The axis of astigmatism generally is measured with patients in the sitting position, whereas refractive surgeries are performed with patients in the supine position. The occurrence of cyclotorsion from the sitting position to the supine position could be an important factor in the correction of astigmatism in refractive surgery. Further, the recent introduction of wavefront-guided ablation has significantly increased the precision required to perform an effective ablation of the cornea. During refractive surgery, cyclotorsion reduces the effectiveness of an astigmatic correction and wavefront ablation, or may even induce astigmatism. An appropriate correction method must be designed to reduce the eye movement and maintain the range of movement within the desired level.
The Torsion Error Detector (TED; NIDEK Co Ltd, Gamagori, Japan) is an automated program that detects the magnitude of cyclotorsion by comparing iris patterns, which are unique to an individual, in sitting and supine positions. This allows surgeons to compensate for the degree of rotational movement during laser ablation. This study analyzes the change in cyclotorsional rotation during excimer laser ablation in LASIK surgery using the NIDEK TED system.
MATERIALS AND METHODS
TORSION ERROR DETECTOR SYSTEM
The Optical Path Difference Scan aberrometer (OPD-Scan; NIDEK Co Ltd) and the charge-coupled device (CCD) camera attached to the microscope of the EC-5000CXII (NIDEK Co Ltd) acquire the iris image using infrared illumination. The torsional position is derived from the natural contrast of iris pigmentation. This eliminates the use of markers for torsion measurement. The angular structure of the iris pigmentation can be addressed most efficiently using a polar coordinate system, with its origin in the pupil center. Cyclotorsion of the eye is represented as a change of angular position of the iris structure encircling the pupil. The TED system attached to the NIDEK EC-5000CXII in our clinic was specially modified to monitor and quantify cyclotorsion throughout surgery, including during laser ablation (Fig 1).
Figure 1. Raw ?mages of the same eye taken by the cameras of the two instruments in the NIDEK Torsion Error Detector system. The image on the left was taken by the OPD-Scan and the image on the right was taken by the camera attached to the microscope port of the EC-5000CXII excimer laser system. The change in cyclotorsion was captured every half-second during laser ablation, therefore the number in the red circle changed every half-second. Ambient illumination was adjusted to induce pupillary constriction to approximately the same pupil size as that obtained with the OPD-Scan.
PREVENTING TORSIONAL MISALIGNMENT
Pharmacological dilation or constriction is not required with the TED system. The TED system constantly monitors the pupil for changes in size (diameter) and the pupil center. If changes occur in pupil size greater than 45% in relation to the reference image, a warning is displayed on the laser screen. Changes 45% or less are compensated for by a proprietary software algorithm. To obviate against significant pupil center shift, ambient illumination in the surgical field can be modulated using the slit-beam illumination or diffuse illumination on the laser device to induce constriction or dilation of the pupil size to roughly correspond with the pupil size of the reference image. If the pupil size change is constantly greater than 45%, then the calculation of the torsion angle may not be possible due to the reduced iris area available for comparison. Rotation is measured around the vertical axis through the center of the pupil during surgery.
A reference eye model with a known amount of cyclotorsion (5?) is imaged by the TED system to check the system for accuracy. The operator can verify whether the TED system measures the correct amount of cyclotorsion.
To ensure pupil size criterion is met, the reference alignment image is acquired immediately after the placido disk turns off during corneal topography acquisition with the OPD-Scan. The illumination of the placido disk is sufficient to temporarily constrict the patient's pupil, thereby allowing adequate iris exposure. Torsional misalignment then can be calculated directly during surgery by comparison to the reference alignment image from the OPD-Scan. The torsion angle is displayed on the laser screen in increments of 0.2?; the data in this study are reported in 0.01? increments as a result of the calculation of the arithmetic mean.
PATIENTS AND INTRAOPERATIVE MEASUREMENTS
A total of 192 eyes of 110 patients who underwent LASIK between January 2004 and February 2004 for the correction of myopic astigmatism were included in this study. The magnitude of cyclotorsion during laser ablation was measured using the TED system. Sixtyfive patients were men and 45 patients were women. Mean patient age was 33.5 ?8.0 years (range: 18 to 58 years). Preoperatively, manifest refractive spherical error ranged from 0 to ?12.75 diopters (D) (mean: ? 6.09?2.72 D), and cylindrical error ranged from 0.50 to 5.00 D (mean: 1.42 ?0.91 D). Manifest refractive spherical equivalent ranged from ?1.00 to -13.75 D (mean: -6.80?2.74 D).
The NIDEK Advanced Vision Excimer Laser (NAVEX; NIDEK Co Ltd) platform was used for all treatments. The NAVEX system consists of the NIDEK EC-5000CXII excimer laser system, Final Fit ablation planning software, OPD-Scan, and MK- 2 000 microkeratome. The OPD-Scan is a combination unit that measures aberrometry, corneal topography, autorefraction, autokeratometry, and pupil size. Aberrometry is performed using dynamic skiascopy, and corneal shape is measured using placido disk topography.
All surgeries were performed by five surgeons (Y.H., LT., ML, T. Y., K.T.) using the same technique. Eyes were anesthetized with topical lidocaine (Xylocaine; Fujisawa Pharmaceutical Co Ltd, Osaka, Japan), and then a sterile drape and lid speculum were placed on the operative eye. Patients were instructed to remain completely still in a recumbent position until the surgery was completed. A corneal flap with a nasal hinge was created using the MK-2000 microkeratome. The iris pattern of the patient's eye in the supine position was imaged via the EC-5000CXII CCD camera, and this pattern was compared to an image obtained during OPD-Scan measurement with the patients in the sitting position (see Fig 1).
If cyclotorsion occurred prior to ablation, the amount and direction (ie, plus, counterclockwise or minus, and clockwise) was displayed in degrees on the laser screen, and the patient's head was gently shifted by the surgeon until the eye rotated to compensate for the initial cyclotorsion. If cyclotorsion greater than or equal to 2? occurred during laser ablation, the surgeon was immediately informed by the laser technician, and the surgeon manually stopped the laser ablation and gently shifted the patient's head to compensate for the cyclotorsion, then resumed ablation by depressing a foot pedal. The laser delivery was accurately centered using a 60-Hz infrared closed loop eye-tracking system and by ensuring that the patient was fixating on a red fixation light throughout the ablation.
Figure 3. Range of change in cyclotorsion during laser ablation. The circles represent the degree of cyclotorsion in each eye after lifting the flap before laser ablation, and the lines represent the range of change in cyclotorsion during laser ablation, with patterns of cyclotorsion varying from eye to eye.
Figure 2. Distribution of the cyclotorsion detected by the NIDEK Torsion Error Detector system after lifting the flap before exposing the eye to the excimer laser ablation. In some cases, cyclotorsion occurred after flap creation when the position adjusted to zero degrees of cyclotorsion at the beginning of surgery.
The change in cyclotorsion was recorded on a videodisk; this change was repeatedly measured after every half-second during laser ablation. Ablation time ranged from 9 to 71 seconds (mean: 43.2 ?12.7 seconds). FoIlowing ablation, the flap was restored to its original position, irrigated using balanced salt solution, and allowed to adhere prior to discharging the patient.
Figure 4. Distribution of the range in change of cyclotorsion during laser ablation. A number of eyes remained within 2? of cyclotorsion, but some eyes had a change of >4? of cyclotorsion.
Figure 2 shows the distribution of the cyclotorsion detected by the TED system after lifting the flap but before exposing the eye to the excimer laser. The absolute degree of cyclotorsion ranged from 0.02? to 3.95? (mean: 1.09?0.92?). Figure 3 shows the range of cyclotorsion measured during laser ablation. During laser ablation, the mean amount of cyclotorsional rotation detected by the TED was 1.33?1.88? (range: -6.33? to 2.99?) in the clockwise direction and 1.00?1.79? (?3.70? to 7.34?) in the counterclockwise direction. The absolute degree of torsion error detected by the TED system was 2.33?1.16? (0 to 6.21?) (Fig 4).
This study investigated the magnitude of ocular cyclotorsion during excimer laser ablation in a cohort of patients undergoing refractive surgery. The degree of cyclotorsion as patients move from a sitting position to a supine position has been measured indirectly using the differences in the Maddox double rod measurements,1 trial frame refraction,2 videokeratography,3 handheld automated keratometer,4 and three-dimensional video-oculography.5 Swami et al6 reported the direct measurement of the error in rotational positioning during LASIK by marking the meridian during slitlamp examination and then using the eye-tracking system after lifting the corneal flap but prior to exposing the eye to the excimer laser. Chernyak7 reported the images from a wavefront analyzer could be captured and compared to the images obtained from a laser system immediately before laser ablation. However, no peer-reviewed studies on measuring the torsional error with the passage of time during laser ablation have been published. In our study, the magnitude of cyclotorsion changed during surgery, and this change differed among individuals.
The treatment of astigmatism represents a significant portion of many conventional and custom ablation treatments. The effectiveness of the cylinder treatment can be reduced as the result of torsional error. It is possible for a cylindrical error to be undercorrected or induced postoperatively as a result of cyclotorsion during LASIK. Several authors1,2,4,5 have reported they found no statistically significant positionally induced cyclotorsion; however, the number of patients in these reports was generally small, ranging from 20 to 30. Several other studies3,6,7 with a larger number of patients, including our study, have reported a mean degree of cyclotorsion ranging from 2? to 4.4?. Although this change is not very large, these studies demonstrated that in some individuals, the deviation was greater than 6? or 10?. Such large changes in ocular cyclotorsion must be accounted for by a continuous cyclotorsion monitoring system.
Among the majority of patients in our study, the cyclotorsion in the sitting and supine positions was approximately Io to 2?. However, in approximately 4% of the patients in our study, the deviation was greater than 5?. Theoretically, a 4? difference would result in a 14% undercorrection of astigmatism. Similarly, a difference of 6? would result in a 20% undercorrection, and a difference of 10? would result in a 35% undercorrection. For wavefront-guided treatments, greater accuracy of alignment is required to fully correct higher-order aberrations. Bueeler et al8 suggested the alignment in the wavefront-guided treatments would have to be performed with a torsional precision of approximately Io or more to achieve the diffraction limit in 95% of the normal eyes with a 7-mm pupil. Even in the case of small pupils, in which alignment becomes less critical, the same quality of result would be obtained in 95% of the 3-mm pupils, thereby providing an accuracy of 3? or more.
Considering the hypothesis of Bueeler et al8 in combination with our results, we conclude torsional error should always be continuously monitored and corrected during laser ablation, particularly when wavefront-guided ablation is performed. Therefore, a monitoring system that measures the torsional error should be established, and the errors should be corrected as minutely as possible during excimer laser ablation.
1. Smith EM Jr, Talamo JH. Cyclotorsion in the seated and supine patient. / Cataract Refract Surg. 1995;21:402-403.
2. Smith EM Jr, Talamo JH, Assil KK, Petashnick DE. Comparison of astigmatic axis in the seated and supine positions. / Refract Corneal Surg. 1994;10:615-620.
3. Suzuki A, Maeda N, Watanabe H, Kiritoshi A, Shimomura Y, Tano Y. Using a reference point and videokeratography for intraoperative identification of astigmatism axis. / Cataract Refract Surg. 1997;23:1491-1495.
4. Tjon-Fo-Sang MJ, de Faber JT, Kingma C, Beekhuis WH. C ye lo - torsion: a possible cause of residual astigmatism in refractive surgery. / Cataract Refract Surg. 2002;28:599-602.
5. Becker R, Krzizok TH, Wassili H. Use of preoperative assessment of positionally induced cyclotorsion: a video-oculographic study. Br J Ophthalmol. 2004;88:417-421.
6. Swami AU, Steinert RF, Osborne WE, White AA. Rotational malposition during laser in situ keratomileusis. Am J Ophthalmol. 2002;133:561-562.
7. Chernyak DA. Cyclotorsional eye motion occurring between wavefront measurement and refractive surgery. / Cataract Refract Surg. 2004;30:633-638.
8. Bueeler M, Mr oc hen M, Seiler T. Maximum permissible torsional misalignment in aberration-sensing and wavefront-guided corneal ablation. / Cataract Refract Surg. 2004;30:17-25.