- Email: [email protected]
Non-subjective cataract analysis and its application in space radiation risk assessment
Adv. Space Res. Vol. 14, No. 10, pp. (10)493--(10)500, 1994 Copyright ~ 1994 COSPAR Printed in Great Britain. All fights re.served. 0273-1177/94 $7.00 + 0.00
Pergamon
NON-SUBJECTIVE CATARACT ANALYSIS AND ITS APPLICATION IN SPACE RADIATION RISK ASSESSMENT B. Wu, C. Medvedovsky and B. V. Worgul Eye Radiation and Environmental Research Laboratory, Columbia University, Department of Ophthalmology, New York, NY 10032, U.S.A.
ABSTRACT Experimental animal studies and human observations suggest that the question is not whether or not prolonged space missions will cause cataracts to appear prematurely in the astronauts, but when and to what degree. Historically the major impediment to radiation cataract follow-up has been the necessarily subjective nature o f assessing the degree of lens transparency. This has spurred the development of instruments which produce video images amenable to digital analysis. One such system, the Zeiss Scheimpflng slit lamp measuring system (SLC), was incorporated into our ongoing studies of radiation cataractogenesis. It was found that the Zeiss SLC measuring system has high resolution and permits the acquisition of reproducible images of the anterior segment of the eye. Our results, based on about 650 images of the rats lens, and followed over a period of 91 weeks of radiation cataract development, showed that the integrated optical density (IOD) of the lens correlated well with conventional assessment with the added advantages of objectivity, permanent and transportable records and linearity as cataracts become more severe. This continuous data acquisition, commencing with cataract onset, can proceed through more advanced stages. The SLC exhibits much greater sensitivity reflected in a continuously progressive severity despite the artifactual plateaus in staging which occur using conventional scoring methods. Systems such as the Zeiss SLC should be used to monitor astronauts frequent visits to low earth orbit to obtain a longitudinal data-base on the influence of this activity on the lens. INTRODUCTION Perhaps the greatest impediment to the use of cataract as a method of monitoring populations at risk to ionizing radiation has been the heretofore subjective and non-linear methodologies which had to be used in their evaluation. Photography, using the conventional slit-lamp biomicroscope, has had limited use resulting primarily from its insufficient depth of focus and poor reproducibility /1,2/. Since the discovery of Scheimpflug principle, various cameras have been developed to obviate these problems in lens transparency assessment/3-5/. The important instrument which uses this concept is the Zeiss Scbeimpflug SLC system. With this system, one can reliably acquire and objectively assess high-depth of field images of the lens crucial elements in epidemiological or pharmacological studies/6-10/. Since 1986, our laboratory has incorporated this system into our ongoing work on radiation cataractogenesis. While no machine can surpass the experienced ophthalmic investigator in detecting the onset of the first changes in the lens, the SLC system is without peer for objective, linear quantification of the loss of transparency once lens opaciflcation (10)493
(10)494
B. Wu et al.
has begun /11/. We have developed a methodology combining conventional slit-lamp biomicroscopy to monitor the onset of lens changes followed by the use of the SLC for longitudinal assessment/12/. This study compares the quantitative analysis of the SLC expressed as integrated optical density (IOD) with the results of the Merriam/Focht (M/F) method/13/of semiquantitative opacity scoring by classical slit-lamp analysis. It supports the long held contention that the well known saltatory nature of radiation cataract development reflects the subjective scoring methodologies and not the actual status of the lens transparency. This study was based on 650 images and followed up to 91 weeks of radiation cataract development. MATERIALS AND METHODS Descriotion of the Scheimnflu~ Slit Lamp Ima~in~ System (SLC) The Zeiss SLC measuring system is a special slit lamp with an electronic image analyzer (Fig. 1).
Fig. 1. The video-based Zeiss Scheimpflug slit lamp system (SLC). It produces complex, objective data on the anterior ocular media. The optomechanical system is designed to satisfy the Scheimpflug principle (Fig. 2). The measuring data are acquired from an image analyzer and instantly available.
(~
fslit imam
L
opt.axis'/'~it ofslit iUumination~ , S ~
J
~
Fig. 2 An illustration of the Scheimpflug principle. According to Scheimpflug /14/ an image of high depth of field is obtained obliquely to the image forming optics L if the obiect Diane (here, the path of the incident slit image in the anterior section of the eye), the nrincipal olane of the image forming optics L, and the image Diane (here, plane of the receiver surface) intersect in a line S, and the resulting angles correspond to each other.
Cataract Analysis and Radiation Risk Assessment
(10)495
The optomechanical unit allows exact optical repositioning relative to the eye, so that reproducible measurements of precisely the same area (___30 lsm) can be carried out. Upon pressing the foot switch, the slit image of the eye is captured by the video camera, digitally stored in a frame buffer and continuously sent to the on-line image analyzer for processing. The "processed" image is displayed on a monitor, and can be archived on a mass storage device for later retrieval and analysis. One can reproducibly reposition the area of interest to within 30 ~m guaranteeing that the data over time is derived from the same site. This is of paramount importance in long term follow-up studies. In addition to these features, binocular stereoscopic observation of the slit image of the eye is possible in the same way as with a conventional slit lamp. Irradiation of Rats The heads of 19 pigmented Sprague-Dawley rats were exposed to plateau 450 MeV/amu 56Fe ions at the BEVALAC facility of the Lawrence Berkeley laboratory, University of California and to 250 kVp X-rays . The rats, which were 28 day (+1 day) old were lightly anesthetized (i.m. injection 25 mg/kg Ketamine and 5 mg/kg xylazine) and held in a specially designed circular jig. The exposure were administered as acute doses of 25 cGy and 50 cGy for 56Fe ions and 200 cGy and 700 cGy for X-rays. The control group was treated in exactly the same manner (transported to the BEVALAC, anesthetized etc.), but were not irradiated. Examination Procedure. Before examination, the pupils of the rats were maximally dilated using 1% mydriacyl and 2.5% neosynephrine. Rats were lightly anesthetized and then placed on a specially designed platform with a device to slightly proptose the eyeball. Before each examination, the light source was calibrated to adjust any drift the system might incur. The calibration procedure guaranteed that densitometric measurement taken at different times produced equivalent density profiles. In our experiment all images were taken at a f'Lxedvoltage (14V). There was good correlation in density with this voltage because even the images of the most severe lens opacities never exceeded the dynamic range in gray levels, The eye was carefully aligned based on the Purkinji images on the lens. The identical spot image was achieved through real-time positioning of the eye on the monitor. When the eyes were properly aligned, images of four 0.5 mm optical slices were captured and stored at the 0 °, 45 °, 90 °, 135 ° meridians respectively. Linear (profile) densitometry was then performed on the upper and lower portions of each meridian. The same locations and spatial parameters were used for each measurement and compared longitudinally. Hence, while for different animal eyes the measured area might not be precisely the same, for each eye of any individual animal the measuring spot was always identical over the course of the entire experimental period. The height of the scanned area was always 150 microns. A proprietary program was designed to allow transfer of the SLC data to an IBM 80486 computer for scaling of all the images against a standard. Using the integrated optical density (IOD) values for quantitative image analysis, at a fixed distance (2.88 mm) from anterior capsule to anterior cortex, the IOD values were then corrected for background. Slit-Lamp Observations The cataracts were followed over the 91 week period by slit-lamp biomicroscopy. Opacification was scored using a modified version of the Merriam-Focht method/13/. This technique has been widely used for scoring the severity of radiation-induced cataract in many species, including humans/15-18/. The scoring, a semiquantitative technique using a 0-4+ range, depends upon the fact that radiation cataracts develop in a characteristically sequential fashion. The earliest changes consist of vacuoles or diffuse opacities around the central suture in the posterior subcapsular region and are gauged as 1 +. When vacuoles are present but number fewer then 4 a 0.5+ cataract is scored. Progression of the posterior subcapsular region and the early involvement of the anterior
(10)496
B. Wu et al.
subcapsular region is termed the 2 + stage. If fewer than four vacuoles or opacities are observed anteriorly a 1.5 + cataract is recorded. A 3 + stage is noted when the anterior opacities progress and the density of the cataract posteriorly does not allow assessment of the vitreous beyond. If the entire posterior cortex is involved, yet the capsule can still be discerned, a 2.5 + cataract is noted. The 4 + stage is one with complete anterior opacification preventing visualization of the remainder of the lens. If the opacity has not become severe enough to prevent passage of the beam to the posterior region, but has made detailed visualization impossible, a 3.5 + stage is scored. Both SLC and slit lamp data were then statistically analyzed and plotted as means, with +/standard errors, as a function of time (weeks). Each point represents the data from at least six eyes. RESULTS An example of the image recorded for measuring is shown in Fig. 3.
Fig. 3 A Scheimpflug image of the anterior segment of a rat eye illustrating a radiation cataract. A line indicating the region to be measured is also shown. The relative transparency is determined microdensitomatrically as IOD for the upper and lower portions of the lens. For various doses the mean IOD values with the standard errors, are presented (Fig. 4) and compared to the equivalent "staging" values for each. As the cataract developed , the IOD correlated well with the conventional cataract scoring stages. This continuous data acquisition, commencing with cataract onset, can proceed through more severe stages. The SLC analysis exhibits much greater sensitivity reflected in a progressive opacification despite the artifactual plateaus in staging which occur using the conventional cataract scoring methods. For example, in the 700 cGy x-ray group, from 14 to 42 weeks, the cataract stages progressed very slowly from 2.5+ to 3 + , although during this apparent plateau, the IOD values increased in an unintercepted fashion from 1.337 to 2.993. For 50 cGy iron ions, from 20 to 68 weeks, it took 48 weeks for the cataract stage to worsen from 2.0+ to 2.5+, while the IOD, during this period, from 27 weeks more than doubled (0.979 to 62 weeks 2.669). Our data, incidentally, also confirmed our previous findings that despite the large dose groups (X-rays = 700 cGy, 56Fe ions = 50 cGy) the minimum latent period for the onset of cataract, gauged either by Merriam/Focht system or Zeiss SLC system, began after 4 weeks post-irradiation. This
Cataract Analysis and Radiation Risk Assessment
(10)497
interval is believed to reflect the minimum time necessary for a sufficient number of aberrantly differentiated fibers to accumulate in the posterior cortex/19/.
(A) 4.0
4.0
7'00 cGy )(-Ray • 200 cGy X-Ray .~ v Control ,/~-- 0""
0
3.5
3.0
!/
2.5
o
2.0
,.
..o
,~""
1.5
~
1.0
,i~'"
0.5
0.5 0.0
I I
a.o
;
1.5 1.0
3.5
0.0 0
40
20
60
80
O0
TIME (weeks)
(B) 4.0 3.5
O 2'5 cGy Iron I o n s ' • 50 cGy Iron Ions V Control -,-./.
a.o
I 0 1""4
•-
2.5
,o
,'"-'i;
_ ~ "
4.0 3.5
//!
J---•
O E--* 03 2.0 [.., 1.5 L)
- - ~-•""
/
3.0
I I I
2.5
-"
Fig. 4 Results comparing the objective (Zeiss SLC System - solid lines) and semi-quantitative (conventional SL - dashed lines) techniques of the assessment of radiation cataract in pigmented rats. Cataract analysis plotted as a function of time post exposure to 250 kVp Xrays (A) or 450 keV/amu 56Fe ions (B). Each point represents the mean and standard error of data from at least six eyes.
1.0 [.., 0.5 ~_.-
0.0 0
,
20
"
,
,
,
40
80
80
TIME
C.)
0.0 100
(weeks)
DISCUSSION The advent of a new generation of imaging systems exploiting Scheimpflug optics has provided the means for objective, quantitative data acquisition of radiation cataract formation with excellent reproducibility and the advantage of a permanent image record of the cataract itself/20/. Our laboratory has been intensely involved in utilizing the Zeiss Scheimpflug slit lamp imaging system for the purpose of its application to human and animal radiation cataract studies. The data reported here support the superiority of SLC as a more sensitive, objective and quantitative technique compared to the conventional slit-lamp scoring system. Previous efforts to collect the measuring data as pixel values expressed in percent reflectance proved difficult and timeconsuming when attempts were made to compare the data from different eyes. This spurred the development of a special program to scale all these data to the same standard integrated optical density, allowing inter, as well as intra, experimental data comparisons. The other important
(10)498
B. Wu et al.
aspect in the study was the use of anesthesia to permit careful positioning of animals and reduce possible fatigue during Scheimpflug image acquisition. The reproducibility which resulted is essential to useful longitudinal analyses. Also, by anesthetizing the animal, only one operator, instead of two, was needed /10,21/. The reproducibility was made possible by the unique superimposed picture of the system displayed on the screen as an enlarged scale, so that even small animal eyes can be properly oriented /21/. The quality of the image can be checked instantly so that it is possible to store only excellent images, together with simultaneous storage of all relevant data. The experimental results clearly showed that the SLC system represents a significant improvement in examining, following, archiving and quantifying cataract progression. This makes the instrument the system of choice for longitudinal assessment of cataract development in human and in experimental animals. The archival capability, wherein graphic and quantitative data can be readily obtained as printout and/or hard copy, recommends its use for permanently recording lens changes in populations subject to decrements in lenticular transparency due to aging, but especially for atomic bomb survivors, victims of accidental radiation exposure, and those exposed to space radiation. As currently configured, the major disadvantage of this system in animal studies is that the posterior portion of the lens may not be included in the slit image resulting in loss of important data. This is particularly onerous in radiation cataract studies because it is in the posterior subcapsular region where opacities first appear and develop. This important data loss is because rats, for example, have essentially spherical lenses and the pupillary opening, even with maximal dilation, is still relatively small. The solution which is now being incorporated into the system is to reduce the c~ angle from 45 ° to 30 ° so one can get the whole image of the anterior segment no matter what kind of animal is being used. This will have the added advantage of allowing the lenses of even partially dilated human eyes to be imaged. The basic requirements of image analysis are: to keep errors as small as possible; to be methodically simple, readily practicable, and exactly reproducible; to enable uncomplicated and unequivocal interpretations/22/. In order to achieve these goals, we are currently developing new methodologies using an automated image acquisition and image storage system. These include software and hardware changes to further reduce the time required for image capture and storage as well as routines to minimize errors in using this system. ACKNOWLEDGEMENTS Funded by grants EY02648 from the NEI, DE-FG02-90ER61009 from the DOE and NAG-9-256/S2 from NASA. REFERENCES .
S.K. West and H.R. Taylor, The detection and grading of cataract, an epidemiologic perspective. Surv. Ophthalmol. 31, 175-184 (1986).
.
A. Laties, E. Keates, E. Lippa, C. Shear, D. Snavely, M. Tupy-Visich and A.N. Chremos, Field test reliability of a new lens opacity rating system utilizing slit-lamp examination. Lens and Eye Toxicity Res., 6, 443-464 (1989).
.
V. Dragomirescu, O. Hockwin, H.R. Koch and K. Sasaki, Development of a new equipment for rotating slit lamp image photography according to Scheimpflug's principle. Interdiscipl. Topics Geront., 13, 118 (1978).
Cataract Analysis and Radiation Risk Assessment
(10)499
O. Hockwin, S. Lerman and C. Ohrloff, Investigations on lens transparency and its disturbances by microdensitometric analyses of Scheimpflug photographs. Curr. Eye Res., 3, 15-22 (1984).
.
.
R.C. Zeimer and Y. Ogura, The quality and sensitivity of lens scatter densitometry with a computerized Scheimpflug video camera. Lens Res., 4, 251-264 (1987).
.
O. Hockwin, V. Dragomirescu, T. Shibata, H. Laser and A. Wegener, Long term followup examination of experimental cataracts in rats by Scheimpflug photography and densitometry. Graefes Arch. Clin. Exp. Ophthalmol., 222, 20-25 (1984).
.
O. Hockwin, H. Laser and A. Wegener, Investigations on rats eyes with diabetic cataract and naphthalene cataract by Zeiss Scheimpflug measuring system SLC. Graefes Arch. Clin. Exp. Ophthalmol., 224, 502-506 (1986). U. Eckerskorn, O. Hockwin, M. Korff, V. Dragomirescu and H. Laser, Evaluation of possible risk factors in senile cataract by a Scheimpflug photo study of lens morphology. Top. Aging Res. Eur., 6, 37-46 (1986).
.
.
A. Wegener, H. Laser and O. Hockwin, Measurement of lens transparency changes in animals: comparison of the Topcon SL-45 with linear microdensitometry and the Zeiss SLC system, in Concepts in Toxicology, ed. Karger, Basel 1987, vol. 4, pp. 263-275.
10.
A. Wegener, H. Laser and O. Hockwin, Reproducibility studies with the Zeiss SLC system and animal cataract models. Ophthalmic Res. 22 (suppl 1), 18-23 (1990).
11.
G.R. Merriam, Jr., C. Medvedovsky and B.V. Worgul, Scheimpflug image analysis of radiation cataract development. Invest. Ophthal. Vis. Sci., 29, 427 (1988).
12.
C. Medvedovsky, G.R. Merriam, Jr., J. David, H. Rosskothen and B.V. Worgul, An integrated scoring system for experimental cataract assessment. Invest. Ophthalmol. Vis. Sci., 30, 144 (1989).
13.
G.R. Merriam, Jr. and E.F. Focht, A clinical and experimental study of the effect of single divided doses of radiation on cataract production. Tr. Am. Ophthalmol. Soc., 60, 35-52 (1962).
14.
T. Scheimpflug, Der photoperspektograph und seine anwendung. Korrespondenz, 43, 516 (1906).
15.
B.V. Worgul, H. Rothstein, C. Medvedovsky and G.R. Merriam, Jr., Radiation cataractogenesis in the amphibian lens. Ophthalmic Res., 14, 73-82 (1982).
16.
G.R. Merriam, Jr. and B.V. Worgul, Experimental radiation cataract - its clinical relevance. Bull. N.Y. Acad. Med., 59, 372-392 (1982).
17.
B.V. Worgul, C. Medvedovsky, P. Powers-Risius and E. Alpen, Accelerated heavy ions and the lens IV. Biomicroscopic and cytopathological analyses of the lenses of mice irradiated with 600 MeV/amu 56Fe ions. Rad. Res., 120, 280-293 (1989).
18.
D.J. Brenner, C. Medvedovsky, Y. Huang, G.R. Merriam, Jr. and B.V. Worgul, Accelerated heavy particles and the lens VI. RBE studies at low doses. Rad. Res., 128, 73-81 (1991).
Photographische
(10)500
B. Wu et al.
19.
B.V. Worgul, G.R. Merriam, Jr. and C. Medvedovsky, Accelerated heavy particles and the lens II. Cytopathological changes. Invest. Ophthalmol. Vis. Sci., 27, 108-114 (1986).
20.
P.A. Edwards, M.B. Datiles and S.B. Green, Reproducibility study on the Scheimpflug cataract video camera. Curr. Eye Res., 10, 955-960 (1988).
21.
M. Kojima, A. Wegener and O. Hockwin, Imaging characteristics of three cameras using the Scheimpflug principle. OphthalmicRes., 22 (suppl 1), 29-35 (1990).
22.
U. Meinel, Comparison between digital and analog image analysis of Scheimpflug photographs. Ophthalmic. Res., 22 (suppl 1), 71-73 (1990).
Pergamon
NON-SUBJECTIVE CATARACT ANALYSIS AND ITS APPLICATION IN SPACE RADIATION RISK ASSESSMENT B. Wu, C. Medvedovsky and B. V. Worgul Eye Radiation and Environmental Research Laboratory, Columbia University, Department of Ophthalmology, New York, NY 10032, U.S.A.
ABSTRACT Experimental animal studies and human observations suggest that the question is not whether or not prolonged space missions will cause cataracts to appear prematurely in the astronauts, but when and to what degree. Historically the major impediment to radiation cataract follow-up has been the necessarily subjective nature o f assessing the degree of lens transparency. This has spurred the development of instruments which produce video images amenable to digital analysis. One such system, the Zeiss Scheimpflng slit lamp measuring system (SLC), was incorporated into our ongoing studies of radiation cataractogenesis. It was found that the Zeiss SLC measuring system has high resolution and permits the acquisition of reproducible images of the anterior segment of the eye. Our results, based on about 650 images of the rats lens, and followed over a period of 91 weeks of radiation cataract development, showed that the integrated optical density (IOD) of the lens correlated well with conventional assessment with the added advantages of objectivity, permanent and transportable records and linearity as cataracts become more severe. This continuous data acquisition, commencing with cataract onset, can proceed through more advanced stages. The SLC exhibits much greater sensitivity reflected in a continuously progressive severity despite the artifactual plateaus in staging which occur using conventional scoring methods. Systems such as the Zeiss SLC should be used to monitor astronauts frequent visits to low earth orbit to obtain a longitudinal data-base on the influence of this activity on the lens. INTRODUCTION Perhaps the greatest impediment to the use of cataract as a method of monitoring populations at risk to ionizing radiation has been the heretofore subjective and non-linear methodologies which had to be used in their evaluation. Photography, using the conventional slit-lamp biomicroscope, has had limited use resulting primarily from its insufficient depth of focus and poor reproducibility /1,2/. Since the discovery of Scheimpflug principle, various cameras have been developed to obviate these problems in lens transparency assessment/3-5/. The important instrument which uses this concept is the Zeiss Scbeimpflug SLC system. With this system, one can reliably acquire and objectively assess high-depth of field images of the lens crucial elements in epidemiological or pharmacological studies/6-10/. Since 1986, our laboratory has incorporated this system into our ongoing work on radiation cataractogenesis. While no machine can surpass the experienced ophthalmic investigator in detecting the onset of the first changes in the lens, the SLC system is without peer for objective, linear quantification of the loss of transparency once lens opaciflcation (10)493
(10)494
B. Wu et al.
has begun /11/. We have developed a methodology combining conventional slit-lamp biomicroscopy to monitor the onset of lens changes followed by the use of the SLC for longitudinal assessment/12/. This study compares the quantitative analysis of the SLC expressed as integrated optical density (IOD) with the results of the Merriam/Focht (M/F) method/13/of semiquantitative opacity scoring by classical slit-lamp analysis. It supports the long held contention that the well known saltatory nature of radiation cataract development reflects the subjective scoring methodologies and not the actual status of the lens transparency. This study was based on 650 images and followed up to 91 weeks of radiation cataract development. MATERIALS AND METHODS Descriotion of the Scheimnflu~ Slit Lamp Ima~in~ System (SLC) The Zeiss SLC measuring system is a special slit lamp with an electronic image analyzer (Fig. 1).
Fig. 1. The video-based Zeiss Scheimpflug slit lamp system (SLC). It produces complex, objective data on the anterior ocular media. The optomechanical system is designed to satisfy the Scheimpflug principle (Fig. 2). The measuring data are acquired from an image analyzer and instantly available.
(~
fslit imam
L
opt.axis'/'~it ofslit iUumination~ , S ~
J
~
Fig. 2 An illustration of the Scheimpflug principle. According to Scheimpflug /14/ an image of high depth of field is obtained obliquely to the image forming optics L if the obiect Diane (here, the path of the incident slit image in the anterior section of the eye), the nrincipal olane of the image forming optics L, and the image Diane (here, plane of the receiver surface) intersect in a line S, and the resulting angles correspond to each other.
Cataract Analysis and Radiation Risk Assessment
(10)495
The optomechanical unit allows exact optical repositioning relative to the eye, so that reproducible measurements of precisely the same area (___30 lsm) can be carried out. Upon pressing the foot switch, the slit image of the eye is captured by the video camera, digitally stored in a frame buffer and continuously sent to the on-line image analyzer for processing. The "processed" image is displayed on a monitor, and can be archived on a mass storage device for later retrieval and analysis. One can reproducibly reposition the area of interest to within 30 ~m guaranteeing that the data over time is derived from the same site. This is of paramount importance in long term follow-up studies. In addition to these features, binocular stereoscopic observation of the slit image of the eye is possible in the same way as with a conventional slit lamp. Irradiation of Rats The heads of 19 pigmented Sprague-Dawley rats were exposed to plateau 450 MeV/amu 56Fe ions at the BEVALAC facility of the Lawrence Berkeley laboratory, University of California and to 250 kVp X-rays . The rats, which were 28 day (+1 day) old were lightly anesthetized (i.m. injection 25 mg/kg Ketamine and 5 mg/kg xylazine) and held in a specially designed circular jig. The exposure were administered as acute doses of 25 cGy and 50 cGy for 56Fe ions and 200 cGy and 700 cGy for X-rays. The control group was treated in exactly the same manner (transported to the BEVALAC, anesthetized etc.), but were not irradiated. Examination Procedure. Before examination, the pupils of the rats were maximally dilated using 1% mydriacyl and 2.5% neosynephrine. Rats were lightly anesthetized and then placed on a specially designed platform with a device to slightly proptose the eyeball. Before each examination, the light source was calibrated to adjust any drift the system might incur. The calibration procedure guaranteed that densitometric measurement taken at different times produced equivalent density profiles. In our experiment all images were taken at a f'Lxedvoltage (14V). There was good correlation in density with this voltage because even the images of the most severe lens opacities never exceeded the dynamic range in gray levels, The eye was carefully aligned based on the Purkinji images on the lens. The identical spot image was achieved through real-time positioning of the eye on the monitor. When the eyes were properly aligned, images of four 0.5 mm optical slices were captured and stored at the 0 °, 45 °, 90 °, 135 ° meridians respectively. Linear (profile) densitometry was then performed on the upper and lower portions of each meridian. The same locations and spatial parameters were used for each measurement and compared longitudinally. Hence, while for different animal eyes the measured area might not be precisely the same, for each eye of any individual animal the measuring spot was always identical over the course of the entire experimental period. The height of the scanned area was always 150 microns. A proprietary program was designed to allow transfer of the SLC data to an IBM 80486 computer for scaling of all the images against a standard. Using the integrated optical density (IOD) values for quantitative image analysis, at a fixed distance (2.88 mm) from anterior capsule to anterior cortex, the IOD values were then corrected for background. Slit-Lamp Observations The cataracts were followed over the 91 week period by slit-lamp biomicroscopy. Opacification was scored using a modified version of the Merriam-Focht method/13/. This technique has been widely used for scoring the severity of radiation-induced cataract in many species, including humans/15-18/. The scoring, a semiquantitative technique using a 0-4+ range, depends upon the fact that radiation cataracts develop in a characteristically sequential fashion. The earliest changes consist of vacuoles or diffuse opacities around the central suture in the posterior subcapsular region and are gauged as 1 +. When vacuoles are present but number fewer then 4 a 0.5+ cataract is scored. Progression of the posterior subcapsular region and the early involvement of the anterior
(10)496
B. Wu et al.
subcapsular region is termed the 2 + stage. If fewer than four vacuoles or opacities are observed anteriorly a 1.5 + cataract is recorded. A 3 + stage is noted when the anterior opacities progress and the density of the cataract posteriorly does not allow assessment of the vitreous beyond. If the entire posterior cortex is involved, yet the capsule can still be discerned, a 2.5 + cataract is noted. The 4 + stage is one with complete anterior opacification preventing visualization of the remainder of the lens. If the opacity has not become severe enough to prevent passage of the beam to the posterior region, but has made detailed visualization impossible, a 3.5 + stage is scored. Both SLC and slit lamp data were then statistically analyzed and plotted as means, with +/standard errors, as a function of time (weeks). Each point represents the data from at least six eyes. RESULTS An example of the image recorded for measuring is shown in Fig. 3.
Fig. 3 A Scheimpflug image of the anterior segment of a rat eye illustrating a radiation cataract. A line indicating the region to be measured is also shown. The relative transparency is determined microdensitomatrically as IOD for the upper and lower portions of the lens. For various doses the mean IOD values with the standard errors, are presented (Fig. 4) and compared to the equivalent "staging" values for each. As the cataract developed , the IOD correlated well with the conventional cataract scoring stages. This continuous data acquisition, commencing with cataract onset, can proceed through more severe stages. The SLC analysis exhibits much greater sensitivity reflected in a progressive opacification despite the artifactual plateaus in staging which occur using the conventional cataract scoring methods. For example, in the 700 cGy x-ray group, from 14 to 42 weeks, the cataract stages progressed very slowly from 2.5+ to 3 + , although during this apparent plateau, the IOD values increased in an unintercepted fashion from 1.337 to 2.993. For 50 cGy iron ions, from 20 to 68 weeks, it took 48 weeks for the cataract stage to worsen from 2.0+ to 2.5+, while the IOD, during this period, from 27 weeks more than doubled (0.979 to 62 weeks 2.669). Our data, incidentally, also confirmed our previous findings that despite the large dose groups (X-rays = 700 cGy, 56Fe ions = 50 cGy) the minimum latent period for the onset of cataract, gauged either by Merriam/Focht system or Zeiss SLC system, began after 4 weeks post-irradiation. This
Cataract Analysis and Radiation Risk Assessment
(10)497
interval is believed to reflect the minimum time necessary for a sufficient number of aberrantly differentiated fibers to accumulate in the posterior cortex/19/.
(A) 4.0
4.0
7'00 cGy )(-Ray • 200 cGy X-Ray .~ v Control ,/~-- 0""
0
3.5
3.0
!/
2.5
o
2.0
,.
..o
,~""
1.5
~
1.0
,i~'"
0.5
0.5 0.0
I I
a.o
;
1.5 1.0
3.5
0.0 0
40
20
60
80
O0
TIME (weeks)
(B) 4.0 3.5
O 2'5 cGy Iron I o n s ' • 50 cGy Iron Ions V Control -,-./.
a.o
I 0 1""4
•-
2.5
,o
,'"-'i;
_ ~ "
4.0 3.5
//!
J---•
O E--* 03 2.0 [.., 1.5 L)
- - ~-•""
/
3.0
I I I
2.5
-"
Fig. 4 Results comparing the objective (Zeiss SLC System - solid lines) and semi-quantitative (conventional SL - dashed lines) techniques of the assessment of radiation cataract in pigmented rats. Cataract analysis plotted as a function of time post exposure to 250 kVp Xrays (A) or 450 keV/amu 56Fe ions (B). Each point represents the mean and standard error of data from at least six eyes.
1.0 [.., 0.5 ~_.-
0.0 0
,
20
"
,
,
,
40
80
80
TIME
C.)
0.0 100
(weeks)
DISCUSSION The advent of a new generation of imaging systems exploiting Scheimpflug optics has provided the means for objective, quantitative data acquisition of radiation cataract formation with excellent reproducibility and the advantage of a permanent image record of the cataract itself/20/. Our laboratory has been intensely involved in utilizing the Zeiss Scheimpflug slit lamp imaging system for the purpose of its application to human and animal radiation cataract studies. The data reported here support the superiority of SLC as a more sensitive, objective and quantitative technique compared to the conventional slit-lamp scoring system. Previous efforts to collect the measuring data as pixel values expressed in percent reflectance proved difficult and timeconsuming when attempts were made to compare the data from different eyes. This spurred the development of a special program to scale all these data to the same standard integrated optical density, allowing inter, as well as intra, experimental data comparisons. The other important
(10)498
B. Wu et al.
aspect in the study was the use of anesthesia to permit careful positioning of animals and reduce possible fatigue during Scheimpflug image acquisition. The reproducibility which resulted is essential to useful longitudinal analyses. Also, by anesthetizing the animal, only one operator, instead of two, was needed /10,21/. The reproducibility was made possible by the unique superimposed picture of the system displayed on the screen as an enlarged scale, so that even small animal eyes can be properly oriented /21/. The quality of the image can be checked instantly so that it is possible to store only excellent images, together with simultaneous storage of all relevant data. The experimental results clearly showed that the SLC system represents a significant improvement in examining, following, archiving and quantifying cataract progression. This makes the instrument the system of choice for longitudinal assessment of cataract development in human and in experimental animals. The archival capability, wherein graphic and quantitative data can be readily obtained as printout and/or hard copy, recommends its use for permanently recording lens changes in populations subject to decrements in lenticular transparency due to aging, but especially for atomic bomb survivors, victims of accidental radiation exposure, and those exposed to space radiation. As currently configured, the major disadvantage of this system in animal studies is that the posterior portion of the lens may not be included in the slit image resulting in loss of important data. This is particularly onerous in radiation cataract studies because it is in the posterior subcapsular region where opacities first appear and develop. This important data loss is because rats, for example, have essentially spherical lenses and the pupillary opening, even with maximal dilation, is still relatively small. The solution which is now being incorporated into the system is to reduce the c~ angle from 45 ° to 30 ° so one can get the whole image of the anterior segment no matter what kind of animal is being used. This will have the added advantage of allowing the lenses of even partially dilated human eyes to be imaged. The basic requirements of image analysis are: to keep errors as small as possible; to be methodically simple, readily practicable, and exactly reproducible; to enable uncomplicated and unequivocal interpretations/22/. In order to achieve these goals, we are currently developing new methodologies using an automated image acquisition and image storage system. These include software and hardware changes to further reduce the time required for image capture and storage as well as routines to minimize errors in using this system. ACKNOWLEDGEMENTS Funded by grants EY02648 from the NEI, DE-FG02-90ER61009 from the DOE and NAG-9-256/S2 from NASA. REFERENCES .
S.K. West and H.R. Taylor, The detection and grading of cataract, an epidemiologic perspective. Surv. Ophthalmol. 31, 175-184 (1986).
.
A. Laties, E. Keates, E. Lippa, C. Shear, D. Snavely, M. Tupy-Visich and A.N. Chremos, Field test reliability of a new lens opacity rating system utilizing slit-lamp examination. Lens and Eye Toxicity Res., 6, 443-464 (1989).
.
V. Dragomirescu, O. Hockwin, H.R. Koch and K. Sasaki, Development of a new equipment for rotating slit lamp image photography according to Scheimpflug's principle. Interdiscipl. Topics Geront., 13, 118 (1978).
Cataract Analysis and Radiation Risk Assessment
(10)499
O. Hockwin, S. Lerman and C. Ohrloff, Investigations on lens transparency and its disturbances by microdensitometric analyses of Scheimpflug photographs. Curr. Eye Res., 3, 15-22 (1984).
.
.
R.C. Zeimer and Y. Ogura, The quality and sensitivity of lens scatter densitometry with a computerized Scheimpflug video camera. Lens Res., 4, 251-264 (1987).
.
O. Hockwin, V. Dragomirescu, T. Shibata, H. Laser and A. Wegener, Long term followup examination of experimental cataracts in rats by Scheimpflug photography and densitometry. Graefes Arch. Clin. Exp. Ophthalmol., 222, 20-25 (1984).
.
O. Hockwin, H. Laser and A. Wegener, Investigations on rats eyes with diabetic cataract and naphthalene cataract by Zeiss Scheimpflug measuring system SLC. Graefes Arch. Clin. Exp. Ophthalmol., 224, 502-506 (1986). U. Eckerskorn, O. Hockwin, M. Korff, V. Dragomirescu and H. Laser, Evaluation of possible risk factors in senile cataract by a Scheimpflug photo study of lens morphology. Top. Aging Res. Eur., 6, 37-46 (1986).
.
.
A. Wegener, H. Laser and O. Hockwin, Measurement of lens transparency changes in animals: comparison of the Topcon SL-45 with linear microdensitometry and the Zeiss SLC system, in Concepts in Toxicology, ed. Karger, Basel 1987, vol. 4, pp. 263-275.
10.
A. Wegener, H. Laser and O. Hockwin, Reproducibility studies with the Zeiss SLC system and animal cataract models. Ophthalmic Res. 22 (suppl 1), 18-23 (1990).
11.
G.R. Merriam, Jr., C. Medvedovsky and B.V. Worgul, Scheimpflug image analysis of radiation cataract development. Invest. Ophthal. Vis. Sci., 29, 427 (1988).
12.
C. Medvedovsky, G.R. Merriam, Jr., J. David, H. Rosskothen and B.V. Worgul, An integrated scoring system for experimental cataract assessment. Invest. Ophthalmol. Vis. Sci., 30, 144 (1989).
13.
G.R. Merriam, Jr. and E.F. Focht, A clinical and experimental study of the effect of single divided doses of radiation on cataract production. Tr. Am. Ophthalmol. Soc., 60, 35-52 (1962).
14.
T. Scheimpflug, Der photoperspektograph und seine anwendung. Korrespondenz, 43, 516 (1906).
15.
B.V. Worgul, H. Rothstein, C. Medvedovsky and G.R. Merriam, Jr., Radiation cataractogenesis in the amphibian lens. Ophthalmic Res., 14, 73-82 (1982).
16.
G.R. Merriam, Jr. and B.V. Worgul, Experimental radiation cataract - its clinical relevance. Bull. N.Y. Acad. Med., 59, 372-392 (1982).
17.
B.V. Worgul, C. Medvedovsky, P. Powers-Risius and E. Alpen, Accelerated heavy ions and the lens IV. Biomicroscopic and cytopathological analyses of the lenses of mice irradiated with 600 MeV/amu 56Fe ions. Rad. Res., 120, 280-293 (1989).
18.
D.J. Brenner, C. Medvedovsky, Y. Huang, G.R. Merriam, Jr. and B.V. Worgul, Accelerated heavy particles and the lens VI. RBE studies at low doses. Rad. Res., 128, 73-81 (1991).
Photographische
(10)500
B. Wu et al.
19.
B.V. Worgul, G.R. Merriam, Jr. and C. Medvedovsky, Accelerated heavy particles and the lens II. Cytopathological changes. Invest. Ophthalmol. Vis. Sci., 27, 108-114 (1986).
20.
P.A. Edwards, M.B. Datiles and S.B. Green, Reproducibility study on the Scheimpflug cataract video camera. Curr. Eye Res., 10, 955-960 (1988).
21.
M. Kojima, A. Wegener and O. Hockwin, Imaging characteristics of three cameras using the Scheimpflug principle. OphthalmicRes., 22 (suppl 1), 29-35 (1990).
22.
U. Meinel, Comparison between digital and analog image analysis of Scheimpflug photographs. Ophthalmic. Res., 22 (suppl 1), 71-73 (1990).