We interfaced color Doppler Fourier domain optical coherence tomography (CD-FDOCT) with a commercial OCT system to perform in vivo studies of human retinal blood flow in real time. FDOCT does not need reference arm scanning and records one full depth and Doppler profile in parallel. The system operates with an equivalent A-scan rate of 25 kHz and allows real time imaging of the color encoded Doppler information together with the tissue morphology at a rate of 2–4 tomograms (40×512 pixel) per second. The recording time of a single tomogram (160×512 data points) is only 6,4ms. Despite the high detection speed we achieve a system sensitivity of 86dB using a beam power of 500µW at the cornea. The fundus camera allows simultaneous view for selection of the region of interest. We observe bi-directional blood flow and pulsatility of blood velocity in retinal vessels with a Doppler detection bandwidth of 12.5 kHz and a longitudinal velocity sensitivity in tissue of 200µm/s.
©2003 Optical Society of America
Ocular perfusion abnormalities have been implicated in the pathogenesis of several important eye diseases including glaucoma, diabetic retinopathy and age related macular degeneration [1,2,3]. Hence, there is considerable interest in new methods for the assessment of ocular blood flow in vivo. Based on the optical Doppler effect techniques have been proposed to study retinal, optic nerve and subfoveal choroidal blood flow . Using bi-directional laser Doppler velocimetry combined with measurement of retinal vessel diameters a method for assessing total retinal blood flow has been proposed  but no information on the velocity profile within the vessel is available. The combination with optical coherence tomography (OCT) gave additional insight by providing cross sectional information on the micrometer scale. Time domain realizations of OCT setups use either the change in heterodyne frequency to calculate the Doppler shift due to moving scattering particles [6,7], or the measurement of the phase change between sequential scans [8,9]. Those methods are already highly advanced with A-Scan rates up to 8 kHz . Recently we introduced a phase sensitive method for measuring Doppler flow using Fourier domain OCT (FDOCT) . Since FDOCT does not need depth scanning and records the full depth information in parallel it allows for high speed detection above 10 kHz A-scan rate as was already demonstrated . Moreover, this method shows increased sensitivity as compared to conventional time domain OCT . In the present contribution we combined the FDOCT system with a commercial fundus camera based scanning apparatus to demonstrate the capability of color Doppler FDOCT (CD FDOCT) to measure in-vivo real-time human retinal tissue perfusion (refractive index of 1,37) with ultra high acquisition speed.
The method of CD FDOCT is based on a local phase analysis of the backscattered sample light. Structural information is obtained via inverse fast Fourier transform of the interference pattern that is recorded as a function of frequency . The transformation result is a complex function and therefore characterized by amplitude and phase. In polar coordinates it reads , where , and tanΦ(z)=Im(Ĩ(z))/Re(Ĩ(z)). The parameter z corresponds to the optical path length difference between reference arm and a particular backscattering sample layer. A non-vanishing phase difference ΔΦ(z) between two recordings at the same transverse position indicates position changes of this layer along the longitudinal axis. Knowing the time difference between adjacent scans τ one can easily calculate the velocity of this change by v(z)=ΔΦ(z,τ)λ (4πτn), where λ is the center wavelength of the light source, and n is the refractive index of the tissue. The time difference τ is inversely proportional to the CCD frame rate which in our case was 25 kHz. The phase change may as well be caused by flowing particles with non-vanishing velocity component in the direction of the illuminating beam. The velocity has an associated Doppler frequency given by fD(z)=2v(z)/λ. The maximal accessible Doppler frequency is limited by the detection speed of the CCD. With τ=40µs and a center wavelength of λ=800nm we have fD,max=12,5 kHz and νmax=±5mm/s. The actual flow speed vf is obtained via vf=v/cos(α) if the exact angle between the incident beam and the flow direction α is known. Normally, the angle α can not be determined in vivo, but absolute velocity information may be obtained by bi-directional detection . The minimum detectable flow velocity is determined by the phase noise in the system. However speckle noise additionally decreases the velocity resolution in tissue. Averaging reduces the speckle noise, thus the formula for calculating the longitudinal velocity component reads
where N is the number of scans at a single transverse point. For the experiments presented here we chose N=4.
We used a fiber based FDOCT setup as shown in Fig. 1. The light source is a superluminescence diode (SUPERLUM; center wavelength at 810nm, FWHM bandwidth of 17nm) that delivers a power of 500µW to the sample. The chopper prevents light from falling onto the CCD and from illuminating the proband eye during the read out cycle of the CCD. According to the American national standard of safe use of lasers (ANSI) such configuration settings would allow for 8h continuous measurement time . The 50/50 fiber coupler splits the light into reference and sample arm. The reference arm mirror together with the focusing lens is mounted on a translation stage for adjusting its relative optical distance to the sample structure. In order to achieve maximal sensitivity the power of the reference arm signal is set close to the saturation level of the CCD detector (6,2nW per pixel) by a neutral density filter. The collimator optics together with the focusing lens approximately matches the dispersion introduced by the sample arm optics. The residual dispersion introduced by the tissue refractive index of the volunteer’s eye is corrected together with the change from wavelength to wavenumber space via software by re-sampling the recorded interference pattern. The sample arm light is coupled into a commercial fundus camera system (Carl Zeiss Meditec Inc., Dublin CA) that allows for a simultaneous view of the probing beam position during the CD FDOCT measurements. At the exit of the interferometer the light passes a beam expander (10×) and is spectrally analyzed via the diffraction grating (1200/mm) and the CCD camera (ANDOR, 1024×250 pixel, pixel size 25×25µm, 1 MHz ADC, fast kinetics mode). The objective lens (achromat f=300mm) images a spectral width of 50nm along the CCD horizontal resulting in a ~0.05nm spectral resolution. The camera records the spectra with an equivalent A-scan rate of 25 kHz, as each spectrum contains the full depth information, i.e. A-scan, along a range of 3,3mm. We achieved a tomogram rate of 4 per second with 160 transversal and 512 longitudinal pixels. Since we use 4 scans at each transversal point for calculating the average flow velocity, the number of transversal points in the color encoded Doppler tomogram reduces to 40. The control scheme is depicted in Fig 1. The chopper triggers the hardware module that synchronizes the CCD camera and the galvo scanners. After recording the spectra that constitute one tomogram the camera transfers the data to the PC (Pentium IV, 2GHz) where it is processed and displayed in a LabView (National Instruments) environment.
Figure 2(b) shows a tomogram taken along the line depicted on the fundus image in Fig 2(a). It is composed of 12 individual tomogram blocks (160×512) that are manually corrected for movement artifacts between adjacent blocks. A single tomogram block is recorded with ultra high speed in only 6,4ms thus they may assumed to be motion artifact free. Lateral motion artifacts that may occur especially during the data transfer of a single tomogram block will still be present. Although the tomogram is taken at an A-scan rate of 25 kHz with a depth range of 3mm we measured a sensitivity of 86dB using a beam power of 500µW at the cornea. The sensitivity was determined by using a well-defined reflecting surface (R=10-4) in the sample arm and measuring the residual signal to noise ratio. The dynamic range in the amplitude tomogram amounts to 36dB (maximum amplitude to noise floor). The red square in Fig. 2(b) indicates the selected region of interest (ROI) for the Doppler analysis that contains two vessel cross sections close to the optic nerve head.
The structure tomograms together with the flow data for the selected ROI’s are displayed with a real time rate of ~ 3 per seconds. It could be easily increased by reducing the number of transverse points. Without displaying the evaluated data the recording can be performed with a tomogram rate of 4 per second for the current settings. In order to reduce residual speckle noise we assessed the heart cycle with an optical pulse plethysmograph placed at the ear lobe and averaged over profiles that correspond to approximately the same phase of the heart cycle. The heart beat frequency of the volunteer was ~85 per minute for all presented measurements. The color encoded Doppler tomogram in Fig. 3(a) represents an average over 10 tomograms. The color code of the Doppler tomogram allows to easily distinguish between artery and vein (right and left vessel in Fig. 3(a), respectively). By viewing the movies in Fig 5 that consist of 10 tomograms of the selected ROI one has additional access to dynamic physiologic processes like the clearly visible pulsations. The graphic in Fig. 3(c) shows cross sections through both vessels in the color encoded Doppler tomogram of Fig. 3(a). The profiles fit well to the parabolic curves (black lines in Fig. 3(c)) as is expected for capillary flow. The longitudinal velocity sensitivity in tissue can be estimated from Fig. 3(c) to 200µm/s by calculating the rms error of the parabolic fit.
It is typical for veins close to the optic nerve head that they show a slight pulsatility which is defined as P=(Vmax,syst -Vmax,dist)/(Vmax,syst+Vmax,dist). Vmax,syst, and Vmax,dist are the maximum velocities for peak systole and end diastole. We chose a ROI that contained a vein of approximately 120µm diameter (blue box in Fig. 2(a)) and recorded a set of tomograms to reconstruct the time course of the capillary flow (Fig. 4(a)). Fig. 4(c) shows profiles that represent averages over 3 profiles each taken at maximum and minimum - systolic and diastolic - flow peak. The pulsatility was found to be 0.25. According to the sampling theorem the heart cycle of ~ 85 beats per minute or 1.3 Hz is already close to the Nyquist limit of 2 Hz given by half the tomogram sampling rate. Higher rates would obviously facilitate the analysis.
In conclusion we demonstrated the ability of CD FDOCT to assess velocity profiles in human retinal vessels in vivo. We achieved an A-scan rate of 25 kHz with a real time tomogram refresh rate of ~ 3 per second and a maximal recording rate of 4 tomograms per second. The color encoded Doppler tomogram allows extraction of flow profiles and monitoring of pulsatile blood flow. The maximum Doppler frequency is currently 12.5 kHz corresponding to half the detection speed of the detector array. The number of transverse points in the flow tomogram is limited by the detector buffer size of the present system of 250 lines. However, there are commercial CCD systems already available that work at rates up to 60 kHz, where the amount of transferred data is only limited by the on board memory of the host computer. The remaining question is whether the transverse scanning apparatus has a small enough step response time to follow the high speed of the detector unit.
We acknowledge the Austrian National Bank (Jubilaeumsfonds grant Nr. 9654), the Austrian Academic Exchange Service OEAD together with the Polish State Committee for Scientific Research (grant 2003/13), the Austrian Fonds zur Foerderung von Wissenschaft und Forschung (FWF grants P14529-PSY, P14218-PSY, FWF Y 159), and the CRAFT program (CRAF-1999-70549) for their financial support. We thank Carl Zeiss Meditec Inc. for providing the OCT system.
References and Links
1. J. Flammer, S. Orgul, V.P. Costa, N. Orzalesi, G.K. Krieglstein, L.M. Serra, J.P. Renard, and E. Stefansson, “The impact of ocular blood flow in glaucoma,” Prog. Retin. Eye Res. 21, 359–393 (2002). [CrossRef] [PubMed]
2. L. Schmetterer and M. Wolzt, “Ocular blood flow and associated functional deviations in diabetic retinopathy,” Diabetologia, 42, 387–405 (1999). [CrossRef]
3. E. Friedman, “A hemodynamic model of the pathogenesis of age-related macular degeneration,” Am.J. Ophthalmol. 124, 677–682 (1997). [PubMed]
4. C.E. Riva, S.D. Cranstoun, J.E. Grunwald, and B.L. Petrig, “Choroidal blood flow in the foveal region of the human disc,” Invest. Ophthalmol. Vis. Sci. 35, 4273–4281 (1994). [PubMed]
5. C.E. Riva, J.E. Grunwald, S.H. Sinclair, and B.L. Petrig, “Blood velocity and volumetric flow rate in human retinal vessels,” Invest.Ophthalmol.Vis.Sci. 26, 1124–1132 (1985). [PubMed]
6. E. Logean, L. Schmetterer, and C. E. Riva, “Optical Doppler velocimetry at various retinal vessel depths by variation of the source coherence length,” Appl. Opt. 39, 2858–2862 (2000). [CrossRef]
7. S. Yazdanfar, A. M. Rollins, and J. A. Izatt, “In vivo imaging of human retinal flow dynamics by color Doppler optical coherence tomography,” Arch. Ophthalmol. 121, 235–239 (2003). [CrossRef] [PubMed]
8. Y. Zhao, Z. Chen, C. Saxer, S. Xiang, J de Boer, and J Nelson, “Phase-resolved optical coherence tomography and optical Doppler tomography for imaging blood f low in human skin with fast scanning speed and high velocity sensitivity,” Opt. Lett. 25, 1358–1360 (2000). [CrossRef]
9. Z. Chen, T. E. Milner, S. Srinivas, X. Wang, A. Malekafzali, M. van Gemert, and J. Nelson, “Noninvasive imaging of in vivo blood flow velocity using optical Doppler tomography,” Opt. Lett. 22, 1119–1121 (1997). [CrossRef] [PubMed]
10. V.X.D. Yang, M. L. Gordon, E. Seng-Yue, S. Lo, B. Qi, J. Pekar, A. Mok, B. C. Wilson, and I. A. Vitkin, “High speed, wide velocity dynamic range Doppler optical coherence tomography (Part II): Imaging in vivo cardiac dynamics of Xenopus laevis,” Opt. Express 11, 1650–1658 (2003); http://www.opticsexpress.org/abstract.cfm?URI=OPEX-11-7-794. [CrossRef] [PubMed]
11. R. Leitgeb, L. Schmetterer, F. Berisha, C. K. Hitzenberger, M. Wojtkowski, T. Bajraszewski, and A. F. Fercher, “Real- time measurements of in-vitro flow by Fourier domain optical coherence tomography,” accepted for publication in Optics Letters.
12. M. Wojtkowski, T. Bajraszewski, P. Targowski, A. Targowski, and A. Kowalczyk, “Real-time in-vivo ophthalmic imaging by ultra fast spectral optical coherence tomography,” SPIE Proc. 4956, 50–58 (2003). [CrossRef]
13. R. A. Leitgeb, C. K. Hitzenberger, and A. F. Fercher, “Performance of Fourier domain vs. time domain optical coherence tomography,” Opt. Express 11, 889–894 (2003). http://www.opticsexpress.org/abstract.cfm?URI=OPEX-11-8-889. [CrossRef] [PubMed]
14. A. F. Fercher, C. K. Hitzenberger, G. Kamp, and S. Y. El-Zaiat, Measurement of intraocular distances by backscattering spectral interferometry,” Opt. Commun. 117, 43–48 (1995). [CrossRef]
16. American National Standards Institute; Safe Use of Lasers; Orlando, Laser Institute of America, ANSI 136.1 - 2000 (2000).