| OCR Text |
Show ORIGINAL CONTRIBUTION Perimetry While Moving the Eyes: Implications for the Variability of Visual Field Defects Armin Toepfer, MD, Erich Kasten, PhD, Tobias Guenther, PhD, and Bernhard A. Sabel, PhD Background: In standard perimetry, subjects fixate so that saccades are reduced and testing precision is increased. However, because vision in daily life requires eye movements, it is appropriate to assess visual fields during eye movement. Methods: Perimetry was carried out in 8 healthy sub-jects and in 16 patients with visual field defects under conditions of a stable and moving fixation spot. Eye movements were simultaneously recorded with an eye tracker. Outcome measures included stimulus detection, variability of visual field border, and saccade amplitudes. Results: Perimetric performance during stable fixa-tion was comparable to that during eye movement. All subjects showed 92%-96% correct detections of the fixation controls and a stable and comparable blind spot position in the stable and moving fixation spot conditions. The eye tracker revealed that 97% of the time the eyes were positioned within 61 from fixation. Conclusions: Visual fields obtained by perimetry while moving the eyes is comparable to standard perimetry in which a stable fixation spot minimizes eye movements. (J Neuro-Ophthalmol 2008;28:308-319) Perimetric results are prone to eye movement artifacts (1,2). In standard perimetry, patients are asked to continuously look at a fixation spot test to reduce the influence of saccadic eye movements. Conventional perim-eters typically include methods to control fixation errors. Despite such measures to control eye movements, visual fields are somewhat unstable, with test-retest variability of approximately 62 of visual angle on conventional and computer-based perimetry (3,4). Institute of Medical Psychology, Otto von Guericke University Magdeburg, Magdeburg, Germany. Address correspondence to Bernhard A. Sabel, PhD, Institute of Medical Psychology, Otto von Guericke University Magdeburg, Leipziger Strasse 44, D-39120 Magdeburg, Germany; E-mail: bernhard.sabel@med. ovgu.de In real-life situations, such as during locomotion or driving a vehicle, it is necessary to scan the surrounding environment with eye movements to avoid collisions and accidents (5-7). Standard perimetry may therefore be considered somewhat artificial. In fact, patients often com-plain about blurred vision, double images, and hallucina-tions, which may result from a lack of eye movements. These side effects of constant fixation may be explained in part by Troxler's fading effect (8). Troxler noticed in 1804 that when one fixates a particular point, a stimulus presented in the peripheral visual field will gradually fade away and disappear after about 20 seconds. The effect is enhanced if the stimulus is small and of low contrast (8). Spillmann et al (8) found that even moving targets will rapidly fade quickly in the field periphery. It is a well-known principle in perception that unvarying stimuli soon disappear from awareness. To avoid this effect, the eyes must constantly move. If they were perfectly motionless, photoreceptors would not be sufficiently stimulated and visual perception would ‘‘bleach.'' Holding the eye position constant during fixation for up to 20 minutes at a time creates discomfort and inattention which could, in turn, produce spurious results (9). To eliminate the need for continuous fixation, oculokinetic perimetry, in which the patient moves the eyes around a central static target to look sequentially at an array of numbers, was developed (10) as a method of visual field assessment. When fixation on a number is accompa-nied by disappearance of the central target, that number is deleted from a recording chart. Inversion of the recording chart gives a plot of the central visual field (11,12). Because the precise measurement of visual fields is critical to determine the size of the deficit and possible recovery of visual function, we have created a perimetric task that uses a moving fixation spot that might be more natural. This moving fixation spot was intended to simulate a more physiological situation by inducing smooth pursuit eye movements and thereby reduce the influence of random eye movements during perimetry. Our experiment follows those of other investigators. A moving fixation spot perimeter, computer-assisted moving eye perimeter (CAMEC), was described by 308 J Neuro-Ophthalmol, Vol. 28, No. 4, 2008 Perimetry While Moving the Eyes J Neuro-Ophthalmol, Vol. 28, No. 4, 2008 Johnston et al (13), whereby the patient looks at a moving fixation target on a high-resolution monitor and tries to keep it inside a circle using a joystick. This test was developed for children to increase their motivation during perimetry. However, this method requires continuous eye movement, which increases the retinal area each stimulus can excite. It also binds attention to the fixation spot, which leads to decreased attention for peripheral target stimuli. Consequently, the position of the blind spot was found at rather variable distances from the fovea (14). Mutlukan and Damato (15) investigated children with the blind-spot program of the CAMEC and the Dicon Auto-Perimeter, which has a moving fixation spot. The blind spot was detected in 75% of the 32 children by the Dicon Auto- Perimeter and in 100% by CAMEC. CAMEC allowed better detection and quantification of scotomas in patients older than 4 years. Mutlukan et al (16,17) investigated the position of the blind spot and detection of glaucomatous visual field loss with a multi-fixation campimeter. This device has a central test stimulus and a series of numbered fixation targets and uses the patient's eye movements to position the stimulus in the visual field. Haarmeier and Thier (18) compared a stationary and a moving fixation spot to test whether smooth pursuit eye movements improve the detection of speed changes. To avoid problems such as the Troxler phenomenon, we have developed a computer program for perimetry with an oscillating fixation spot. We hypothesized that this moving spot would 1) reduce saccades into the periphery, 2) improve stability of fixation, 3) increase precision of perimetry, and 4) improve comfort. Because nothing is known about the optimal amplitude and frequency of oscillation of this moving fixation spot, we first carried out a pilot study to address this issue. We then compared visual fields under stationary and moving fixation spot conditions. METHODS Participants We recruited 8 healthy subjects (5 men and 3 women, mean SD age 42 6 12.2 years) and 16 patients with visual field defects (11 men and 5 women, age 55 6 17 years) through newspaper advertisements (Table 1). Visual field defects were due to glaucoma (n = 2), cerebral bleeding (n = 1), retrochiasmal ischemic stroke (n = 6), surgery for brain tumor (n = 1) and epilepsy (n = 1), retinal detachment (n = 1), optic neuritis (n = 1), multiple sclerosis (n = 2), and central retinal artery occlusion (n = 1). Figure 1 displays the visual fields obtained with high-resolution perimetry (HRP) for all study participants. Because the aim of our study was to compare performance under different perimetric testing conditions and not to describe a particular patient sample, we believed that a more homogeneous patient population might have been more adequate, but it would have reduced our ability to generalize our findings to a heterogeneous patient population as typically seen in clinical practice. Furthermore, for the determination of reaction time and fluctuation of visual field defects, the underlying cause of the deficit did not matter. The control subjects and the patients were compa-rable with respect to age. Inclusion criteria for the control subjects were no detectable visual field defects and no known neurological or ophthalmological illnesses. The inclusion criterion for the patients was the documentation of a visual field defect by at least one perimetric evaluation before study entry. Exclusion criteria for all participants were age <18 years, visual acuity <0.15, hemispatial neglect, severe cognitive deficits making them unable to comply with instructions, photosensitive epilepsy, psycho-sis, attention deficits, impairments in motor performance, and a history of having undergone more than five perimetry tests during the last 12 months. Perimetric Procedure Suprathreshold HRP tests were carried out mono-cularly with the eyes at a 30-cm distance from a 17-inch computer monitor. A grid of 475 points within ±27 horizontally and ±22 vertically was tested. The stimulus size was 0.15° with 76 cd/m2 luminescence and 38 cd/m2 background luminescence (19). We compared perimetric performance under stationary and moving fixation spot conditions. For the stationary condition, we carried out three repeated tests. For the moving fixation spot condition, we carried out two repeated tests with a slower speed (2°/s) and two repeated tests with a faster moving fixation spot (speed 3°/s). Each perimetry session lasted about 20 minutes. To reduce head movement artifacts, the head was fixated on a chin-forehead rest. During the test, horizontal eye movements were measured by an infrared eye tracker with high temporal and spatial resolution (Chronos Vision Eye Tracker, measuring 200 frames/s with accuracy of approximately ±0.1°). We have used the stationary fixation spot in our studies previously (19-23). It has a visual angle of 0.14. The fixation control consists of a color change detection task. At random intervals, the fixation spot changes its color from light green to light yellow. The subject must fixate continuously and respond to these color changes by pressing a key on the computer key board each time the color change is detected. The number of correct detections of such color changes is used as a measure of fixation ability. In contrast, the moving fixation spot swings horizontally ±1° in each direction at a speed of 2 or 3/s. Subjects were instructed to follow this spot with eye movements while holding the head steady. At the inflection point, the fixation spot took a 200 ms ‘‘break'' (stationary 309 J Neuro-Ophthalmol, Vol. 28, No. 4, 2008 Toepfer et al TABLE 1. Demographic and etiologic data of study sample Case C1 C2 C3 C4 C5 C6 C7 C8 P1 P2 P3 P4 P5 P6 P7 P8 P9 P10 P11 P12 P13 P14 P15 P16 Sex MM M F F F MM M F F FM M M F M M M M FM M M Age (years) 45 64 44 29 45 23 43 44 78 54 42 65 30 74 73 68 72 63 41 40 58 24 39 63 Tested Eye RR L L L L LL R R L LL R R R R R R L RR R L Acuity Right Eye 0.9 0.8 1.40 1.24 0.63 1.00 0.7 1.25 0.4 0.8 0.20 0.60 0.80 0.15 0.80 1.00 0.60 0.15 .60 0.80 0.40 0.80 0.60 0.80 Acuity Left Eye Lesion Side 0.8 0.8 1.40 1.00 0.80 1.00 0.8 1.25 0.6 0.8 0.50 0.16 0.80 0.15 0.80 1.00 1.00 0.15 .50 0.80 0.40 0.80 0.60 1.00 -- - - - - -- R B B L R R L B L B R R R L R R Lesion Location -- - - - - -- Optic radiation Optic nerve Retina Retina Optic radiation Retina Cortical Optic radiation and cortical Cortical Optic nerve Retinal Cortical Optic nerve Optic radiation and cortical Optic radiation and cortical Cortical Etiology Normal Normal Normal Normal Normal Normal Normal Normal Surgery due to brain tumor Multiple sclerosis Glaucoma Glaucoma Hippocampectomy right (surgery for epilepsy) Right retinal thrombosis Ischemic infarct Ischemic infarct Ischemic infarct Multiple sclerosis Right retinal ablation Ischemic infarction Neuritis nervi optici Intracerebral bleeding Ischemic infarct Ischemic infarct F, female; M, male; L, left; R, right; B, both sides of the visual field. position) while a perimetric target stimulus was presented for 150 ms elsewhere on the screen at one of 475 positions. To control for possible false alarms, the stimulus was given only in 80% of all inflexions. To validate the results of the suprathreshold HRP with conventional near-threshold perimetry, we also carried out 3 examinations on the Tuebinger Automated Perimeter 2000 (24). Here the stimulus size was 0.14° and the intensity varied stepwise between an individual threshold level and 1000 cd/m2. The inner 30 of the visual field were tested with a 90-point test grid. Background light was 10 cd/m2. Each target point detected at a luminescence of 100 cd/ m2 or below was defined as being ‘‘intact.'' Between perimetric sessions, participants were allowed a break of at least 15 minutes. No more than four tests were carried out on a given day. After each test, a questionnaire was used to assess the level of discomfort (exertion) and possible side effects during the test. Each patient had to carry out all perimetric tests within a maximum period of 2 weeks. We decided against the use of a refractive error correction because contact lenses or spectacles would have interfered with the proper function of an eye tracker (see below), which uses cornea reflections to measure eye position. This step is permissible because, although the detection threshold is elevated when the refractive error is not corrected, neither the localization of the visual field defect nor the variability of perimetric performance is altered. Eye Movement Recordings During HRP eye movements were recorded with a Chronos Vision Eye Tracker (23). Before each 310 q 2008 Lippincott Williams & Wilkins Perimetry While Moving the Eyes J Neuro-Ophthalmol, Vol. 28, No. 4, 2008 FIG. 1. Visual field charts obtained by high-resolution perimetry showing detection performance in normal subjects (C1-C8) and patients with visual field defects (P1-P16). examination, this system was individually calibrated using a fixation spot and 4 points at a distance of 10s from the center. The coordinates of the eye position were obtained by detecting the center of the pupil 200 times/s. With use of this technology, smooth pursuit eye movements created by the subjects as they followed the uniform motion of the horizontally swinging fixation spot could be visualized as a regular sinus curve in the x-axis eye tracker data (Fig. 2). In contrast, saccades leaving the path of the swinging fixation spot were seen as clearly visible spikes away from the sinus-like baseline curve. By quantifying such spikes, we were able to differentiate between smooth pursuit eye movements and aberrant saccadic eye movements. The 311 J Neuro-Ophthalmol, Vol. 28, No. 4, 2008 Toepfer et al FIG. 2. Smooth pursuit eye movements following the moving fixation spot of the horizontal x-axis measured by an eye tracker. The point is swinging at an amplitude of 2° (=2 boxes in ordinate); 1 box in abscissa = 2.5 seconds. number of such aberrant saccadic drifts was counted as an indicator of fixation instability. To be counted as a deviation from the sinus-like baseline curve, the following criteria (25-34) applied: • Saccadic duration between 15 and 1000 ms; • Saccadic velocity of at least 15°/s up to 750°/s; • Deviation from the (moving or stationary) fixation spot for at least 100 ms. We classified all saccadic eye movements within these limits as aberrant drifts. Movements that did not qualify for this definition were defined as head or body movements or other artifacts and were ignored. The analysis of eye tracker raw data was done by TETGaze-count, a program developed in our laboratory by one of the authors (TG). Outcome Measures One goal of our experiment was to show the degree to which visual field variability in perimetry is influenced by eye movements. Therefore we analyzed the variability of the visual field border between the intact and the defective area by measuring the horizontal distance from the 0-vertical meridian at positions 0°, 5°, 10°, 15°, and 20° of visual angle above and below the visual field center (Fig. 3). In patients with unilateral visual field defects, these measurements refer to the defective half-field only. In patients with visual field defects in both hemifields, we calculated the mean distance between the vertical meridian and the left and right visual field borders. Many patients had no clear border between the intact and the defective visual fields but a large transition zone and scattered relative defects. Because single ‘‘blind'' positions in an otherwise intact field may be artifacts due to blinking, we FIG. 3. Results of high-resolution perimetry. Black squares indicate areas of the visual field in which the subject did not respond to stimulus presentation. White indicates response in the predetermined time window. The arrows show how the visual field borderline was determined at 9 different positions. This example was taken from the left eye from a patient with a left homonymous hemianopia. The blind spot was only measured in healthy subjects (circle, fixation spot; black, blind; white, intact visual areas). The x- and y-axes show degrees of visual angle 312 q 2008 Lippincott Williams & Wilkins Perimetry While Moving the Eyes J Neuro-Ophthalmol, Vol. 28, No. 4, 2008 devised the following definitions of a visual field border. We measured the distance between the vertical meridian and the first stimulus location with a minimum of two undetected stimuli in a row. The standard error of the variability of the visual field borders between different tests was taken as an indicator of test reliability. In healthy subjects without defective visual field areas, the x-coordinate of the blind spot was measured and its fluctuation was compared between the tests. Subjective Reports In our past studies, patients frequently reported side effects of visual field testing. Therefore we developed a 7-item questionnaire to assess the following side effects of testing: Tears in eyes during the perimetry Burning of eyes during testing Double images of the central fixation spot Blurring of the fixation spot Visual hallucinations Foggy images Illusion of a bright frame around the test field Subjects had to rate each side effect on a scale from 0 (never occurring) to 5 (occurring frequently). In addition, we asked the subjects to rate their overall exertion after every separate perimetric investigation. The participants did not know which test method was the newly developed one (single-masked condition). Statistical Analysis Variability of visual field borders in patients and positions of the blind spot coordinates in subjects were measured by calculating the mean variances of these coordinates for each subject during each of the following three conditions: perimetry with stable fixation, with a slower moving fixation spot, or with a faster moving fixation spot We used analysis of variance (ANOVA) for repeated measurements to compare the variability between the three test conditions. Pair-wise post hoc comparisons between conditions were done using Bonferroni adjust-ments. To analyze the correlation between eye movements and visual field variability, we classified the saccadic drifts as small (1-2°) or large ( 2°) and calculated the Pearson's rank correlation coefficient with measures of visual field variability (standard errors of measurement). All subjects were assigned in stratified groups which determined the order in which the tests were carried out. Randomization was done by computer-generated group assignments. Possible differences in comfort during testing and intensity of side effects among the different perimetric procedures were checked by using ANOVA for repeated measurements. RESULTS Fixation Stability In HRP fixation ability was quantified by recording detections of color changes of the fixation spot. With the stationary fixation spot, we found 96.3 ± 0.57% (mean SD) correct detections of this color change. With the slower moving fixation spot, we found 92.3 ± 2.55% correct detections of this color change. With the faster moving fixation spot, we found 93.9 ± 1.38% correct detections of this color change, results that were not significantly different (SEM between 1.1 and 2.3, significance between 0.32 and 0.67). In the Tuebinger automated perimetry (TAP) test, we found only 89.26 ± 14.04% correct responses to the fixation task, and this was not statistically different from the HRP-fixation performance. Eye Movements During Perimetry In the eye tracking system, we found that in 97% of the test durations, the eyes fixated within a window of ± 1. We found no significant differences between patients and control subjects in the stationary (patients 98.0% and control subjects 97.5%) or moving fixation spot tests (patients, slower moving fixation spot 97.4%; control subjects, slower moving fixation spot 96.8%; patients, faster moving fixation spot 97.8%; and control subjects, faster moving fixation spot 97.0%). Variability of Perimetric Results We studied the general variability of perimetric results by measuring the mean fluctuation of visual field borders within each test group. Average variability of the visual field border within the 16 patients (expressed as the SEM of the eye positions from midline) was between ±2.5° (slower moving fixation spot) and ±3.3° (faster moving fixation spot). The stable fixation spot had an average ±2.6° SEM of the visual field borderline (Fig. 4). The differences were not significant. Average variability of the center x-coordinate of the blind spot was ±0.49° (conventional perimetry on TAP), ±0.56° (slower moving fixation spot), or ±0.35° degrees (faster moving fixation spot). The stable fixation spot in HRP had a SEM of 0.43°. Differences were not significant (Wilks lambda multivariance test 0.785, X = 0,882). Relationship of Visual Field Variability and Eye Movements Visual field fluctuations were significantly correlated with eye movements in the condition using the stable fix-ation spot only (Spearman's r = 0.645, P = 0.032) (Table 2). There were no significant correlations in the conditions with the moving fixation spots. 313 J Neuro-Ophthalmol, Vol. 28, No. 4, 2008 Toepfer et al FIG. 4. Intertest variability (expressed as standard error) of the visual fields in patients and control subjects under different perimetric conditions: stable fixation point, slower or faster moving fixation point on a computer-based high resolution perimetry and when assessed by Tuebinger automated perimetry. Left. Average variability of the blind spot x-axis in healthy subjects. Right. Average variability of the visual field borderline (x-axis) in patients as determined by the method shown in Fig. 2. The y-axis displays the variability (standard error of the mean) in degrees of visual angle. The figure shows that the variability is greater in patients that in healthy control subjects. Reliability of Different Perimetric Methods in the Blind Spot Measurements In control subjects, only the blind spot could be used as a measure of variability of visual field investigations. In the stable fixation spot condition, the average position of the blind spot in healthy subjects was 16.35 ± 0.55° from the midline at the x-coordinate and 1.73 ± 1.36° at the y-coordinate. This value did not differ significantly from the other methods using the slower (x 16.69 ± 0.60° and y 1.83 ± 1.68°) or faster moving fixation spot (x 16.51 ± 0.78° andy 1.85 ± 1.58°) and TAP (x 16.11 ± 0.88° andy 1.74 ± 1.63°). Correlation Between Variability of Blind Spot Measurements and Eye Movements The correlation between the variability of the blind spot coordinates and saccadic eye movements was small and negative (stable fixation spot: r= 0.434, NS) (Table 3). TABLE 2. Visual field fluctuations in relation to eye movements Correlation Between Visual Field Fluctuations and Total Eye Movements Correlation Between Visual Field Fluctuations and Eye Movements >2 degrees Test Stable fixation spot Slower moving fixation spot Faster moving fixation spot Spearman Rho 0.645 0.517 0.2 P < 0.032 0.154 0.534 Spearman Rho 0.609 0.142 0.042 P < 0.047 0.715 0.897 314 q 2008 Lippincott Williams & Wilkins Perimetry While Moving the Eyes J Neuro-Ophthalmol, Vol. 28, No. 4, 2008 TABLE 3. Correlation betweeen blind spot field fluctuations and eye movements Correlation Between Blind Spot Field Correlation Between Blind Spot Field Fluctuations and x-Axis of Total Fluctuations and x-Axis and Eye Eye Movements Movements >2 degrees Test Spearman Rho P < Spearman Rho P < Stable fixation spot 0.434 0.282 0.587 0.126 Slower moving fixation spot 0.437 0.207 0.039 0.916 Faster moving fixation spot 0.459 0.156 0.510 0.109 This result shows that a higher variability of the blind spot position was not related to the number of saccadic eye movements. Subjective Comfort During Tests and Side Effects The trials with a moving fixation spot did not lead to less exertion. Exertion showed values of 1.92 6 1.41 with the slower moving fixation spot and 2.12 6 1.24 with the faster moving fixation spot in the control subject group. Exertion values in the visual impaired group were between 1.86 6 0.87 (slower moving fixation spot) and 2.36 6 1.23 (faster moving fixation spot). There were no significant differences between the tests with stable and with a moving fixation spots (Wilks lambda multivariance test l = 0.809, P = 0.226). The control subject and patient groups did not differ (Fig. 5). Side Effects Side effects did not differ significantly between the three perimetric methods. The average score for all seven items was between 3.3 6 3.4 (faster moving fixation spot) and 4.5 6 5.0 (stable fixation spot). The slower moving fixation spot had an item score of 4.7 6 5.2. TAP had a 3.7 6 3.1 item score. There was no significant advantage for the moving fixation spot in any of the items. Thus, the moving fixation spot produced as much exertion during perimetry and the side effects could not be reduced by this new fixation method (Fig. 6). DISCUSSION The main goal of our experiment was to determine whether visual fields obtained by perimetry with and without eye movements differed from each other. Another goal of the study was to evaluate whether computer-based perimetry with a moving fixation spot might be more convenient for patients and reduce possible influences of continuous fixation of a stationary spot as used during standard perimetry. Moving fixation spot perimetry was found not to be more comfortable for control subjects or patients than perimetry where the eyes have to be suppressed by fixating a stable fixation point. There were no significant differences either in the assessment of comfort or in the reduction of uncomfortable side effects. Our finding is not in agreement with that of Wong et al (35), who noted that use of the Humphrey Field Analyzer and the Dicon TKS 4000, with a moving fixation target during Dicon testing, makes visual field testing more comfortable for patients. Asman et al (36) also evaluated the fixation accuracy of static (Humphrey Field Analyzer) and kinetic fixation (Dicon) perimetry and determined their ability to detect the absolute scotoma of the physiologic blind spot. In patients with glaucoma, the frequency of fixation errors was significantly greater for kinetic (17.2%) than for static (10.2%) methods. The authors concluded that kinetic perimetry was associated with greater fixation inaccuracy and underestimation of the absolute scotoma at the physiologic blind spot. Our method of presenting target stimuli at the inflection points of the moving fixation spot led to visual detection that was comparable to perimetry with a stable fixation spot. This result implies that introducing eye movements has little or no effect on perimetric perfor-mance. This result also contrasts with prior observations by Demirel (4), who found a significant influence of eye movements on perimetry results but only if the eye positions exceeded ±1° amplitude more than 20% of the time. Also, in patients with glaucoma, Henson et al (37,38) found that fixation was within 0.5° of the target in only 7% of the presentations, whereas it was at best within this range in about 60% of presentations. By contrast, in our study we found that fixation was within an area of ±1° about 97% of the time (during only 3% of the testing time were saccades larger than 1°). This result is similar to that of Kasten et al (23), who found eye positions ±1° away from a stable fixation spot in 81.6% and up to 2 away in 94.8% of the overall test-ing time. We clearly found that saccades have smaller effects on perimetric performance than other studies have suggested. 315 J Neuro-Ophthalmol, Vol. 28, No. 4, 2008 Toepfer et al FIG. 5. Results of the exertion question. Test conditions were the same as in the legend to Figure 4. Exertion was rated by the subjects on a scale from 0 (easy) to 5 (very hard). The bars display the different fixation point conditions for the control subjects and patients combined: stable fixation spot, slower and faster moving fixation spot, and exertion for the standard perimeter. In previous investigations (20), variability of the visual border position was ±2.2° in perimetric tests repeated five times in each patient). In our study, the SEM was ±2.6° in three independent measurements. A significant positive correlation between eye movements and this SEM of visual field borders was found only within tests using a stable fixation spot. Tests with the moving fixation spot did not show any significant influence of eye movements on visual field fluctuation. We conclude that eye movements only slightly influence perimetric results. To be able to estimate the variability of visual field testing under the different perimetry conditions, we assessed the variability of the blind spot position in the control subjects. It was between ±0.42° and ±1.04°, less than half of what we found in the group of patients with a variety of visual field defects. In control subjects, the blind spot variability and in patients the variability of visual field defects was relatively independent of eye movements because we did not find a positive correlation between eye movements and visual field fluctuations. Apart from eye movements, then, there are other influences such as atten¬ tional fluctuations (39,40) that contribute to the fluctuation of visual field border positions in patients. Another influ-ence may be the existence of ‘‘areas of residual vision'' (also termed ‘‘relative defects'' or ‘‘transition zones''), which are typically located between the visual field defect and the intact visual field. In these areas, stimuli are seen unreliably, perhaps as a functional expression of residual visual structures in the damaged brain (21). The existence of areas of residual vision in patients might also be corroborated by the fact that variation of the blind spot in healthy persons is significantly lower (0.77°) than variation 316 q 2008 Lippincott Williams & Wilkins Perimetry While Moving the Eyes J Neuro-Ophthalmol, Vol. 28, No. 4, 2008 FIG. 6. Sum score for side effects of a questionnaire with 7 items, which could be rated on a scale from 0 (never) to 5 (very often). The x-axis is the same as that in Figure 4. There were no significant differences between the test conditions. of the visual field border positions (2.6°). Based on these values, we estimate that only up to one fourth of the var-iability of visual field border is caused by eye movements. Several additional factors contribute to visual field variability in normal subjects and in patients with optic neuritis, glaucoma, or ocular hypertension. Henson et al (38) found decreased brightness sensitivity, whereas stimulus eccentricity, patient age, fixation loss rate, and false-positive rate did not predict response variability. In our study, a higher variability of the position of the center of the blind spot was not related with an increasing number of saccadic eye movements. This result could be explained by the fact that saccades exceeding 2° of visual angle occurred only 1% of the time. Visual field fluc-tuations were significantly correlated with eye movements in the stable fixation condition (R = 0.645), whereas there were no significant correlations in the conditions with the moving fixation spots. Therefore, it remains unclear whether patients with large areas of residual vision make more saccadic eye movements and if they are directed toward or away from the hemianopic field. This issue requires additional research. It has been argued that patients with visual field defects compensate for their deficit by making more frequent eye movements toward the hemianopic field and that visual field enlargements found after vision restoration therapy (VRT) may be an artifact of such eye movements (41). On the other hand, Henson et al (37,38) concluded that fixation errors, although contributing to variability, are not the major cause of the increased variability seen at locations with reduced sensitivity. In our study, patients with visual field defects did not show more eye movements than visually healthy subjects. Large saccades were very rare. This result argues against the proposal that areas of residual vision (‘‘relative defects'') or their enlargement are exclusively an artifact of eye movements (41,42). Even so, when studies on recovery and restoration of visual field 317 J Neuro-Ophthalmol, Vol. 28, No. 4, 2008 Toepfer et al defects are carried out (43-56), eye movement recordings are needed to estimate how much eye movement has influenced the test results. In summary, perimetry during ongoing eye move-ments induced by a moving fixation spot leads to results comparable to those of conventional perimetry. We found similar position, extent, and fluctuation of visual field defects with both methods in patients and also comparable positions and fluctuations of the blind spot in healthy control subjects. Fixation was within ±1° of the visual field center during 97% of the perimetric testing time in patients and control subjects. The moving fixation spot did not lead to more stable test results or to fewer side effects, and it neither increased nor decreased the number of large sac-cades. Thus, moving the eyes has no significant effect on perimetry, leading us to conclude that rudimentary visual stimulus detection does not depend on eye movements. REFERENCES 1. Heijl A. Studies on computerized perimetry. Acta Ophthalmol Suppl 1977;132:1-42. 2. Heijl A, Krakau CE. A note of fixation during perimetry. Acta Ophthalmol (Copenh) 1977;55:854-61. 3. Eizenman M, Trope GE, Fortinsky M, et al. Stability of fixation in healthy subjects during automated perimetry. Can J Ophthalmol 1992;27:336-40. 4. Demirel S. Optimizing the reliability of automated perimetry for the early detection of visual disorders.. Masters Thesis. Melbourne, Australia: Department of Optometry and Vision Sciences, University of Melbourne; 1995. 5. Lencer RM, Clarke AH. Influence of optokinetic and vestibular stimuli on the performance of smooth pursuit eye movements: implications for a clinical test. Acta Otolaryngol 1998;118:161-9. 6. McAuley JH, Rothwell JC, Marsden CD. Human anticipatory eye movements may reflect rhythmic central nervous activity. Neurosci-ence 1999;94:339-50. 7. Goldberg ME, Bisley JW, Powell KD, et al. Saccades, salience and attention: the role of the lateral intraparietal area in visual behavior. Prog Brain Res 2006;155:157-75. 8. Hunzelmann N, Spillmann L. Movement adaptation in the peripheral retina. Vision Res 1984;24:1765-9. 9. Wolffe M. Visual field assessment-a comparative study of four perimeters. Optician 2001;222:16-8. 10. Damato BE. Oculokinetic perimetry: a simple visual field test for use in the community. Br J Ophthalmol 1985;69:927-31. 11. Alvarez E, Damato BE, Jay JL, et al. Comparative evaluation of oculokinetic perimetry and conventional perimetry in glaucoma. Br J Ophthalmol 1988;72:258-62. 12. Felius J, Langerhorst CT, van den Berg TJ, et al. Oculokinetic perimetry compared with standard perimetric threshold testing. Int Ophthalmol 1992;16:221-6. 13. Johnston SC, Damato BE, Evans AL, et al. Computerised visual field test for children using multiple moving fixation targets. Med Biol Eng Comp 1989;27:612-6. 14. Mutlukan E. Topographical reproducibility of small scotoma using computerised dynamic fixation target. Doc Ophthalmol 1995;91: 223-31. 15. Mutlukan E, Damato BE. Computerised perimetry with moving and steady fixation in children. Eye 1993;7:554-61. 16. Mutlukan E, Damato BE, Tavadia SM. Clustering of fixation targets in multifixation campimetry. Eye 1993;7:131-3. 17. Mutlukan E, Damato BE, Jay JL. Clinical evaluation of a multi-fixation campimeter for the detection of glaucomatous visual field loss. Br J Ophthalmol 1993;77:332-8. 18. Haarmeier T, Thier P. Detection of speed changes during pursuit eye movements. Exp Brain Res 2006;170:345-57. 19. Kasten E, Strasburger H, Sabel BA. Programs for diagnosis and therapy of visual deficits in vision rehabilitation. Spat Vis 1997;10: 499-503. 20. Kasten E, Wuest S, Sabel BA. Variability of stimulus detection, form discrimination and color recognition with suprathreshold campimetry in brain damaged patients. Neuroophthalmology 1998;20:161-76. 21. Kasten E, Wuest S, Sabel BA. Residual vision in transition zones in patients with cerebral blindness. J Clin Exp Neuropsychol 1998; 20: 581-98. 22. Kasten E, Poggel DA, Muller-Oehring E, et al. Restoration of vision II: residual functions and training-induced visual field enlargement in brain-damaged patients. Restorative Neurol Neurosci 1999;15: 273-87. 23. Kasten E, Bunzenthal U, Sabel BA. Visual field recovery after vision restoration therapy (VRT) is independent of eye movements: an eye tracker study. Behav Brain Res 2006;175:18-26. 24. Lachenmayr BJ, Vivell PM. Perimetrie. Stuttgart, Germany: Thieme Verlag; 1992. 25. Aivar MP, Hayhoe MM, Chizk CL, et al. Spatial memory and saccadic targeting in a natural task. J Vis 2005;5:177-93. 26. Welchman AE, Harris JM. Task demands and binocular eye movements. J Vis 2003;3:817-30. 27. Cho S, Lee C. Expansion of visual space after saccadic eye movements. J Vis 2003;3:906-18. 28. Rotman G, Brenner E, Smeets JB. Mislocalization of targets flashed during smooth pursuit depends on the change in gaze direction after the flash. J Vis 2004;4:564-7. 29. Yang Q, Kapoula Z. Binocular coordination of saccades at far and at near in children and in adults. J Vis 2003;3:554-61. 30. Pracejus L. Bewegungsnacheffekt und glatte Augenbewegungen. In: Fachbereich Psychologie Giessen, Germany: Uni Giessen; 2003. 31. Vickers JN. Toward defining the role of gaze control in complex targetting skills. In: Visual Search 2. Bristol, PA: Taylor Francis; 1993:265-85. 32. Wang L. Glissadic saccades: a possible measure of vigilance. Ergonomics 1998;41:721-32. 33. Traisk F, Bolzani R, Ygge Y. A comparison between the magnetic scleral search coil and infrared reflection methods for saccadic eye movement analysis. Graefes Arch Clin Exp Ophthalmol 2005;243: 791-7. 34. Boghen D, Troost BT, Daroff RB, et al. Velocity characteristics of normal human saccades. Invest Ophthalmol 1974;13:619-23. 35. Wong AY, Dodge RM, Remington LA. Comparing threshold visual fields between the Dicon TKS 4000 automated perimeter and the Humphrey Field Analyzer. J Am Optom Assoc 1995;66:706-11. 36. Asman P, Fingeret M, Robin A, et al. Kinetic and static fixation methods in automated threshold perimetry. J Glaucoma 1999;8: 290-6. 37. Henson DB, Evans J, Chauhan BC, et al. Influence of fixation accuracy on threshold variability in patients with open angle glaucoma. Invest Ophthalmol Vis Sci 1996;37:444-50. 38. Henson DB, Chaudry S, Artes PH, et al. Response variability in the visual field: comparison of optic neuritis, glaucoma, ocular hypertension, and normal eyes. Invest Ophthalmol Vis Sci 2000; 41:417-21. 39. Poggel DA, Kasten E, Mu¨ller-Oehring EM, et al. Verbesserung residualer Sehfa¨higkeit bei Patienten mit homonymen Gesichtsfel-dausfa ¨llen durch ra¨umliche Aufmerksamkeit. Z Med Psychol 2005; 14:119-32. 40. Kasten E, Bunzenthal U, Mu¨ller-Oehring EM, et al. Vision restoration therapy does not benefit from co-stimulation: a pilot study. J Clin Exp Neuropsychol 2007;29:569-84. 41. Jamara RJ, Van De Velde F, Peli E. Scanning eye movements in homonymous hemianopia documented by scanning laser ophthal-moscope retinal perimetry. Optom Vis Sci 2003;80:495-504. 318 q 2008 Lippincott Williams & Wilkins Perimetry While Moving the Eyes J Neuro-Ophthalmol, Vol. 28, No. 4, 2008 42. Sabel BA, Trauzettel-Klosinksi S. Improving vision in a patient with homonymous hemianopia. J Neuroophthalmology 2005;25:143-9. 43. Sabel BA, Kenkel S, Kasten E. Vision restoration therapy (VRT) efficacy as assessed by comparative perimetric analysis and subjective questionnaires. Restorative Neurol Neurosci 2004;22: 399-420. 44. Mueller I, Mast H, Sabel BA. Recovery of visual field defects: a large clinical observational study using vision restoration therapy. Restorative Neurol Neurosci 2007;25:563-72. 45. Kasten E, Guenther T, Sabel BA. Inverse stimuli in perimetric performance reveal larger visual field defects: implications for vision restoration. Restorative Neurol Neurosci 2008;26:355-64. 46. Gudlin J, Mueller I, Thanos S, et al. Computer based vision restoration training in glaucoma patients-a small open pilot study. Restorative Neurol Neurosci 2008;26:403-12. 47. Schumacher EH, Jacko JA, Pimo SA, et al. Reorganization of visual processing is related to eccentric viewing in patients with macular degeneration. Restorative Neurol Neurosci 2008;26:391-402. 48. Mitchell DE. A special role for binocular visual input during development and as a component of occlusion therapy for treatment of amblyopia. Restorative Neurol Neurosci 2008;26:425-34. 49. Achtman RL, Green CS, Bavelier D. Video games as a tool to train visual skills. Restorative Neurol Neurosci 2008;26:435-46. 50. Stoerig P. Functional rehabilitation of partial cortical blindness? Restorative Neurol Neurosci 2008;26:291-303. 51. Bergsma DP, Van der Wildt GJ. Properties of the regained visual field after visual detection training of hemianopia patients. Restorative Neurol Neurosci 2008;26:365-75. 52. Polat U. Restoration of underdeveloped cortical functions: evidence from treatment of adult amblyopia. Restorative Neurol Neurosci 2008;26:413-24. 53. Werth R. Cerebral blindness and plasticity of the visual system in children. Restorative Neurol Neurosci 2008;26:377-89. 54. Chokron S, Perez C, Obadia M, et al. From blindsight to sight: cognitive rehabilitation of visual field defects. Restorative Neurol Neurosci 2008;26:305-20. 55. Gall C, Mueller I, Gudlin J. Vision-and health-related quality of life before and after vision restoration training in cerebrally damaged patients. Restorative Neurol Neurosci 2008;26:341-53. 56. Poggel AD, Mueller I, Kasten E. Multifactorial predictors and outcome variables of vision restoration training in patients with post-geniculate visual field-loss. Restorative Neurol Neurosci 2008;26: 321-39. 319 |
