- Email: [email protected]
High precision biometry of pseudophakic eyes using partial coherence interferometry
High precision biometry of pseudophakic eyes using partial coherence interferometry Oliver Findl, MD, Wolfgang Drexler, PhD, Rupert Menapace, MD, Christoph K. Hitzenberger, PhD, Adolf F. Percher, PhD ABSTRACT Purpose: To investigate the applicability of the scanning version of dual-beam partial coherence interferometry (PCI) for measuring the anterior segment and axial length of pseudophakic eyes in a clinical setting and to determine the achievable precision with this biometry technique. Setting: Department of Ophthalmology, Vienna General Hospital , and Institute of Medical Physics, University of Vienna, Austria. Methods: Partial coherence interferometry was performed in 39 pseudophakic eyes of 39 patients after implantation of a foldable acrylic intraocular lens (IOL). Results: Effective lens position (ELP), IOL thickness, and lens-capsule distance (LCD) were determined with a precision of 2 to 3 )lm ; corneal thickness and axial eye length, with a precision of 0.8 and 5.0 )lm , respectively. The mean ELP of the IOL was 4.093 mm ± 0.290 (SO). In 7 eyes (18%), a positive LCD of 68 ± 40 )1m was detected with PCI. Mean corneal thickness was 526.4 ± 31 .5 )lm ; mean IOL thickness, 791 .5 ± 40.2 )lm ; and mean axial length, 23.388 ± 0.824 mm. Conclusion: The scanning version of PCI enables high precision (:55 )lm) and high resolution ( -12 )lm) biometry of pseudophakic eyes that is better than conventional ultrasound by a factor of more than 20. For the first time, positive LCD, a possible risk factor for posterior capsule opacification , could be detected and quantified. Furthermore, this technique offers a high degree of comfort for the patient since it is a noncontact method with no need for local anesthesia or pupil dilation and has a reduced risk of corneal infection. J Cataract Refract Surg 1998; 24: 1087-1093
ccurate biometry of pseudophakic eyes is important for precise determination of the effective lens position (ELP) after cataract surgery. The ELP is needed to determine the A-constant, which is used in several intraocular lens (IOL) power calculation formu-
A
From the Universitiitsklinik for Augenheilkunde, Allgemeines Krankenhaus Wien (Find!, Menapace), and the Institut for Medizinische Physik, Universitiit Wien (Drexler, Hitzenberger, Fercher), Austria. Reprint requests to Wolfgang Drexler, PhD, Institut for Medizinische Physik, Universitiit Wien, Wiihringer Strafe 13, A -1090 Vienna, Austria.
las along with axial length and corneal power. 1-4 Biometry is also important for determining axial length symmetry for cataract surgery in the second eye when preoperative measurements of the first eye are unavailable. 5 Accurate biometry helps identify the source of error in 6 postoperative refraction after IOL implantation. •7 Ophthalmologic biometry of pseudophakic eyes is routinely performed with conventional A-scan ultrasound, typically using a 10 MHz transducer, enabling a resolution of approximately 200 J.Un and a precision of 150 J.Un. 8·9 Recently, a noninvasive optical biometry technique, partial coherence interferometry (PCI), based
J CATARACf REFRACT SURG-VOL 24, AUGUST
1998
1087
HIGH PRECISION BIOMETRY OF PSEUDOPHAKIC EYES
on an optical coherence-domain reflectometry principle, has been developed for high precision measurement of intraocular distances 10 and high resolution tomographic imaging of the human fundus. 11 ·12 Using a classic interferometric setup, it has been used in a variety of ophthalmologic studies. 13- 15 A special dual-
(LCD) in pseudophakic eyes, a possible risk factor for posterior capsule opacification (PCO).
beam version of this technique, which is insensitive to longitudinal eye movements during measurements, has been demonstrated to perform high precision biometry16-22 as well as tomographic imaging. 23-26
Two approaches of PCI have been used: In the reflectometer approach, the object forms one of the two mirrors of a classic Michelson interferometer. 28 ·29 Because of the inherent instability of living objects, highspeed scanning techniques must be used. In the second approach, the object can be illuminated by a dual beam, generated by a Michelson interferometer for example. 9·30 This approach is insensitive to movements of the entire object. The dual-beam technique is depicted in Figure 1. The two components (1, 2) of the dual beam exiting from the external Michelson interferometer illuminate the eye. Both beam components are reflected at various interfaces within the eye. If the delay (20D) caused by the eye coincides with the initial delay (2d) of the beam
Patients and Methods Dual-Beam Partial Coherence Interferometry
During ultrasonography of pseudophakic eyes, the IOL material causes multiple artefacts at the posterior lens surface and in the vitreous, disturbing the interpretation of the A-scan and therefore the precise determination of intraocular distancesY Since PCI uses light instead of sound, PCI of pseudophakic eyes does not experience this problem. The purpose of this study was to demonstrate the potential ofPCI for performing high precision measurements of IOL-dependent ELP and axial length, as well as for detecting and quantifying lens-capsule distance
~
Coaxial Dual Beam
~;
c
Measurement Mirror(2)
lF -+-
I I.
c 2
AIOLS
1
:~2dA .• _, " ·
T
:- d-:
External Michelson Interferometer
Reference Mirror(! )
OD
Detection Unit
- ' - - - - ' -....... d
00
PC! Signal
Figure 1. (Findl) Schematic of the scanning version of the dual-beam partial coherence interferometer. The eye is illuminated via an external interferometer that produces a coaxial, dual beam. Reflected signals from the eye, for example, C,, C2 , AIOLS,, AIOLS 2 , are superimposed on and detected by a photodetector. A PCI signal of the optical distance (OD) indicating the sum of corneal thickness and anterior chamber depth is depicted, showing the envelope of the measured signal that is recorded and displayed by a personal computer. Measurements at a specific angle between vision axis and measurement direction or along a completely linear or circular scan are performed using a computer-controlled scanning mirror and special scanning optics.
1088
J CATARACT REFRACT SURG-VOL 24, AUGUST 1998
HIGH PRECISION BIOMETRY OF PSEUDOPHAKIC EYES
components, two reflected waves, for example component 1 reflected at the anterior surface of the intraocular lens (AIOLS 1) and component 2 reflected at the cornea (C 2), will interfere at the interferometer exit. This technique incorporates the Doppler principle: one of the interferometer mirrors (mirror 2 in the figure) is translated with constant speed. 15 Hence the interference phenomenon at the interferometer exit is oscillating with the corresponding Doppler frequency. This oscillating light intensity is detected by the photodetector. The electronic photodetector signal is amplified and bandpass filtered to improve the signal-to-noise ratio. After digitization, the signal (PCI signal) is recorded on a PC. The longer the component waves (so-called coherence length), the less defined the path difference at which interference occurs. Hence, the resolution of this technique is determined by the coherence length (lc) of the light used. The coherence length itself depends on the spectral width (M) of the light:
=!J.')... '),}
lc
where A is the mean wavelength and i'lA, the spectral bandwidth of the light. In the present measurements, we used a mean wavelength of A = 855 nm and a spectral bandwidth of i'lA = 20 to 25 nm, which leads to a resolution of approximately 15 J.Un in air. Axial resolution of PCI is determined by the coherence length and is independent of the numerical aperture of the illuminating beam. Attached to the interferometer is a computercontrolled scanning unit to direct the measurement beam toward the eye at arbitrary angles. 23 ·31 This device makes it possible to measure intraocular distances at various angles to the vision axis or along linear or circular scans, i.e., a series of equidistant points, for tomographic imaging of the retina. To stabilize the vision axis, a fixation light is offered directly to the investigated eye (Figure 1).
Patients and Study Design All research and measurements followed the tenets of the Helsinki agreement. Informed consent was obtained from all patients after the nature and possible consequences of the study had been explained. Furthermore, the study was approved by the ethics committee of the Vienna University School of Medicine.
Corneal thickness, ELP, IOL thickness, LCD (if existent), and vitreous length were measured by PCI in 39 eyes of 39 patients. A minimum of 10 measurements were obtained for calculating mean values and standard deviation. By adding the means of the measured intraocular distances, the total axial length was calculated. Measurement of the anterior eye segment is performed approximately parallel to the optical axis. 20 In this case, the four Purkinje images of the reflecting interfaces of cornea and IOL can be superimposed to achieve optimal signal quality. The third and fourth Purkinje image of the highly reflective IOL can be observed by the examiner on an observation monitor. The scanning version ofPCI enables measurements not only parallel to the vision axis but at arbitrary angles to it. Hence, the angle between measurement beam and vision axis is altered until the four Purkinje images are superimposed. Measurement of the anterior eye segment is then performed along this direction. Patients were asked to fixate on a green LED that was previously adjusted coaxial to the vision axis. The mean age of the patients was 72 years :::!: 11 (SD) (range 50 to 93 years). The AcrySofMA60BM lens (Alcon) had been implanted through a temporal, self-sealing incision after phacoemulsification in all eyes. All operations were done by the same surgeon (R.M.). Partial coherence interferometry was performed 6 :::!: 4 months postoperatively (range 3 to 22 months). None of the eyes had had neodymium:YAG capsulotomy. During in vivo measurements of the human eye, laser safety regulations must be considered. The light source used in this study has a center wavelength of A1 = 855 nm with a power of about 220 j..tW at the cornea or an intensity of approximately 572 j..tW/cm2 (averaged over a 7 mm aperture). Such light intensity is allowed for about 28 minutesY The times needed for 1 single measurement of the axial length and the entire anterior segment are 0.5 and 2.0 seconds, respectively. By performing 8 to 10 longitudinal scans for statistical purposes, the maximum time of continuous illumination varies between 4 and 20 seconds. This is far below the safety limits.
Data Analysis Data are presented as mean value :::!: standard deviation and range, indicating minimum and maximum. Partial coherence interferometry measures opti-
J CATARACT REFRACT SURG-VOL 24, AUGUST 1998
1089
HIGH PRECISION BIOMETRY OF PSEUDOPHAKIC EYES
cal distances within the eye. To obtain geometric distances, the optical distances have to be divided by the group refractive indices of the respective ocular media. 15 •19 In this study, values of 1.3440, 1.3454, and 1.3851 were used for the group refractive indices of vitreous, aqueous, and cornea, respectively. 19 The group refractive index of the IOL was obtained by measuring its optical thickness and by dividing this value by the central geometric thickness according to the manufacturer's data. A group refractive index of 1. 7482 was used for the AcrySof lens. 19 The precision of PCI is defined as the standard deviation of multiple recorded consecutive measurements of the ocular distance under investigation. Correlations were investigated using linear regresston; P < .05 was considered the level of significance.
both figures, the PCI signal intensity is plotted versus the optical distance to the anterior corneal surface. Four main peaks, arising from light reflected at the anterior and posterior corneal surfaces and the anterior and posterior crystalline lens (top) and IOL (bottom) surfaces, can be distinguished in Figure 2, indicating the optical corneal thickness, anterior chamber depth, cataractous lens (top), and IOL thickness (bottom), respectively. In the cataractous lens (top), peaks probably caused by light reflected at the cortex-nucleus interfaces are also detected. In contrast to ultrasound biometry, no artefacts from the IOL material are detected by PCI in the pseudophakic eye (bottom). Figure 3 shows the measurement of axial length parallel to the vision axis, since the patient fixates on the infrared measurement beam. This is made possible by using the broad emission spectrum of the light source that also contains faintly visible wavelengths. Two main peaks, caused by light reflection at the internal limiting membrane and the retinal pigment epithelium, can be distinguished. The latter was used, together with the peak reflected at the posterior IOL surface (Figure 2, bottom), to determine vitreous length. Another example of the anterior segment measurement of a pseudophakic eye is depicted in Figure 4. Because of the high resolution of the scanning version
Results Typical PCI signals-so-called optical A-scans-of the anterior segment of the same eye before and after cataract surgery as well as the axial length are shown in Figures 2 and 3, respectively. These optical A-scans are similar to those obtained by conventional ultrasound but show very narrow signal peaks, allowing much higher precision and resolution of measurement. In
10 . - --
- -- - - - -- --
Cornea
1.0
2.0
3.0
-
- -- -,
Crystalliro: lens
Anterior chamber depth
Cornea
0 .0
- - - -- - - -- --
Anterior chamber depth
4 .0
Figure 2. (Findl) Measurement of the anterior segment of an eye 1 day before (top) and 12 weeks after (bottom) cataract surgery. The interference fringe contrast (intensity) is plotted as a function of the optical distance to the anterior corneal surface. Signal peaks indicating the corneal thickness, anterior chamber depth, crystalline (cataractous) lens, as well as the IOL are depicted. In the cataractous lens (top), peaks probably caused by light reflected at the cortex-nucleus interfaces are also detected. In contrast to ultrasound biometry, no artefacts from the IOL material are detected by the PCI technique in the pseudophakic eye (bottom).
!,..,.ocular lens
5.0
6 .0
7.0
8.0
9.0
10.0
11.0
Optical distance to anterior corneal surface (mm)
1090
J CATARACT REFRACT SURG-VOL 24, AUGUST 1998
12.0
HIGH PRECISION BIOMETRY OF PSEUDOPHAKIC EYES 1.0 - , - - - - - - - - - - - - - - -- - - - --
- - -- - - - - -----,
Retinal pigment epithelium
0.5
Figure 3. (Findl) Partial coherence interferometry measurement (optical A-scan) of the axial length of a pseudophakic eye parallel to the vision axis. The optical distances from the anterior corneal surface to peaks originating from fundus layers are indicated (optical distances are shown). Two main peaks can be distinguished caused by light reflected at the internal limiting membrane and the retinal pigment epithelium. The latter is used to determine axial eye length.
Internal limiting membrane
0.0 + - - - - - , - - - - - - - , - -- - - - - , - - - - - -- , - - - - - --,--------j 32.00
31.50
31.00
30.50
Optical distance to anterior corneal surface (mm)
LCD, 3.1 ± 4.0 J.Uil (range 0.6 to 7.6 J.Un); and axial length, 4.6 ± 1.6 J.Uil (range 2.0 to 8.7 J.Un). Hence, the precision of PCI for axial length is better than that achieved with conventional ultrasound biometry in the literature by a factor of more than 20.7 The mean ELP in a pseudophakic eye with an AcrySof lens was 4.093 ± 0.290 mm (range 3.537 to 4.710 mm) and the mean IOL thickness, 791.5 ± 40.2 J.Un (range 718.3 to 883.9 J.Un). A mean LCD of
of PCI, an additional peak, indicating a signal arising from the posterior lens capsule (compare with Figure 2), enables the detection and precise quantification of an LCD. Corneal thickness was determined with a precision of 0.8 ± 0.3 J.Uil (range 0.2 to 1.6 J.Un); ELP, the distance from posterior corneal to anterior IOL surface, with a precision of2.6 ± 1.2 J.Uil (range 0.8 to 5.9 J.Un); IOL thickness, 2.3 ± 0.9 J.Uil (range 0.9 to 5.3 J.Un);
1.0
Anterior chamber depth
Cornea
Intraocular lens
Lens-capsule
j distance
Figure 4. (Findl) Measurement of the anterior segment of a pseudophakic eye. Because of the high resolution of the scanning version of PCI, an additional peak (compare with Figure 2), indicating a signal arising from the posterior lens capsule, enables the detection and precise quantification of an LCD.
0.0 0.0
1.0
2.0
3.0
4.0
5.0
6.0
7.0
8.0
Optical distance to anterior corneal surface (mm)
J CATARACT REFRACT SURG-VOL 24, AUGUST 1998
1091
HIGH PRECISION BIOMETRY OF PSEUDOPHAKIC EYES
68.5 ± 40.2 J.lm (range 21.6 to 113.9 J.Un) was detected by PCI in 7 eyes (18%). In addition, a mean corneal thickness of 526.4 ± 31.5 J.lm (range 466.3 to 596.2 J.Un), as well as a mean axial length of 23.388 ± 0.824 mm (range 21.856 to 25.443 mm), was measured. A statistically significant correlation between ELP and axial length was found (R = 0.32; P < .005). Measured IOL thickness and IOL diopters showed a highly significant correlation (R = 0.93; P < .0001).
7. 8.
9.
10. 11.
Discussion We have demonstrated that biometry using the scanning version of PCI can be performed in pseudophakic eyes. The precision and resolution was more than 20 times better than with conventional ultrasound. Therefore, accurate determination of the ELP after cataract surgery is possible. Hence, IOL-dependent A-constants, required for most IOL power calculation formulas, can be determined more precisely. Furthermore, LCD, a possible risk factor for PCO, was not only detected but also quantified with this technique. Evaluation of the IOL-specific tendency to produce a positive LCD is possible. 33 This noncontact technique offers a high degree of comfort to the patient since biometry can be performed within a short time without the need of local anesthesia. It also reduces the risk of corneal infection.
12. 13.
14.
15.
16.
17.
18.
References 1. Sanders 0, Retzlaff J, Kraff M, et al. Comparison of the
2.
3.
4.
5.
6.
accuracy of the Binkhorst, Colenbrander, and SRKTM implant power prediction formulas. Am Intra-Ocular Implant Soc J 1981; 7:337-340 Sanders DR, Retzlaff J, Kraff MC. Comparison of the SRK IITM formula and other second generation formulas. J Cataract Refract Surg 1988; 14:136-141 RetzlaffJA, Sanders DR, KraffMC. Development of the SRK/T intraocular lens implant power calculation formula. J Cataract Refract Surg 1990; 16:333-340; correction p 528 Holladay JT. Standardizing constants for ultrasonic biometry, keratometry, and intraocular lens power calculations. J Cataract Refract Surg 1997; 23:1356-1370 Holladay JT, Prager TC, Chandler TY, eta!. A three-part system for refining intraocular lens power calculations. J Cataract Refract Surg 1988; 14:17-24 Holladay JT, Prager TC, Ruiz RS, et a!. Improving the
1092
19.
20.
21.
22.
23.
predictability of intraocular lens power calculations. Arch Ophthalmol 1986; 104:539-541 Olsen T. Sources of error in intraocular lens power calculation. J Cataract Refract Surg 1992; 18:125-129 Olsen T. The accuracy of ultrasonic determination of axial length in pseudophakic eyes. Acta Ophthalmol 1989; 67:141-144 Bamber JC, Trstam M. Diagnostic ultrasound. In: Webb S, ed, The Physics of Medical Imaging. Bristol and Philadelphia, Adam Hilger 1988; 319-388 Percher AF, Roth E. Ophthalmic laser interferometer. Proc SPIE 1986; 658:48-51 Huang D, Swanson EA, Lin CP, eta!. Optical coherence tomography. Science 1991; 254:1178-1181 Percher AF. Optical coherence tomography. J Biomed Opt 1996; 1:157-173 Puliafito CA, Hee MR, Lin CP, et a!. Imaging of macular diseases with optical coherence tomography. Ophthalmology 1995; 102:217-229 Hee MR, Puliafito CA, Wong C, et a!. Quantitative assessment of macular edema with optical coherence tomography. Arch Ophthalmol 1995; 113:1019-1029 Schuman JS, Hee MR, Puliafito CA, et al. Quantification of nerve fiber layer thickness in normal and glaucomatous eyes using optical coherence tomography. Arch Ophthalmol 1995; 113:586-596 Hitzenberger CK. Optical measurement of the axial eye length by laser Doppler interferometry. Invest Ophthalmol Vis Sci 1991; 32:616-624 Drexler W, Baumgartner A, Find! 0, eta!. Submicrometer precision biometry of the anterior segment of the human eye. Invest Ophthalmol Vis Sci 1997; 38:13041313 Drexler W, Baumgartner A, Find! 0, et a!. Biometric investigation of changes in the anterior eye segment during accommodation. Vision Res 1997; 37:27892800 Drexler W, Find! 0, Schmetterer L, eta!. Eye elongation during accommodation in humans-differences between emmetropes and myopes. In press, Invest Ophthalmol Vis Sci Drexler W, Hitzenberger CK, Baumgartner A, et a!. Investigation of dispersion effects in ocular media by multiple wavelength partial coherence interferometer. Exp Eye Res 1998; 66:25-33 Hitzenberger CK, Drexler W, Dolezal C, eta!. Measurement of the axial length of cataract eyes by laser Doppler interferometry. Invest Ophthalmol Vis Sci 1993; 34: 1886-1893 Drexler W, Find! 0, Menapace R, et a!. Partial coherence interferometry: a novel approach to biometry in cataract surgery. In press, Am J Ophthalmol Percher AF, Hitzenberger CK, Drexler W, et al. In vivo optical coherence tomography (letter). Am J Ophthalmol 1993; 116:113-114
J CATARACT REFRACT SURG-VOL 24, AUGUST 1998
HIGH PRECISION BIOMETRY OF PSEUDOPHAKIC EYES
24. Drexler W, Hitzenberger CK, Sattmann H, Fercher AF. Measurement of the thickness of fundus layers by partial coherence tomography. Opt Eng 1995; 34:701-710 25. Drexler W, Find! 0, Menapace R, et a!. Dual beam optical coherence tomography: signal identification for ophthalmologic diagnosis. J Biomed Opt 1998; 3:5565 26. Baumgartner A, Hitzenberger CK, Sattmann H, et a!. Signal and resolution enhancements in dual beam optical coherence tomography of the human eye. J Biomed Opt 1998; 3:45-54 27. Naeser K, Naeser A, Boberg-Ans J, Bargum R. Axial length following implantation of posterior chamber lenses. J Cataract Refract Surg 1989; 15:673-675 28. Takada K, Noda J, Okamoto K. Measurement of spatial distribution of mode coupling in birefringent polarization-maintaining fiber with new detection scheme. Opt Lett 1986; 11 :680-682 29. Huang D, Wang J, Lin CP, et a!. Micron-resolution ranging of cornea anterior chamber by optical reflectometry. Lasers Surg Med 1991; 11:419-425 30. Fercher AF, Mengedoht K, Werner W Eye length measurement by interferometry with partially coherent light. Optics Lett 1988; 13:186-188
31. Drexler W, Hitzenberger CK, Fercher AF. A scanninglaser interferometer for fundus profile measurement of the eye. Proc SPIE 1993; 2083:363-371 32. American National Standards Institute. Safe Use of Lasers, ANSI Z 136.1. New York, NY, American National Standards Institute, 1986 33. Find! 0, Drexler W, Menapace R, et a!. Accurate measurement of lens-capsule distance and effective lens position: comparison of different intraocular lenses. J Cataract Refract Surg 1998; 24:1094-1098
Presented in part at the annual meeting of the Deutschsprachige Gesellschaftfor Intraokularlinsen Implantation und Refraktive Chirurgie, Frankfurt, Germany, March 1997 None of the authors has a proprietary or financial interest in any product mentioned. Supported by the Austrian Fonds zur Forderung der wissenschaftlichen Forschung (grant P 9781-MED). Ing. H. Sattmann and Mr. L. Schachinger, Institite ofMedical Physics, University of Vienna, Austria, constructed the instrument's electronics and software and provided technical support, respectively.
J CATARACT REFRACT SURG-VOL 24, AUGUST 1998
1093
ccurate biometry of pseudophakic eyes is important for precise determination of the effective lens position (ELP) after cataract surgery. The ELP is needed to determine the A-constant, which is used in several intraocular lens (IOL) power calculation formu-
A
From the Universitiitsklinik for Augenheilkunde, Allgemeines Krankenhaus Wien (Find!, Menapace), and the Institut for Medizinische Physik, Universitiit Wien (Drexler, Hitzenberger, Fercher), Austria. Reprint requests to Wolfgang Drexler, PhD, Institut for Medizinische Physik, Universitiit Wien, Wiihringer Strafe 13, A -1090 Vienna, Austria.
las along with axial length and corneal power. 1-4 Biometry is also important for determining axial length symmetry for cataract surgery in the second eye when preoperative measurements of the first eye are unavailable. 5 Accurate biometry helps identify the source of error in 6 postoperative refraction after IOL implantation. •7 Ophthalmologic biometry of pseudophakic eyes is routinely performed with conventional A-scan ultrasound, typically using a 10 MHz transducer, enabling a resolution of approximately 200 J.Un and a precision of 150 J.Un. 8·9 Recently, a noninvasive optical biometry technique, partial coherence interferometry (PCI), based
J CATARACf REFRACT SURG-VOL 24, AUGUST
1998
1087
HIGH PRECISION BIOMETRY OF PSEUDOPHAKIC EYES
on an optical coherence-domain reflectometry principle, has been developed for high precision measurement of intraocular distances 10 and high resolution tomographic imaging of the human fundus. 11 ·12 Using a classic interferometric setup, it has been used in a variety of ophthalmologic studies. 13- 15 A special dual-
(LCD) in pseudophakic eyes, a possible risk factor for posterior capsule opacification (PCO).
beam version of this technique, which is insensitive to longitudinal eye movements during measurements, has been demonstrated to perform high precision biometry16-22 as well as tomographic imaging. 23-26
Two approaches of PCI have been used: In the reflectometer approach, the object forms one of the two mirrors of a classic Michelson interferometer. 28 ·29 Because of the inherent instability of living objects, highspeed scanning techniques must be used. In the second approach, the object can be illuminated by a dual beam, generated by a Michelson interferometer for example. 9·30 This approach is insensitive to movements of the entire object. The dual-beam technique is depicted in Figure 1. The two components (1, 2) of the dual beam exiting from the external Michelson interferometer illuminate the eye. Both beam components are reflected at various interfaces within the eye. If the delay (20D) caused by the eye coincides with the initial delay (2d) of the beam
Patients and Methods Dual-Beam Partial Coherence Interferometry
During ultrasonography of pseudophakic eyes, the IOL material causes multiple artefacts at the posterior lens surface and in the vitreous, disturbing the interpretation of the A-scan and therefore the precise determination of intraocular distancesY Since PCI uses light instead of sound, PCI of pseudophakic eyes does not experience this problem. The purpose of this study was to demonstrate the potential ofPCI for performing high precision measurements of IOL-dependent ELP and axial length, as well as for detecting and quantifying lens-capsule distance
~
Coaxial Dual Beam
~;
c
Measurement Mirror(2)
lF -+-
I I.
c 2
AIOLS
1
:~2dA .• _, " ·
T
:- d-:
External Michelson Interferometer
Reference Mirror(! )
OD
Detection Unit
- ' - - - - ' -....... d
00
PC! Signal
Figure 1. (Findl) Schematic of the scanning version of the dual-beam partial coherence interferometer. The eye is illuminated via an external interferometer that produces a coaxial, dual beam. Reflected signals from the eye, for example, C,, C2 , AIOLS,, AIOLS 2 , are superimposed on and detected by a photodetector. A PCI signal of the optical distance (OD) indicating the sum of corneal thickness and anterior chamber depth is depicted, showing the envelope of the measured signal that is recorded and displayed by a personal computer. Measurements at a specific angle between vision axis and measurement direction or along a completely linear or circular scan are performed using a computer-controlled scanning mirror and special scanning optics.
1088
J CATARACT REFRACT SURG-VOL 24, AUGUST 1998
HIGH PRECISION BIOMETRY OF PSEUDOPHAKIC EYES
components, two reflected waves, for example component 1 reflected at the anterior surface of the intraocular lens (AIOLS 1) and component 2 reflected at the cornea (C 2), will interfere at the interferometer exit. This technique incorporates the Doppler principle: one of the interferometer mirrors (mirror 2 in the figure) is translated with constant speed. 15 Hence the interference phenomenon at the interferometer exit is oscillating with the corresponding Doppler frequency. This oscillating light intensity is detected by the photodetector. The electronic photodetector signal is amplified and bandpass filtered to improve the signal-to-noise ratio. After digitization, the signal (PCI signal) is recorded on a PC. The longer the component waves (so-called coherence length), the less defined the path difference at which interference occurs. Hence, the resolution of this technique is determined by the coherence length (lc) of the light used. The coherence length itself depends on the spectral width (M) of the light:
=!J.')... '),}
lc
where A is the mean wavelength and i'lA, the spectral bandwidth of the light. In the present measurements, we used a mean wavelength of A = 855 nm and a spectral bandwidth of i'lA = 20 to 25 nm, which leads to a resolution of approximately 15 J.Un in air. Axial resolution of PCI is determined by the coherence length and is independent of the numerical aperture of the illuminating beam. Attached to the interferometer is a computercontrolled scanning unit to direct the measurement beam toward the eye at arbitrary angles. 23 ·31 This device makes it possible to measure intraocular distances at various angles to the vision axis or along linear or circular scans, i.e., a series of equidistant points, for tomographic imaging of the retina. To stabilize the vision axis, a fixation light is offered directly to the investigated eye (Figure 1).
Patients and Study Design All research and measurements followed the tenets of the Helsinki agreement. Informed consent was obtained from all patients after the nature and possible consequences of the study had been explained. Furthermore, the study was approved by the ethics committee of the Vienna University School of Medicine.
Corneal thickness, ELP, IOL thickness, LCD (if existent), and vitreous length were measured by PCI in 39 eyes of 39 patients. A minimum of 10 measurements were obtained for calculating mean values and standard deviation. By adding the means of the measured intraocular distances, the total axial length was calculated. Measurement of the anterior eye segment is performed approximately parallel to the optical axis. 20 In this case, the four Purkinje images of the reflecting interfaces of cornea and IOL can be superimposed to achieve optimal signal quality. The third and fourth Purkinje image of the highly reflective IOL can be observed by the examiner on an observation monitor. The scanning version ofPCI enables measurements not only parallel to the vision axis but at arbitrary angles to it. Hence, the angle between measurement beam and vision axis is altered until the four Purkinje images are superimposed. Measurement of the anterior eye segment is then performed along this direction. Patients were asked to fixate on a green LED that was previously adjusted coaxial to the vision axis. The mean age of the patients was 72 years :::!: 11 (SD) (range 50 to 93 years). The AcrySofMA60BM lens (Alcon) had been implanted through a temporal, self-sealing incision after phacoemulsification in all eyes. All operations were done by the same surgeon (R.M.). Partial coherence interferometry was performed 6 :::!: 4 months postoperatively (range 3 to 22 months). None of the eyes had had neodymium:YAG capsulotomy. During in vivo measurements of the human eye, laser safety regulations must be considered. The light source used in this study has a center wavelength of A1 = 855 nm with a power of about 220 j..tW at the cornea or an intensity of approximately 572 j..tW/cm2 (averaged over a 7 mm aperture). Such light intensity is allowed for about 28 minutesY The times needed for 1 single measurement of the axial length and the entire anterior segment are 0.5 and 2.0 seconds, respectively. By performing 8 to 10 longitudinal scans for statistical purposes, the maximum time of continuous illumination varies between 4 and 20 seconds. This is far below the safety limits.
Data Analysis Data are presented as mean value :::!: standard deviation and range, indicating minimum and maximum. Partial coherence interferometry measures opti-
J CATARACT REFRACT SURG-VOL 24, AUGUST 1998
1089
HIGH PRECISION BIOMETRY OF PSEUDOPHAKIC EYES
cal distances within the eye. To obtain geometric distances, the optical distances have to be divided by the group refractive indices of the respective ocular media. 15 •19 In this study, values of 1.3440, 1.3454, and 1.3851 were used for the group refractive indices of vitreous, aqueous, and cornea, respectively. 19 The group refractive index of the IOL was obtained by measuring its optical thickness and by dividing this value by the central geometric thickness according to the manufacturer's data. A group refractive index of 1. 7482 was used for the AcrySof lens. 19 The precision of PCI is defined as the standard deviation of multiple recorded consecutive measurements of the ocular distance under investigation. Correlations were investigated using linear regresston; P < .05 was considered the level of significance.
both figures, the PCI signal intensity is plotted versus the optical distance to the anterior corneal surface. Four main peaks, arising from light reflected at the anterior and posterior corneal surfaces and the anterior and posterior crystalline lens (top) and IOL (bottom) surfaces, can be distinguished in Figure 2, indicating the optical corneal thickness, anterior chamber depth, cataractous lens (top), and IOL thickness (bottom), respectively. In the cataractous lens (top), peaks probably caused by light reflected at the cortex-nucleus interfaces are also detected. In contrast to ultrasound biometry, no artefacts from the IOL material are detected by PCI in the pseudophakic eye (bottom). Figure 3 shows the measurement of axial length parallel to the vision axis, since the patient fixates on the infrared measurement beam. This is made possible by using the broad emission spectrum of the light source that also contains faintly visible wavelengths. Two main peaks, caused by light reflection at the internal limiting membrane and the retinal pigment epithelium, can be distinguished. The latter was used, together with the peak reflected at the posterior IOL surface (Figure 2, bottom), to determine vitreous length. Another example of the anterior segment measurement of a pseudophakic eye is depicted in Figure 4. Because of the high resolution of the scanning version
Results Typical PCI signals-so-called optical A-scans-of the anterior segment of the same eye before and after cataract surgery as well as the axial length are shown in Figures 2 and 3, respectively. These optical A-scans are similar to those obtained by conventional ultrasound but show very narrow signal peaks, allowing much higher precision and resolution of measurement. In
10 . - --
- -- - - - -- --
Cornea
1.0
2.0
3.0
-
- -- -,
Crystalliro: lens
Anterior chamber depth
Cornea
0 .0
- - - -- - - -- --
Anterior chamber depth
4 .0
Figure 2. (Findl) Measurement of the anterior segment of an eye 1 day before (top) and 12 weeks after (bottom) cataract surgery. The interference fringe contrast (intensity) is plotted as a function of the optical distance to the anterior corneal surface. Signal peaks indicating the corneal thickness, anterior chamber depth, crystalline (cataractous) lens, as well as the IOL are depicted. In the cataractous lens (top), peaks probably caused by light reflected at the cortex-nucleus interfaces are also detected. In contrast to ultrasound biometry, no artefacts from the IOL material are detected by the PCI technique in the pseudophakic eye (bottom).
!,..,.ocular lens
5.0
6 .0
7.0
8.0
9.0
10.0
11.0
Optical distance to anterior corneal surface (mm)
1090
J CATARACT REFRACT SURG-VOL 24, AUGUST 1998
12.0
HIGH PRECISION BIOMETRY OF PSEUDOPHAKIC EYES 1.0 - , - - - - - - - - - - - - - - -- - - - --
- - -- - - - - -----,
Retinal pigment epithelium
0.5
Figure 3. (Findl) Partial coherence interferometry measurement (optical A-scan) of the axial length of a pseudophakic eye parallel to the vision axis. The optical distances from the anterior corneal surface to peaks originating from fundus layers are indicated (optical distances are shown). Two main peaks can be distinguished caused by light reflected at the internal limiting membrane and the retinal pigment epithelium. The latter is used to determine axial eye length.
Internal limiting membrane
0.0 + - - - - - , - - - - - - - , - -- - - - - , - - - - - -- , - - - - - --,--------j 32.00
31.50
31.00
30.50
Optical distance to anterior corneal surface (mm)
LCD, 3.1 ± 4.0 J.Uil (range 0.6 to 7.6 J.Un); and axial length, 4.6 ± 1.6 J.Uil (range 2.0 to 8.7 J.Un). Hence, the precision of PCI for axial length is better than that achieved with conventional ultrasound biometry in the literature by a factor of more than 20.7 The mean ELP in a pseudophakic eye with an AcrySof lens was 4.093 ± 0.290 mm (range 3.537 to 4.710 mm) and the mean IOL thickness, 791.5 ± 40.2 J.Un (range 718.3 to 883.9 J.Un). A mean LCD of
of PCI, an additional peak, indicating a signal arising from the posterior lens capsule (compare with Figure 2), enables the detection and precise quantification of an LCD. Corneal thickness was determined with a precision of 0.8 ± 0.3 J.Uil (range 0.2 to 1.6 J.Un); ELP, the distance from posterior corneal to anterior IOL surface, with a precision of2.6 ± 1.2 J.Uil (range 0.8 to 5.9 J.Un); IOL thickness, 2.3 ± 0.9 J.Uil (range 0.9 to 5.3 J.Un);
1.0
Anterior chamber depth
Cornea
Intraocular lens
Lens-capsule
j distance
Figure 4. (Findl) Measurement of the anterior segment of a pseudophakic eye. Because of the high resolution of the scanning version of PCI, an additional peak (compare with Figure 2), indicating a signal arising from the posterior lens capsule, enables the detection and precise quantification of an LCD.
0.0 0.0
1.0
2.0
3.0
4.0
5.0
6.0
7.0
8.0
Optical distance to anterior corneal surface (mm)
J CATARACT REFRACT SURG-VOL 24, AUGUST 1998
1091
HIGH PRECISION BIOMETRY OF PSEUDOPHAKIC EYES
68.5 ± 40.2 J.lm (range 21.6 to 113.9 J.Un) was detected by PCI in 7 eyes (18%). In addition, a mean corneal thickness of 526.4 ± 31.5 J.lm (range 466.3 to 596.2 J.Un), as well as a mean axial length of 23.388 ± 0.824 mm (range 21.856 to 25.443 mm), was measured. A statistically significant correlation between ELP and axial length was found (R = 0.32; P < .005). Measured IOL thickness and IOL diopters showed a highly significant correlation (R = 0.93; P < .0001).
7. 8.
9.
10. 11.
Discussion We have demonstrated that biometry using the scanning version of PCI can be performed in pseudophakic eyes. The precision and resolution was more than 20 times better than with conventional ultrasound. Therefore, accurate determination of the ELP after cataract surgery is possible. Hence, IOL-dependent A-constants, required for most IOL power calculation formulas, can be determined more precisely. Furthermore, LCD, a possible risk factor for PCO, was not only detected but also quantified with this technique. Evaluation of the IOL-specific tendency to produce a positive LCD is possible. 33 This noncontact technique offers a high degree of comfort to the patient since biometry can be performed within a short time without the need of local anesthesia. It also reduces the risk of corneal infection.
12. 13.
14.
15.
16.
17.
18.
References 1. Sanders 0, Retzlaff J, Kraff M, et al. Comparison of the
2.
3.
4.
5.
6.
accuracy of the Binkhorst, Colenbrander, and SRKTM implant power prediction formulas. Am Intra-Ocular Implant Soc J 1981; 7:337-340 Sanders DR, Retzlaff J, Kraff MC. Comparison of the SRK IITM formula and other second generation formulas. J Cataract Refract Surg 1988; 14:136-141 RetzlaffJA, Sanders DR, KraffMC. Development of the SRK/T intraocular lens implant power calculation formula. J Cataract Refract Surg 1990; 16:333-340; correction p 528 Holladay JT. Standardizing constants for ultrasonic biometry, keratometry, and intraocular lens power calculations. J Cataract Refract Surg 1997; 23:1356-1370 Holladay JT, Prager TC, Chandler TY, eta!. A three-part system for refining intraocular lens power calculations. J Cataract Refract Surg 1988; 14:17-24 Holladay JT, Prager TC, Ruiz RS, et a!. Improving the
1092
19.
20.
21.
22.
23.
predictability of intraocular lens power calculations. Arch Ophthalmol 1986; 104:539-541 Olsen T. Sources of error in intraocular lens power calculation. J Cataract Refract Surg 1992; 18:125-129 Olsen T. The accuracy of ultrasonic determination of axial length in pseudophakic eyes. Acta Ophthalmol 1989; 67:141-144 Bamber JC, Trstam M. Diagnostic ultrasound. In: Webb S, ed, The Physics of Medical Imaging. Bristol and Philadelphia, Adam Hilger 1988; 319-388 Percher AF, Roth E. Ophthalmic laser interferometer. Proc SPIE 1986; 658:48-51 Huang D, Swanson EA, Lin CP, eta!. Optical coherence tomography. Science 1991; 254:1178-1181 Percher AF. Optical coherence tomography. J Biomed Opt 1996; 1:157-173 Puliafito CA, Hee MR, Lin CP, et a!. Imaging of macular diseases with optical coherence tomography. Ophthalmology 1995; 102:217-229 Hee MR, Puliafito CA, Wong C, et a!. Quantitative assessment of macular edema with optical coherence tomography. Arch Ophthalmol 1995; 113:1019-1029 Schuman JS, Hee MR, Puliafito CA, et al. Quantification of nerve fiber layer thickness in normal and glaucomatous eyes using optical coherence tomography. Arch Ophthalmol 1995; 113:586-596 Hitzenberger CK. Optical measurement of the axial eye length by laser Doppler interferometry. Invest Ophthalmol Vis Sci 1991; 32:616-624 Drexler W, Baumgartner A, Find! 0, eta!. Submicrometer precision biometry of the anterior segment of the human eye. Invest Ophthalmol Vis Sci 1997; 38:13041313 Drexler W, Baumgartner A, Find! 0, et a!. Biometric investigation of changes in the anterior eye segment during accommodation. Vision Res 1997; 37:27892800 Drexler W, Find! 0, Schmetterer L, eta!. Eye elongation during accommodation in humans-differences between emmetropes and myopes. In press, Invest Ophthalmol Vis Sci Drexler W, Hitzenberger CK, Baumgartner A, et a!. Investigation of dispersion effects in ocular media by multiple wavelength partial coherence interferometer. Exp Eye Res 1998; 66:25-33 Hitzenberger CK, Drexler W, Dolezal C, eta!. Measurement of the axial length of cataract eyes by laser Doppler interferometry. Invest Ophthalmol Vis Sci 1993; 34: 1886-1893 Drexler W, Find! 0, Menapace R, et a!. Partial coherence interferometry: a novel approach to biometry in cataract surgery. In press, Am J Ophthalmol Percher AF, Hitzenberger CK, Drexler W, et al. In vivo optical coherence tomography (letter). Am J Ophthalmol 1993; 116:113-114
J CATARACT REFRACT SURG-VOL 24, AUGUST 1998
HIGH PRECISION BIOMETRY OF PSEUDOPHAKIC EYES
24. Drexler W, Hitzenberger CK, Sattmann H, Fercher AF. Measurement of the thickness of fundus layers by partial coherence tomography. Opt Eng 1995; 34:701-710 25. Drexler W, Find! 0, Menapace R, et a!. Dual beam optical coherence tomography: signal identification for ophthalmologic diagnosis. J Biomed Opt 1998; 3:5565 26. Baumgartner A, Hitzenberger CK, Sattmann H, et a!. Signal and resolution enhancements in dual beam optical coherence tomography of the human eye. J Biomed Opt 1998; 3:45-54 27. Naeser K, Naeser A, Boberg-Ans J, Bargum R. Axial length following implantation of posterior chamber lenses. J Cataract Refract Surg 1989; 15:673-675 28. Takada K, Noda J, Okamoto K. Measurement of spatial distribution of mode coupling in birefringent polarization-maintaining fiber with new detection scheme. Opt Lett 1986; 11 :680-682 29. Huang D, Wang J, Lin CP, et a!. Micron-resolution ranging of cornea anterior chamber by optical reflectometry. Lasers Surg Med 1991; 11:419-425 30. Fercher AF, Mengedoht K, Werner W Eye length measurement by interferometry with partially coherent light. Optics Lett 1988; 13:186-188
31. Drexler W, Hitzenberger CK, Fercher AF. A scanninglaser interferometer for fundus profile measurement of the eye. Proc SPIE 1993; 2083:363-371 32. American National Standards Institute. Safe Use of Lasers, ANSI Z 136.1. New York, NY, American National Standards Institute, 1986 33. Find! 0, Drexler W, Menapace R, et a!. Accurate measurement of lens-capsule distance and effective lens position: comparison of different intraocular lenses. J Cataract Refract Surg 1998; 24:1094-1098
Presented in part at the annual meeting of the Deutschsprachige Gesellschaftfor Intraokularlinsen Implantation und Refraktive Chirurgie, Frankfurt, Germany, March 1997 None of the authors has a proprietary or financial interest in any product mentioned. Supported by the Austrian Fonds zur Forderung der wissenschaftlichen Forschung (grant P 9781-MED). Ing. H. Sattmann and Mr. L. Schachinger, Institite ofMedical Physics, University of Vienna, Austria, constructed the instrument's electronics and software and provided technical support, respectively.
J CATARACT REFRACT SURG-VOL 24, AUGUST 1998
1093