[A new approach for studying the retinal and choroidal circulation].Nippon Ganka Gakkai Zasshi 2004; 108(12):836-61; discussion 862NG
One advantage of advanced computer technology is the high throughput with which the retinal and the choroidal circulation can be evaluated from new aspects. To study the choroidal circulation, we first reevaluated indocyanine green video angiography to improve the visualization of indocyanine green (ICG) images, then applied computer technology to analyze images obtained by an ICG video camera system. We also developed a new instrument to measure oxygen saturation levels in the fundus using spectral retinal imaging technology. I. Choroidal circulation. 1. Reevaluation of ICG video camera system: For this purpose, the bio-chemical nature of ICG was studied. 1) Spectral absorption of ICG: The peak absorption of ICG in distilled water was 780 nm as measured with a spectrophotometer. Its maximum absorption shifted from 780 nm to 805 nm after gradually mixing ICG with human serum protein. Conjugation time of ICG as well as fluorescein sodium with human serum protein was then measured by a stopped flowmeter. It was found that fluorescein sodium conjugated with human serum protein within a few milliseconds, while ICG required more than 600 seconds before equilibrium of the binding was reached. From these observations, we developed a new ICG video system with dual light sources; one, a 780 nm diode laser for the early dye filling phase, and the other, a 805 nm diode laser for the later phase of ICG angiography. 2) Binding properties of ICG in human blood: Blood samples were obtained from three healthy volunteers after intravenous administration of ICG. The resulting plasma samples were fractionated by agarose gel immunoelectrophoresis and polyacrylamide gel DISC electrophoresis. The electrophoretic pattern obtained by each method was observed with an ICG fundus video system. We also studied the affinity of ICG for lipids that are common molecular components of lipoproteins such as high-density lipoprotein (HDL) and low-density lipoprotein (LDL). Four kinds of ICG solutions mixed with phospholipid, free cholesterol, esterified cholesterol, and triacrylglycerol were observed with the ICG fundus video system. Both electrophoretic studies showed that ICG bound intensely to HDL and moderately to LDL, and only the solution with phospholipid fluoresced brightly when observed with the ICG fundus video system. 2. Residual fundus ICG fluorescence: Residual fundus fluorescence observed in the late phase of ICG angiography may be delineated differently in normal subjects and in patients with age-related macular degeneration (ARMD). We performed ICG angiography on 8 normal subjects aged below 36 years (8 eyes), 9 normal subjects aged above 62 years (9 eyes), and 21 patients with ARMD aged 50 to 88 years (37 eyes). The intensity and pattern of fluorescence from angiograms obtained in the ultra-late phase, 24 hours after dye injection, was recorded and analyzed. In the ultra-late phase, 95% of ARMD eyes with choroidal neovascularization (CNV) showed geographic hypofluorescent lesions. These hypofluorescent lesions occurred in 73% of ARMD eyes without CN, while age-matched normal subjects had no hypofluorescent lesions. The mean intensity of fluorescence in the normal elder subject group was significantly higher than that seen in the normal younger subject group. These findings may reflect aging change and bio-distribution of lipid on the Bruch-RPE complex. 3. The early dye filling pattern of the choroid: We performed ICG angiography on 10 healthy young volunteers aged 22 to 26 years (23.4+/-1.3; mean+/-standard deviation) using an improved ICG video camera system. ICG (50 mg) dissolved in 2 ml in distilled water was injected through the antecubital vein. Although the choroidal dye filling varied among subjects, it always began in the macular area. In the 10 subjects, initial dye filling had two patterns: reticular (n=8) and flush (n=2). The choroidal circulation filled completely before the retinal circulation did. Bright fluorescence in the macula and fast blood flow may be correlated with ample blood volume and abundant blood vessels in the macular area. 4. The spreading pattern of ICG fluorescence in the choroid: The ICG images obtained for observing the early dye filling pattern of the choroid were further processed with a computer-assisted image analyzer. Subtracted images were made using the early ICG frames with a time interval of 0.12 second. Ninety frames of time-sequential images for 3 seconds starting from the initial dye appearance in the choroid were prepared to construct an animated image. In the initial phase of eye filling, patchy fluorescence appeared in the fovea. The fluorescence then spread centrifugally in all directions in a wave-like pulsatile manner towards the peripheral fundus with increasing brightness. Thus an animated video of subtracted images allowed us to evaluate flow dynamics at the level of the choriocapillaris. Using this new approach, pathogenic involvement of choroidal circulation in varied chorioretinal diseases such as central serous chorioretinopathy can be studied with precision. II. Retinal circulation. We developed a new device to study the oxygen saturation (OS) levels in a wider fundus area. We call this device a spectral retinal image (SRI) system. We are pursuing the possibility of this instrument being in clinics to evaluate chorioretinal diseases. 1. Introduction of the device: The instrument consists of a Sagnac interferometer that has been mounted on top of a fundus camera, and a software module which consists of an acquisition module and an analysis module. The image acquisition process takes 6 seconds during which the fundus is illuminated by the white incandescent light of the fundus camera at the regular power settings. OS values in each pixel of the fundus image with a 35-degree view can then be estimated from the respective spectrum that is obtained by Fourier-transforming the interferometer signals. Each OS value is represented by a specific color, and each pixel of the fundus image is painted accordingly. 2. Clinical applications. 1) Control study: OS levels of both retinal arteries and veins within a 1 disc diameter (DD) area around the disc were measured from 20 SRIs obtained from 10 healthy volunteers, and were 96.65+/-3.30% and 56.05+/-4.69%, respectively. Then 30 healthy volunteers were recruited for further study in which the OS values were calculated in five retinal regions: (1) juxta-papillary area within 1.75 DD, (2) fovea within 1.0 DD, (3) papillomacular region within 1.0 DD, (4) superior area of the posterior fundus within 1.0 DD, and (5) inferior posterior area of the posterior fundus within 1.0 DD. The OS level of the juxta-papillary area was the highest, while that of the fovea was the lowest and the other three posterior retinal regions were in the middle. Thus OS levels differed at various areas in the retina with statistical significance. In spite of abundant choroidal circulation in the fovea, the overlying retina may have a relative by low oxygen level. As the retinal pigment epithelium may be efficient enough to block the effect of the choroid optically, our results may indicate that the OS levels represent the OS of the retina. 2) Measurement of OS levels in eyes with retinal circulatory disturbances: Eleven eyes of 10 patients with central retinal vein occlusion (CRVO), which showed various degrees of severity, and 4 fellow unaffected eyes of selected patients were examined by both fluorescein angiography (FAG) and the new SRI system. The fluroescein angiograms were correlated with OS maps that were calculated from the SRI. OS grading demonstrated by OS maps correlated well with severity of CRVO, as estimated by FAG. Thus our SRI system is noninvasive with reproducible results, and may prove to be a useful clinical tool to evaluate the degree of retinal ischemia. 3) Measurement of OS levels in eyes with glaucoma: Forty-seven eyes with open angle glaucoma (OAG) of 47 patients and 21 eyes of 21 age-matched normal subjects were recuited for the study. Twelve eyes with low-tension glaucoma (LTG) were included in the OAG eyes and the rest of the OAG eyes had primary OAG. All patients and normal subjects were examined by SRI. Visual field tests for OAG eyes were done with a Humphrey Field Analyzer using the 30-2 program, Swedish interactive threshold argorithm (SITA). OS levels in the retina at 5 different points: superior, inferior, superio- and inferio-temporal, and nasal region within a juxta-papillary area of 200 microm in diameter were calculated from the SRI. OS levels of retinal arteries were also measured and there were no significant differences between OAG and the control group. OAG eyes showed reduced OS levels in the inferio-temporal retina with statistical significance. This observation was more prominent in LTG eyes. The reduced OS levels observed in OAG eyes correlated well with mean deviation (MD) and the sum of total deviation of the 17 points in corresponding areas in the visual field analysis.