yþ ¼ P 64 k¼1 Iðx; y; kþ 64 ð5þ where I (x, y, k) is the pulse height of the output signal obtained at the horizontally x-th and vertically y-th pixel (x, y; x =1,2,3,, 100; y =1,2, 3,, 100) in the k-th scan (k; k =1, 2, 3,, 64). d 2,ofI(x, y, k), is calculated as follows: P 64 ½Iðx; y; kþš 2 d 2 k¼1 ¼ fi mean ðx; yþg 2 64 ð6þ Thus, <I> 2 d 2 is calculated as (I mean [x, y]) 2 d 2. The square blur ratio (SBR) at the x-th and y-th pixel, SBR (x, y), is defined as follows: SBR ðx; yþ ¼C fi mean ðx; yþg 2 =d 2 Offset ð7þ where C is a constant for calibration obtained by an experiment using a rotating ground-glass disc at known speeds or model capillary tubes through which blood flows at known mean velocities, and offset is the measured SBR value when there is no blood flow. The SBR at each pixel is divided into 50 colour-coded levels, which are displayed as colour graphics on a colour monitor showing the twodimensional variation in the SBR level over the field of interest. The average SBR level (SBR av ) in any rectangular field of interest on a displayed colour map is calculated and its change over a maximum of 5 seconds is monitored during one measurement. The maximum exposure of the retina with the present apparatus is 90 mw cm 2 for 10 seconds, well below the permissible variable limits (460 mw cm 2 for 10-second exposures) determined by the American National Standards Institute (ANSI) (2000). All SBR data are digitally recorded for later analysis. Scanning using coherent light such as laser light inherently induces a speckle phenomenon, which can theoretically influence the measurement. This effect, however, should be negligible because it is far below the range of effects attributable to the movement of blood cells in ocular tissues. To facilitate stabilization of the subject s head, the instrument is equipped with a band for head fixation. If the fixation of the subject s eye, achieved using an internal target, is precise enough, the software can track the measurement point so that the NB value is obtained from the same site when a masked investigator later analyses the results recorded digitally on a magneto-optical disc. Recently, the CCD for an ordinary CCD camera was installed instead of the area sensor described above; the number of pixels has increased from to or more (Konishi et al. 2002). In 2008 this instrument became commercially available in Japan (Softcare Ltd, Iizuka, Japan). 725

4 Clinical protocol Preparation of the subject Before measurement, adequate dilatation of the subject s pupil is preferable. Because alpha-adrenergic agonists such as phenylephrine may influence circulation in the ONH, choroid or retina (Sugiyama et al. 1992), a muscarinic antagonist (i.e. tropicamide) is recommended for pupil dilatation (see also Reporting). Before starting the measurement, a rest period of 5 10 mins should be scheduled in order to obtain stable systemic haemodynamic conditions, which should be verified by repeated measurements of systemic blood pressure and pulse rate. In addition, as a rise in blood glucose has been reported to increase ONH blood flow (Kida et al. 2001, 2007), clinical studies should be performed after a 2-hour fasting period and a 6-hour period of abstinence from coffee and alcohol. Given these research protocols, it may be advisable to instruct subjects and patients to refrain from having a heavy meal or doing heavy exercise the day before haemodynamic measurements. Evaluation procedure The fundus camera must be aligned with the patient s (dilated) pupil in order to obtain an optimal fundus picture on an infrared CCD camera. This Fig. 4. A real-time measurement of optic nerve head circulation (square blur ratio values) using the laser speckle flowgraphy. The software tracked the measurement field (the open square), although a small eye movement occurred during measurement. can be ensured by watching the fundus picture on the monitor. The subject s eye should be positioned by means of an internal fixation target so that the site of interest is in the central area of +the fundus image. When taking measurements in the ONH rim tissue, it is recommended that a rim area free from any visible vessels (usually the temporal region) is selected. The operator will then start the measurement. The blood velocity in the selected region is continuously calculated, displayed and recorded on a magnetooptical disc. Typical images representative of measurements of ONH circulation are shown in Figs 3 and 4. Potential stimuli applicable during measurement The effects of potential stimuli, such as flicker stimulation or hyperoxia, on the SBR or NB in LSFG have not been studied and there are no published reports on the types of stimuli potentially applicable during LSFG measurement. Specific parameters Fig. 3. Colour-coded map showing the circulation (square blur ratio values) of the ocular fundus around the optic nerve head on the display of the laser speckle flowgraphy instrument. Vascular bed(s) studied If we use a diode laser as the probing beam, it is impossible to measure retinal and choroidal circulation separately, other than in areas free of any retinal vessels, where the laser speckle represents only the choroidal 726

5 circulation (e.g. the foveal area in the human or primate fundus or the area outside the wing-medullated area in the rabbit fundus). When an area free of visible retinal vessels is measured in primate eyes, the contribution of the retinal circulation is thought to be < 10% (Isono et al. 1997). Thus, this system allows for the investigation of circulation in the ONH, choroid and the choroid retina and of blood velocity through retinal veins of small diameters (Nagahara et al. 1999). The measurement point for analysis can be anywhere within a mm or mm area (45-degree or 30-degree visual angles of the fundus camera, respectively). As mentioned under Basic technology, iris circulation can also be measured if the focusing distance of the system and the laser power are modified (Tomidokoro et al. 1998). In addition, retinal circulation can be measured in animals if a blue-component argon laser is substituted for the diode laser (Tamaki et al. 1994, 1996b, 1999a, 2003). It is difficult to apply this system in humans because argon illumination causes too much glare to allow for adequate steady fixation during the measurements, even when an area outside the vascular arcade is being measured. The current system has been used in vascular beds with low blood flow velocity, such as capillary beds or retinal veins with small diameters, although SBR values can theoretically be applied in higher-velocity vascular beds, such as major retinal vessels. An in vitro experiment demonstrated that SBR values showed a linear correlation with the velocity of particles diffusing at 200 mm second (Konishi et al. 2002). We regret that no previous studies have addressed the upper limit of the blood velocity that can be measured using the laser speckle method. Measured parameters The main outcome variables of this method are the SBR or NB values of the selected region in arbitrary units. Both variables are originally quantitative indices of blood flow velocity, but NB values have also been shown to demonstrate a significant and good correlation with blood flow data simultaneously determined with the hydrogen gas clearance method in the ONH or the coloured microspheres technique in the iris, choroid or retina, as mentioned above. Although no direct comparison between SBR values and results obtained by the coloured microspheres technique or hydrogen gas clearance method in the same site of an eye has been published, SBR values obtained in vitro (Fujii et al. 1996) and in vivo (Nagahara et al. 2003, unpublished results) were very well correlated with NB values obtained in vitro and in vivo, respectively. Methods of calculation are described in Basic technology: underlying physical principles and devices available. Different sites in the same eye or in different eyes should have different C and Offset values in eqn (7). Although it is impossible to determine C and Offset values at each site of each eye, it is assumed that C and Offset values are similar at the same site in the same eye. Thus, the C and Offset values themselves should have little effect on the extent of relative change in the SBR values obtained in the same site in the same eye. However, for this reason, the values obtained with this technology should not be used to compare velocity between fellow eyes, different eyes or different sites in one eye. By contrast, these values are suited to monitor the time course of change in the tissue blood velocity at the same site in the same eye at various intervals ranging from seconds to months (Tamaki et al. 1996a, 1996b, 1999a, 2003; Tomidokoro et al. 1998; Tomita et al. 1999, 2000; Takayama et al. 2003). These outcomes can also be expressed as percentage changes from baseline values. Calibration Preparation of the device The instrument is calibrated using a ground-glass disc rotating at a certain speed because the spectral distribution of its speckle resembles that of the speckle obtained from the ocular fundus. Reproducibility in healthy subjects Reproducibility of the LSFG in healthy subjects has been published previously (Tamaki et al. 1997a; Nagahara et al. 1999). In a group of 12 eyes in six healthy volunteers, the coefficients of reproducibility ( jm 1 m 2 j=fðm 1 þ m 2 Þ=2g) of 1-min and 24-hour interval measurements of NB in the ONH rim tissue were 11.7 ± 3.3% and 13.0 ± 3.0%. Measurements of NB in the choroid retina were 8.7 ± 1.5% and 9.7 ± 2.5% (mean ± standard error of the mean [SEM]; n = 12). The data indicate that reproducibility is slightly worse for the ONH than for the choroid retina (Tamaki et al. 1997a). In another group of 18 eyes in nine healthy volunteers, the coefficients of reproducibility for measurements of NB in the ONH rim tissue taken at 21-day intervals were 8 18% for each time of measurement (11.00 hours, hours, hours, hours) (Tamaki et al. 1997b). By contrast, the coefficient of reproducibility for 5-min interval measurements of NB in retinal veins in 16 normal human eyes was 2.5 ± 0.9% (Nagahara et al. 1999), which is obviously better than those for the choroid retina and the ONH. Reproducibility in patients The reproducibility of measurements in the ocular fundus of a diseased eye, such as an eye with extensive chorioretinal atrophy, in which the reflectivity of laser light is expected to differ greatly from that in a normal ocular fundus, has not yet been examined. However, the reproducibility of measurements in eyes with glaucoma does not appear to differ significantly from that in normal eyes (Tomita et al. 1999; Tamaki et al. 2001a; Okuno et al. 2004). Main limitations One of the major limitations of the instrument is that the quality of the measurements is mainly dependent on the clarity of the ocular media. Thus, measurements in patients with cataract or corneal problems are sometimes difficult to interpret. Furthermore, fixation is crucial to obtain optimal measurements, although, as stated above, the LSFG system s software can adjust for small eye movements. Measurements in subjects with central vision loss or poor visual acuity will result in greater variability. Thus, it is generally recommended that the co-operability of poorer subjects be considered in the sample size calculation when planning a study in patients 727

6 Table 1. Summary of previous reports using laser speckle flowgraphy. Site of measurement Changes in blood flow Condition, drugs or treatment* Optic nerve head Increased Cigarette smoking (habitual smokers); alcohol; glucose; carteolol; betaxolol; latanoprost; unoprostone; nipradilol; latanoprost + carteolol (NTG); lomerizine; nilvadipine (NTG) No change Cigarette smoking (light smokers); timolol; latanoprost + timolol (NTG); trabeculectomy or needling (POAG); scleral buckling and encircling procedure (RRD); diurnal variation Decreased Caffeine; phenylephrine; diurnal variation at night (NTG) Choroid-retina Increased Cigarette smoking habitual smokers); dynamic exercise; unoprostone; stellate ganglion block (Bell s palsy); nilvadipine (NTG) No change Diurnal variation Decreased Cigarette smoking (light smokers); caffeine; scleral buckling and encircling procedure (RRD) *Subjects were normal volunteers unless a parenthesis is added. NTG = normal-tension glaucoma; POAG = primary open-angle glaucoma; RRD = rhegmatogenous retinal detachment. or subjects with impaired vision. Lastly, it must be noted that, as the SBR and NB values are in arbitrary units, it is difficult to make direct comparisons among data obtained from different sites of measurement or in the eyes of different subjects. Reporting Laser speckle flowgraphy is reported to be a useful tool for the assessment of circulation changes in the iris, ONH, choroid and choroid retina. Table 1 summarizes results in humans reported so far, which demonstrate the acute effects of cigarette smoking (Tamaki et al. 1999b, 2004), alcohol consumption (Kojima et al. 2000), caffeine consumption (Okuno et al. 2002), dynamic exercise (Okuno et al. 2006), glucose loading (Kida et al. 2007), various topical drugs (Tamaki et al. 1997b, 1997c, 1999c, 2001b, 2001c; Ishii et al. 2001; Makimoto et al. 2002; Mizuno et al. 2002; Sugiyama et al. 2009; Takayama et al. 2009), oral calcium antagonists (Tomita et al. 1999; Tamaki et al. 2003; Koseki et al. 2008), trabeculectomy (Tamaki et al. 2001a), retinal detachment surgery (Nagahara et al. 2000), stellate ganglion block Nagahara et al. 2001), and diurnal variation (Okuno et al. 2004). Care must be taken in interpreting the results because LSFG provides velocity data in arbitrary units. Therefore, the differences in ONH blood flow between the right and left eyes and between the superior and inferior temporal neuroretinal rims in normal volunteers (Yaoeda et al. 2000a) must be interpreted with caution. Similarly, findings of a weak correlation between blood flow indices as measured by LSFG and scanning laser Doppler flowmetry (Yaoeda et al. 2000b), and of a correlation between ONH rim circulation and functional damage in the corresponding visual field in patients with glaucoma (Yaoeda et al. 2003), need to be carefully interpreted. Others have compared LSFG and indocyanine green angiography performed in volunteers with choroidal diseases with the aim of confirming whether LSFG may be used to evaluate choroidal haemodynamics (Isono et al. 2003). Unresolved issues and open questions Despite many efforts, several aspects of LSFG measurement remain to be clarified. (1) One unresolved question relates to pupil dilatation. Because pupil width is a major determinant of the image quality obtained by the instrument, information on pupil size and how it might relate to measurement is important. No study has systematically looked into the relationship between pupil size and the SBR or NB. Furthermore, we cannot exclude the possibility that topical tropicamide used for pupil dilatation may influence ocular fundus circulation. (2) It is desirable to separately measure choroidal and retinal blood flow without using the blue-component of the argon laser. The incorporation of a confocal scanning system may allow separate estimations of choroidal and retinal circulations. (3) One shortcoming of the current LSFG system is that the obtained SBR or NB values are given in arbitrary units and inter-eye comparison is practically impossible. With the SBR, it should be theoretically possible to measure mean blood flow velocity through major retinal vessels in absolute units, if the results were carefully calibrated (Nagahara et al. 1999), and agreement of the results with those achieved with laser Doppler velocimetry (Yoshida et al. 2003) in the same site in the same eye might be obtained. (4) The current instrument could not measure blood velocity in the iris of pigmented rabbits (Tomidokoro et al. 1998), probably because melanin granules greatly impede the penetration and scattering of laser light, which limits the application of the current system in the iris of Japanese people or in others with heavily pigmented irides. Incorporating a coherent light source with a longer wavelength may improve the feasibility of such measurements. References American National Standards Institute (2000): American National Standard for the Safe Use of Laser (ANSI Z ). Washington, DC: American National Standards Institute. Briers JD & Fercher AF (1982): Retinal blood flow visualization by means of laser speckle photography. Invest Ophthalmol Vis Sci 22: Fercher AF & Briers JD (1981): Flow visualization by means of single-exposure speckle photography. Opt Commun 37: Flammer J, Orgu l S, Costa VP, Orzalesi N, Krieglstein GK, Serra LM, Renard JP & Stefa nsson E (2002): The impact of ocular blood flow in glaucoma. Prog Retin Eye Res 21: Fujii H, Konishi N, Okamoto K & Dobashi Y (1996): Measurement of ocular circulation by laser speckle method. J Eye (Atarashiiganka) 13: Harris A, Rechtman E, Siesky B, Jonescu-Cuypers C, McCranor L & Garzozi HJ (2005): The role of optic nerve blood flow in the pathogenesis of glaucoma. Ophthalmol Clin North Am 18:

7 Ishii K, Tomidokoro A, Nagahara M, Tamaki Y, Kanno M, Fukaya Y & Araie M (2001): Effects of topical latanoprost on optic nerve head circulation in rabbits, monkeys and humans. Invest Ophthalmol Vis Sci 42: Isono H, Kimura Y & Aoyagi K (1997): Analysis of choroidal blood flow by laser speckle flowgraphy. Nippon Ganka Gakkai Zasshi (J Jpn Ophthalmol Soc) 101: Isono H, Kishi S, Kimura Y, Hagiwara N, Konishi N & Fujii H (2003): Observation of choroidal circulation using index of erythrocytic velocity. Arch Ophthalmol 121: Kida T, Oku H, Sugiyama T & Ikeda T (2001): The mechanism and change in the optic nerve head (ONH) circulation in rabbits after glucose loading. Curr Eye Res 22: Kida T, Sugiyama T, Oku H, Harino S & Ikeda T (2007): Plasma endothelin-1 levels depress optic nerve head circulation detected during the glucose tolerance test. Graefes Arch Clin Exp Ophthalmol 245: Kojima S, Sugiyama T, Kojima M, Azuma I & Ito S (2000): Effect of the consumption of ethanol on the microcirculation of the human optic nerve head in the acute phase. Jpn J Ophthalmol 44: Konishi N, Tokimoto Y, Kohra K & Fujii H (2002): New laser speckle flowgraphy system using CCD camera. Opt Rev 9: Koseki N, Araie M, Tomidokoro A, Nagahara M, Hasegawa T, Tamaki Y & Yamamoto S (2008): A placebo-controlled 3-year study of a calcium blocker on visual field and ocular circulation in glaucoma with low normal pressure. Ophthalmology 115: Makimoto Y, Sugiyama T, Kojima S & Azuma I (2002): Longterm effect of topically applied isopropyl unoprostone on microcirculation in the human ocular fundus. Jpn J Ophthalmol 46: Mizuno K, Koide T, Saito N, Fujii M, Nagahara M, Tomidokoro A, Tamaki Y & Araie M (2002): Topical nipradilol: effects on optic nerve head circulation in humans and periocular distribution in monkeys. Invest Ophthalmol Vis Sci 43: Nagahara M, Tamaki Y, Araie M & Fujii H (1999): Real-time blood velocity measurements in human retinal vein using the laser speckle phenomenon. Jpn J Ophthalmol 43: Nagahara M, Tamaki Y, Araie M & Eguchi S (2000): Effects of scleral buckling and encircling procedures on human optic nerve head and retinochoroidal circulation. Br J Ophthalmol 84: Nagahara M, Tamaki Y, Araie M & Umeyama T (2001): The acute effects of stellate ganglion block on circulation in human ocular fundus. Acta Ophthalmol Scand 79: Ohtsubo J & Asakura T (1976): Velocity measurement of a diffuse object by using time-varying speckles. Opt Quant Electron 8: Okuno T, Sugiyama T, Tominaga M, Kojima S & Ikeda T (2002): Effects of caffeine on microcirculation of the human ocular fundus. Jpn J Ophthalmol 46: Okuno T, Sugiyama T, Kojima S, Nakajima M & Ikeda T (2004): Diurnal variation in microcirculation of ocular fundus and visual field change in normal-tension glaucoma. Eye 18: Okuno T, Sugiyama T, Kohyama M, Kojima S, Oku H & Ikeda T (2006): Ocular blood flow changes after dynamic exercise in humans. Eye 20: Sugiyama K, Bacon DR, Cioffi GA, Fahrenback WH & Van Buskirk EM (1992): The effect of phenylephrine on the ciliary body and optic nerve head microvasculature in rabbits. J Glaucoma 1: Sugiyama T, Utsumi T, Azuma I & Fujii H (1996): Measurement of optic nerve head circulation: comparison of laser speckle and hydrogen clearance methods. Jpn J Ophthalmol 40: Sugiyama T, Kojima S, Ishida O & Ikeda T (2009): Changes in optic nerve head blood flow induced by the combined therapy of latanoprost and beta-blockers. Acta Ophthalmol (in press). Takayama J, Tomidokoro A, Ishii K, Tamaki Y, Fukaya Y, Hosokawa T & Araie M (2003): Time course of the change in optic nerve head circulation after an acute increase in intraocular pressure. Invest Ophthalmol Vis Sci 44: Takayama J, Mayama C, Mishima A, Nagahara M, Tomidokoro A & Araie M (2009): Topical phenylephrine decreases blood velocity in the optic nerve head and increases resistive index in the retinal arteries. Eye 23: Tamaki Y, Araie M, Kawamoto E & Fujii H (1994): Non-contact, two-dimensional measurement of retinal microcirculation using laser speckle phenomenon. Invest Ophthalmol Vis Res 35: Tamaki Y, Araie M, Kawamoto E, Eguchi S & Fujii H (1995): Non-contact, two-dimensional measurement of tissue circulation in choroid and optic nerve head using laser speckle phenomenon. Exp Eye Res 60: Tamaki Y, Araie M, Tomita K & Tomidokoro A (1996a): Time change in nicardipine effect on choroidal circulation in rabbit eyes. Curr Eye Res 15: Tamaki Y, Araie M, Tomita K & Tomidokoro A (1996b): Time course of changes in nicardipine effects on microcirculation in retina and optic nerve head in living rabbit eye. Jpn J Ophthalmol 40: Tamaki Y, Araie M, Tomita K, Nagahara M, Tomidokoro A & Fujii H (1997a): Real-time measurement of human optic nerve head and choroid circulation, using the laser speckle phenomenon. Jpn J Ophthalmol 41: Tamaki Y, Araie M, Tomita K, Nagahara M & Tomidokoro A (1997b): Effect of topical beta-blockers on tissue blood flow in the human optic nerve head. Curr Eye Res 16: Tamaki Y, Araie M, Tomita K, Tomidokoro A & Nagahara M (1997c): Effects of topical adrenergic agents on tissue circulation in rabbit and human optic nerve head evaluated with laser speckle tissue circulation analyser. Surv Ophthalmol 42 (Suppl. 1): Tamaki Y, Araie M, Tomita K & Urashima H (1999a): Effects of pranidipine, a new calcium antagonist, on circulation in the choroid, retina and optic nerve head. Curr Eye Res 19: Tamaki Y, Araie M, Nagahara M & Tomita K (1999b): Acute effects of cigarette smoking on tissue circulation in human optic nerve head and choroid-retina. Ophthalmology 106: Tamaki Y, Araie M, Tomita K & Nagahara M (1999c): Effect of topical betaxolol on tissue circulation in the human optic nerve head. J Ocul Pharmacol Ther 15: Tamaki Y, AraieM, Hasegawa T & Nagahara M (2001a): Optic nerve head circulation after intraocular pressure reduction achieved by trabeculectomy. Ophthalmology 108: Tamaki Y, Nagahara M, Araie M, Tomita K, Sandoh S & Tomidokoro A (2001b): Topical latanoprost and optic nerve head and retinal circulation in humans. J Ocul Pharmocol Ther 17: Tamaki Y, Araie M, Tomita K, Nagahara M, Sandoh S & Tomidokoro A (2001c): Effect of topical unoprostone on circulation of human optic nerve head and retina. J Ocul Pharmocol Ther 17: Tamaki Y, Araie M, Fukaya Y, Nagahara M, Imamura A, Honda M, Obata R & Tomita K (2003): Effects of lomerizine, a calcium channel antagonist, on retinal and optic nerve head circulation in rabbits and humans. Invest Ophthalmol Vis Sci 44: Tamaki Y, Araie M, Nagahara M, Tomita K & Matsubara M (2004): The acute effects of cigarette smoking on human optic nerve and posterior fundus circulation in light smokers. Eye 14: Tomidokoro A, Araie M, Tamaki Y & Tomita K (1998): In vivo measurement of iridal circulation using laser speckle phenomenon. Invest Ophthalmol Vis Res 39: Tomita K, Araie M, Tamaki Y, Nagahara M & Sugiyama T (1999): Effects of nilvadipine, a calcium antagonist, on rabbit ocular circulation and optic nerve head circulation in NTG subjects. Invest Ophthalmol Vis Sci 40: Tomita K, Tomidokoro A, Tamaki Y, Araie M, Matsubara M & Fukuya Y (2000): Effects of semotiadil, a novel calcium antagonist, on the retina and optic nerve head circulation. J Ocul Pharmacol Ther 16: Yaoeda K, Shirakashi M, Funaki S, Funaki H, Nakatsue T, Fukushima A & Abe H (2000a): Measurement of microcirculation in optic nerve head by laser speckle flowgraphy in normal volunteers. Am J Ophthalmol 130: Yaoeda K, Shirakashi M, Funaki S, Funaki H, Nakatsue T & Abe H (2000b): Measurement of microcirculation in the optic nerve head by laser speckle flowgraphy and scanning laser Doppler flowmetry. Am J Ophthalmol 129: Yaoeda K, Shirakashi M, Fukushima A, Funaki S, Funaki H, Abe H & Tanabe N (2003): Relationship between optic nerve head circulation and visual field loss in glaucoma. Acta Ophthalmol Scand 81: Yoshida A, Feke GT, Mori F, Nagaoka T, Fujio N, Ogasawara H, Konno S & Mcmeel JW (2003): Reproducibility and clinical application of a newly developed stabilized retinal laser Doppler instrument. Am J Ophthalmol 135: Received on July 6th, Accepted on February 11th, Correspondence: Tetsuya Sugiyama MD, PhD Department of Ophthalmology Osaka Medical College 2-7 Daigaku-machi Takatsuki Japan Tel: Fax: tsugiyama@poh.osaka-med.ac.jp 729