Voluntary Saccadic Oscillations, Resembling Ocular Flutter and Opsoclonus

Two human subjects, who had no signs of other neurologic disorders, produced large amplitude, to-and-fro saccadic oscillations. One subject generated these oscillations intentionally. Eye movement recordings by DC electro-oculography and/or magnetic scleral search coil showed that the movements were bursts of conjugate saccades in opposing directions with no intersaccadic intervals. They were multidirectional (horizontal, vertical, or oblique), had amplitudes up to 40 degrees and had linear or curvilinear trajectories. These characteristics are similar to those of ocular flutter and opsoclonus in patients with brainstem and/or cerebellar disorders. Our observations show that fixation instabilities resembling ocular flutter and opsoclonus can be produced voluntarily.

Journal of Neuro- Ophthalmology 14( 2): 95- 101, 1994. 1994 Raven Press, Ltd., New York Voluntary Saccadic Oscillations, Resembling Ocular Flutter and Opsoclonus Robert D. Yee, M. D., Peter H. Spiegel, M. D., Tetsuto Yamada, M. D., Larry A. Abel, Ph. D., David A. Suzuki, Ph. D., and David S. Zee, M. D. Two human subjects, who had no signs of other neuro­logic disorders, produced large amplitude, to- and- fro saccadic oscillations. One subject generated these oscil­lations intentionally. Eye movement recordings by DC electro- oculography and/ or magnetic scleral search coil showed that the movements were bursts of conjugate saccades in opposing directions with no intersaccadic intervals. They were multidirectional ( horizontal, verti­cal, or oblique), had amplitudes up to 40 degrees and had linear or curvilinear trajectories. These characteris­tics are similar to those of ocular flutter and opsoclonus in patients with brainstem and/ or cerebellar disorders. Our observations show that fixation instabilities resem­bling ocular flutter and opsoclonus can be produced vol­untarily. Key Words: Saccades- Fixation instability- Eye move­ments- Opsoclonus- Ocular flutter- Voluntary nystag- From the Department of Ophthalmology ( R. D. Y., P. H. S., T. Y., L. A. A., D. A. S.), Indiana University School of Medicine, Indianapolis, Indiana; Departments of Neurology and Ophthal­mology ( D. S. Z.), Johns Hopkins School of Medicine, Baltimore, Maryland, U. S. A. This research was supported in part by an Unrestricted Grant and a Research Fellowship Grant to the Department of Oph­thalmology, Indiana University School of Medicine, from Re­search To Prevent Blindness, Inc., New York, New York. Address correspondence and reprint requests to Dr. Robert D. Yee, Department of Ophthalmology, Indiana University School of Medicine, 702 Rotary Circle, Indianapolis, IN 46202, U. S. A. Fixation instabilities can be categorized as nystagmus or saccadic oscillations ( 1). Physiologic nystagmus and pathologic nystagmus consist of involuntary, rhythmic, to- and- fro movements of the eyes. A smooth eye movement moves the eyes away from the fixation target, and usually a fast eye movement or saccade carries them back to­ward the target. In saccadic oscillations a saccade takes the eyes away from the target, and another saccade carries them back toward the target. Sac­cadic oscillations have been categorized based in part on whether or not a pause or intersaccadic interval, during which the eyes are stationary, ex­ists between the two opposing saccades. Square-wave jerks, macrosquare wave jerks, and macro-saccadic oscillations have intersaccadic intervals, whereas voluntary " nystagmus," ocular flutter, and opsoclonus do not. Saccadic oscillations, except voluntary " nystag­mus," are involuntary. That is, they cannot be pro­duced intentionally by normal subjects. Voluntary " nystagmus" consists of burst of very small, back-to- back saccades with no intersaccadic intervals that some normal subjects can generate for a few to several seconds ( 2). It can be distinguished easily from voluntary, repetitive refixations between tar­gets, since the latter are usually much larger in amplitude, have a much lower frequency, and have intersaccadic intervals of about 200 ms. Ocu­lar flutter and opsoclonus occur involuntarily in patients with damage to the brainstem and/ or cer­ebellum ( 3). Ocular flutter is mainly horizontal, whereas opsoclonus is multidirectional. They have larger amplitudes and lower frequencies than vol­untary " nystagmus," and persist for longer peri­ods of time. We describe eye movements in two persons that resemble ocular flutter and opsoclonus. They pro­duced large amplitude saccades in opposing direc- 95 96 R. D. YEE ET AL. tions that had no intersaccadic intervals. We sus­pected that one subject made the eye movements voluntarily, and knew that the other made them intentionally. SUBJECTS AND METHODS Subjects Subject 1 was a man who was referred to us for evaluation of abnormal eye movements. He had a whiplash injury when the rear of his stationary car was struck by another car. He bumped his fore­head on the windshield, did not lose conscious­ness, and had soreness of the posterior neck and back. According to the patient, the next day some­one said that his eyes looked abnormal. Several days later another observer told him that his eyes were jumping around. He complained that he could read for only 10 minutes at a time because he lost his place on the page. However, he could drive a car satisfactorily. Physicians who examined him immediately after the accident and over the next few months did not find abnormal eye move­ments or other ocular or neurologic abnormalities. Several months after the accident an optome­trist, an ophthalmologist, and a neurologist found a fixation instability that they called " nystagmus" or " opsoclonus." There was no history of a pre­ceding viral illness, exposure to toxic chemicals or drugs, ocular disorders, neurologic diseases, or other illnesses. He did not complain of oscillopsia, diplopia, paresis, sensory loss, imbalance, vertigo, tinnitus, or hearing loss. Several neurologic eval­uations revealed no abnormalities, other than the fixation instability. He did not have ataxia, incoor­dination, paresis, or sensory loss. Computed to­mography scans and magnetic resonance imaging scans of the head and neck showed no abnormal­ities. Examination of the spinal fluid, including measurements of immunoglobulins, was normal. Subject 2 was one of the authors ( R. D. Y.). He had no neurologic or ocular disorders and had ex­tensive experience as a normal subject in research using electronic eye movement recordings. He could produce typical voluntary " nystagmus," and tried intentionally to simulate the fixation in­stability demonstrated by Subject 1. Informed con­sent was obtained from both subjects. Recording and Analysis Methods The eye movements of both subjects were re­corded with DC electro- oculography ( EOG). Hor­izontal eye movements were recorded with Ag- AgCl skin electrodes placed at the inner and outer canthi of both eyes. Electrodes above and below the eyes recorded vertical eye movements. The low- pass bandwidth of the EOG system was 0 to 42 Hz. The EOG system has been described in de­tail previously ( 4). The eye movements of Subject 2 were also recorded by a magnetic scleral search coil method. The method was based on the technique described by Robinson for measuring eye move­ments in experimental animals ( 5) and on modifi­cations described by Collewijn and colleagues ( 6) for use in human subjects. The search coil was an annulus of soft plastic contact lens material in which several turns of a fine copper wire had been imbedded ( Skalar, Delft, Netherlands). The con­tact lens was placed on the eye after a topical an­esthetic had been used. The subject sat within a weak magnetic field generated within three pairs of 6- ft diameter coils ( CNC Engineering, Seattle, Washington). The search coil method used a low bandpass filter of 0 to 1,600 Hz and had a linear range of 40 degrees centered about the primary position of gaze. The subjects' heads were firmly secured by a head rest. They were asked to fixate a bright, 0.25- degree diameter, red light produced by a helium-neon laser ( Spectraphysics, Mountain View, Cali­fornia) and projected onto a dim, flat screen in front of them. Analog eye movement signals were digitized at 200 Hz by an on- line microcomputer and displayed on a polygraph. Eye velocity was calculated by the two- point difference formula. The microcomputer analysis system has been de­scribed in detail previously ( 4). RESULTS Subject 1 The neuro- ophthalmic examination of Subject 1 was normal, except for the fixation instability. The instability was intermittent, and was not present when he described the history of his medical prob­lems and during the measurement of Snellen vi­sual acuity, which was 20/ 20 in each eye. How­ever, when he was instructed to fixate a stationary or moving target, large amplitude, rapid, to- and-fro, horizontal and oblique, conjugate eye move­ments occurred. The EOG recordings could not be calibrated ac­curately because his eyes did not follow the laser target. The gain settings of the amplifiers were placed at those used for most normal subjects in our laboratory. At these settings an eye movement of 1 degree produces a polygraph pen deflection of 1 mm. Figure 1 shows his eye movements when he was asked to fixate the stationary target located at the center of his visual fields. I Neuro- Ophthalmol Vol. 14, No. 2, 1994 VOLUNTARY SACCADIC OSCILLATIONS 97 15 .:;.. " - - U " ^ A : :- T\^ , A H • • • • - \ r ' If ^ ^ F :, ••>• : : | • .. A ! • / iVh i , \ U •: J" r= •• r- • • • • . . / • . : ' . : • • ( . : :" • : . , - | : • • • [• j . . ] w- i 1... : • n - - :- i I n 1 1 1 . i is I I - I-•, 1 : r: . | i ! i i ! • ! , - : -: • •• ' ' . , , - . ! - . ! I i i ! i ^_ r- n i i j ^ • i \ i ! i L _ hi_!_ KM i ; | ! i j i V • F m i i I i ' l !• '! •: >•...- ^ i ! i j I I : : •''..:•• H^- l^- V^ Wy r ^ T ^ ^ f c ~ T u T : i w v ' • U ' A ' ; . i I i' •'. . • I I M i M . *< 15^ I:" • : : ; • .: i- 7:: • ' \-\ l '£•: , • : : P :: i : ! •; ::- l :-.\. - . i . I'!:. •"-• p[ A' 7 V ;:: V•\:~~\ • | , | - : i - - ' I I :• -•• I- ~ W~ iiiM/ ih. ; * 1 - T - • v l- l . ' • I . 1 -^ U-' f tiLr1r " 3_ I "= T t drrt : • • • • ! : \ \ i i 1 i • ~ T i " I i. /^ I • v ! : • M1 L \ S^ s~"\ . I N • •. ••: '•;:•• I I I ! ' " 1 i,!: ' ^ A ; - ! r • • A " 1 \ i ! * r : : xi7 V - I U / :, :•*/ A M J U i^ \ n r - l\- - ! r vi llivl p" fvt' I: F M' 1 : 1 ! i ! 1 ! 1 !• 1 sec FIG. 1. Subject 1 " fixating" target at center. Top: One- second intervals. Middle: Vertical eye movements. Bottom: Horizontal movements. Couplets and triplets of to- and- fro saccades have short or no intersaccadic intervals. Black arrows show oblique, back- to- back saccades resembling opsoclonus. White arrow shows small saccadic oscilla­tions resembling voluntary " nystagmus." The eye movements were oblique ( horizontal and vertical), to- and- fro saccades. Although we could not accurately measure their velocities, they were clearly greater than those of most smooth eye movements. Figure 2 shows the peak velocity-amplitude relationships of horizontal eye move­ments in the right eye. The precise amplitudes and velocities could not be measured, but the velocity-amplitude relationships had the curvilinear shapes that are characteristic of saccades ( 7). The saccades were conjugate. Figure 1 shows EOG recordings when the subject was asked to fixate the target at center. Some to- and- fro sac­cades had intersaccadic intervals varying from about 80 to 280 ms. There was no consistent posi­tion of the eyes during the intervals to identify center gaze. Several pairs of to- and- fro saccades had no intersaccadic intervals and resembled beats of ocular flutter. A continuous series of to-and- fro saccades had no intersaccadic intervals and a frequency of about 7 to 8 Hz, resembling ocular flutter and opsoclonus. In Fig. 1 a horizontal pendular oscillation of very small amplitude ( about 1 degree) and high frequency ( about 25 Hz) is present. These characteristics are similar to those of voluntary " nystagmus." The saccadic oscillations were present during tests of smooth pursuit, optokinetic nystagmus, and vestibulo- ocular responses. However, the in­tersaccadic intervals between some saccades show­ing smooth movements in the expected directions, for example, in the direction of the smooth pursuit target, in the direction of optokinetic drum mo­tion, and in the opposite direction to that of rotary chair rotation. Subject 2 Subject 2 observed Subject l's eye movements and tried to produce similar saccadic oscillations. He attempted to move his eyes continuously back-o o 0) • • • * % '.% * left right amplitude ( deg) FIG. 2. Peak velocity vs amplitude of saccadic oscil­lations in Subject 1. The curvilinear relationships re­semble those of saccades in normal subjects. / Neuro- Ophlhalmol, Vol. 14, No. 2, 1994 98 R. D. YEE ET AL. and- forth in all directions, so that he constantly perceived oscillopsia and could not see any object in the visual surround clearly. Figure 3 shows his eye movements recorded with EOG when he used this strategy. The large- amplitude, conjugate, rapid to- and- fro movements were oblique, had fre­quencies of 2 to 5 Hz and usually showed no in-tersaccadic intervals. Figure 4 presents his eye movements recorded with magnetic search coil during another attempt to generate saccadic oscil­lations. The eye movements were oblique, conju­gate saccades. Intersaccadic intervals between some to- and- fro saccades varied between 60 and 160 ms. Many saccades in opposing directions had no intersaccadic intervals. The subject could pro­duce the saccadic oscillations continuously for sev­eral minutes. Small- amplitude ( 1- 3 degrees), high- frequency ( 25 Hz) primarily horizontal oscillations were present during some intersaccadic intervals and were occasionally superimposed on large- am­plitude saccades. The oscillations sometimes had smaller, conjugate, vertical components. They rep­resented voluntary " nystagmus" in this subject. The peak velocity- amplitude relationships of hori­zontal ( Fig. 5) and vertical ( Fig. 6) eye movements were characteristic of saccades. Figure 7 is a graph of horizontal and vertical eye position during a segment of the saccadic oscillations. The saccades usually had both horizontal and vertical compo­nents. Pairs of opposing saccades tended to have closed- loop trajectories, that is, the second saccade carried to eyes to a point near the starting point of first saccade. The saccades had curvilinear or cres-centic shapes. A few oscillations of voluntary " nystagmus" ( vertical component) were superim­posed on a large, horizontal saccade. COMMENT Saccades are abrupt, rapid eye movements of short duration that comprise refixations, fast com­ponents of physiologic and pathologic nystagmus, and a class of fixation instability, called saccadic oscillations. They have high peak velocities, up to 800 degrees/ s for large saccades, which are pro­duced by forceful contraction of extraocular muscle fibers. The abrupt increase in the firing rate of the ocular motor neurons, causing the contraction, is generated by stimulation of the motoneurons from the sudden discharge of excitatory burst neurons. The burst neurons can fire at rates up to 800 spikes per second during large saccades in monkeys. Burst neurons are located in the paramedian pon­tine reticular formation ( PPRF) for horizontal sac­cades and in the rostral interstitial nucleus of the medial longitudinal fasciculus ( riMLF) for vertical saccades. Burst neurons are tonically inhibited by another group of cells in the pontine tegmentum ( nucleus raphe interpositus) called omnidirectional pause neurons. The pause neurons inhibit burst neurons that generate saccades in all directions. During fixation of a stationary target, tonic firing of the pause neurons inhibits firing of excitatory burst neurons, preventing the occurrence of extra­neous saccades. To generate a saccade, an inhibi- 15^ 15" FIG. 3. Voluntary saccadic oscillations in Subject 2. Electro- oculographic record­ings on curvilinear graph paper produce artifactual curvature of large pen deflec­tions. Top: Vertical movements of the left eye. Middle: Horizontal movements of the right eye. Bottom: Horizontal movements of the left eye. The subject begins oscilla­tions at arrows. Most oblique, to- and- fro saccades have no intersaccadic intervals and resemble opsoclonus. 1 sec / Neuro- Ophthalmot, Vol. 14, No. 2, 1994 VOLUNTARY SACCADIC OSCILLATIONS 99 15" 15" FIG. 4. Voluntary saccadic oscillations in Subject 2. Coil recordings on rectilinear graph paper. Top: One- second intervals. Middle: Vertical movements of the right eye. Bottom Two Lines: Horizontal move­ments of the right eye, and horizontal movements of the left eye. Most couplets and triplets of oblique, to- and- fro sac-cades have no intersaccadic intervals. White arrows show low- amplitude, hori­zontal oscillations representing voluntary " nystagmus." Black arrows show artifact when pen deflection saturated recording channel. 1 sec tory signal decreases the tonic firing of pause neu­rons, removing inhibition of the burst neurons. The sources of excitatory and inhibitory input to the pause neurons during fixation and saccades is not known. The pause neurons receive projections from the frontal eye fields, supplementary eye fields, superior colliculus, and mesencephalic re­ticular formation. The anatomy and physiology of these pathways were summarized recently ( 1,7,8). Zee and Robinson ( 9) described a model for sac­cadic pulse generation that could be modified to produce saccadic oscillations. Their neural circuit drives the eyes to a calculated position in the or­bits, whereas other models produce a prepro­grammed pulse of specified intensity and duration based on the position of the new target. The firing of burst neurons in their model is inherently un­stable because of a small delay in the eye position feedback loop. By decreasing the activity of pause neurons, they simulated a saccadic oscillation that resembled voluntary " nystagmus," which they had recorded in a normal subject. Increasing the delay in the eye position feedback loop and de­creasing pause neuron activity produced a saccadic o I 400 » _ 300 > H O 200 o - J UJ > •• • . ft * I u o 600 - 500 - 400- 300 - 200- 100- 0 - • • . * • • • • • t • • • . * • •-•••;• • • • •„ •• i : - i - * i - * i - • - i - • i - 40 - 20 20 40 60 AMPLITUDE ( cleg ) FIG. 5. Horizontal peak velocity vs amplitude of vol­untary saccadic oscillations in Subject 2. Positive am­plitudes are to the right, negative to the left. The ve­locity- amplitude relationships are similar to those of horizontal, refixation saccades in this subject. AMPLITUDE ( deg ) FIG. 6. Vertical peak velocity vs amplitude of volun­tary saccadic oscillations in Subject 2. Positive ampli­tudes are upward, negative downward. The velocity-amplitude relationships are similar to those of vertical, refixation saccades in this subject. / Neuro- Ophthalmol, Vol. 14, No. 2, 1994 200 R. D. YEE ET Ah. „ " 10.0 " O 2 0 .0 Z 1 O 2 J A. in s^ / \ > g o ooo ( I \ p-\ 3 a \ _--^- y w l_^-_^-• f/^^ ~~^~ >- 1/ UJ yf - I - 20.0 r < u t- H I - tr > - < o . o J 1 1 1 1 1 > 1 ' - 40.0 - 20.0 .000 20.0 HORIZONTAL EYE POSITION ( deg ) FIG. 7. Graph of vertical vs horizontal eye positions during voluntary saccadic oscillations in Subject 2. Positive deflections are to the right or upward, nega­tive deflections to the left or downward. Numbers in­dicate beginnings of to- and- fro, saccadic couplets. Note couplets 1 and 3 begin and end near same posi­tions [ closed loop) and small, vertical oscillations dur­ing couplet 2 shown by arrow ( voluntary " nystag­mus"). oscillation that was similar to ocular flutter re­corded in a patient with encephalitis. Presumably, similar changes in the neural circuit can be pro­duced intentionally in voluntary " nystagmus" and result from damage from disorders causing ocular flutter and opsoclonus. A variety of disorders cause ocular flutter and opsoclonus ( 3). In children they include prenatal and perinatal central nervous system injury ( pre­sumably due to anoxia), paraneoplastic effects of neuroblastoma, and viral encephalitis. In adults they are postviral encephalitis, paraneoplastic ef­fects of neoplasms, infarction, trauma, multiple sclerosis, intracranial tumors, and toxins. How­ever, attempts to identify consistent anatomic loci of damage in opsoclonus have been unsuccessful. Nevertheless, most patients have signs and symp­toms consistent with damage to the brainstem and/ or cerebellum, suggesting that abnormalities in the brainstem system that generates saccades are present. Examples of normal subjects generating back- to-back saccades have been reported previously. About 5% of normal subjects can produce volun­tary nystagmus ( 2,10- 12). Presumably they can voluntarily inhibit pause cells. Hain and colleagues ( 13) described a patient with a progressive neuro­degenerative disorder who generated large dy­namic overshoots at the end of horizontal saccades when he blinked. A dynamic overshoot is a hyper- / Neuro- Ophthalmol, Vol. 14, No. 2, 1994 metric saccade, followed immediately by a saccade in the opposite direction with no intersaccadic in­terval between the saccades. The patient also pro­duced short bursts of horizontal ocular flutter when he blinked. One of the authors of that report and of our report ( D. S. Z.) can generate small am­plitude dynamic overshoots and ocular flutter with blinks. They hypothesized that the blink might somehow inhibit pause cells or directly stimulate burst cells. Examination of recordings from our two subjects showed no correlation between their saccadic oscillations and blinking. Our subjects' eye movements had many features characteristic of voluntary " nystagmus," ocular flutter, and opsoclonus. Reported cases of volun­tary " nystagmus" demonstrated frequencies of 8 to 23 Hz, amplitudes varying between 3 and 8 de­grees and maximum durations of 20- 30 s2. Both of our subjects produced short periods of 1 degree, 25 Hz, pendular oscillations for short periods ( Figs. 1 and 4). Digre ( 3) reviewed opsoclonus in adults and summarized it in children. She defined opso­clonus as " involuntary, arrhythmic, chaotic, ir­regular eye movements, that are predominantly horizontal but with vertical and rotating jerks at a frequency of six to 12 per second." The eye move­ments in our subjects fit this definition, except for the issue of volition. Recordings of ocular flutter and opsoclonus have shown single to- and- fro os­cillations and bursts of oscillations with frequen­cies of 5 to 13 Hz ( 9,14- 16). The large- amplitude saccadic oscillations in our subjects had frequen­cies of 2 to 8 Hz. Gresty and colleagues ( 16) de­scribed opsoclonus in a patient with multiple scle­rosis that had vertical and horizontal oscillations, with closed- loop, crescentic waveforms. The mag­netic search coil recordings of Subject 2 ( Fig. 7) showed similar features. We believe that our first subject produced his saccadic oscillations voluntarily. The absence of other objective, abnormal findings on neurologic examination, the normal visual acuity, the absence of oscillopsia, and the apparent lack of effort in trying to fixate the laser target are inconsistent with findings in patients with ocular flutter and opsoclonus, whom we have examined, and in pa­tients described in the literature. Head trauma has caused opsoclonus ( 3,17,18). However, the trauma was severe and other signs indicative of neurologic damage were present. Our second subject voluntarily produced his saccadic oscillations. They closely resembled those of the first subject and had many of the character­istics of opsoclonus. These included lack of an in­tersaccadic interval between many to- and- fro sac- VOLUNTARY SACCADIC OSCILLATIONS 101 cades, varying directions, high frequency, large amplitude, and curvilinear, crescentic waveforms. He could generate saccadic oscillations on his first attempt after observing the first subject's eye movements. He used oscillopsia and continuously blurred vision as feedback. When a normal subject is instructed to track a jumping target or look back and forth between two stationary targets as rapidly as possible, he or she produces saccades with a latency ( delay between target movement and eye movement) or an intersaccadic interval of about 200 ms. During these tasks, the subject's feedback about the accuracy of the saccades is the clarity of the targets. Our second subject tried to avoid see­ing any object clearly. Voluntary " nystagmus," in addition to large amplitude oscillations, was pro­duced by this strategy. Interestingly, the first sub­ject also generated voluntary " nystagmus." We believe that other normal subjects can learn to pro­duce similar, large amplitude, saccadic oscilla­tions. Eye movement records from our second subject occasionally showed small amplitude, vertical os­cillations superimposed on large horizontal sac­cades ( Fig. 7). Zee and coworkers ( 19) recently de­scribed eye movements in normal subjects who were required to track target movements calling for combinations of saccades and vergence. They observed small- amplitude, saccadic oscillations su­perimposed on large saccades during pure sac­cades and saccades combined with vergence eye movements. The oscillations were orthogonal to the direction of the large saccade, that is, vertical oscillations with horizontal saccades and horizon­tal oscillations with vertical saccades, and had am­plitudes and frequencies characteristic of " volun­tary nystagmus." They hypothesized that inhibi­tion of omnidirectional pause neurons during a refixation saccade in one direction, for example, horizontal, produced disinhibition of inherently unstable burst neurons that produce saccades in the orthogonal direction, for example, vertical. This would produce saccadic oscillations in the or­thogonal direction. Ocular flutter and opsoclonus are dramatic signs which indicate that damage to the brainstem and/ or cerebellum has occurred. However, when sac­cadic oscillations similar to ocular flutter and op­soclonus are found in patients who do not have other neurologic signs and symptoms, the possi­bility of a nonorganic cause should be considered. Nevertheless, a thorough evaluation, including a search for systemic disorders, for example, viral infections and occult neoplasms, neurologic con­sultation, and neuroradiologic examination, for ex­ample, magnetic resonance imaging of the head and posterior fossa, is needed. REFERENCES 1. Leigh RJ, Zee DS. Diagnosis of central disorders of ocular motility. In: The neurology of eye movements, 2nd ed. Phila­delphia: FA Davis; 1991: 378- 407. 2. Aschoff JC, Becker W, Rettelbach R. Voluntary nystagmus in five generations. / Neurol Neurosurg Psychiatry 1976; 39: 300- 4. 3. Digre KB. Opsoclonus in adults: report of three cases and review of the literature. Arch Neurol 1986; 43: 1165- 75. 4. Baloh RW, Langhofer L, Honrubia V, Yee RD. On- line anal­ysis of eye movements using a digital computer. Aviat Space Environ Med 1980; 51: 563- 7. 5. Robinson DA. A method of measuring eye movements us­ing a scleral search coil in a magnetic field. 7EEE Trans Bio- Med Electron 1963; 10: 137- 45. 6. Collewijn H, Van Der Mark F, Jansen TC. Precise recording of human eye movements. Vision Res 1975; 15: 447- 50. 7. Becker W. Metrics. In: Wurtz RH, Goldberg ME, eds. The neurobiology of saccadic eye movements. Amsterdam: Elsevier; 1989: 13- 67. 8. Buttner U, Buttner- Ennever JA. Present concepts of oculo­motor organization. In: Buttner- Ennever JA, ed. Neuroanat­omy of the oculomotor system. Amsterdam: Elsevier; 1988: 3- 32. 9. Zee DS, Robinson DA. A hypothetical explanation of sac­cadic oscillations. Ann Neurol 1979; 5: 405- 14. 10. Jarrett A, Emery JN, Coats AC, et al. Voluntary nystagmus. Ann Ophthalmol 1977; 9: 853- 9. 11. Krohel G, Friffin JF. Voluntary vertical nystagmus. Neurol­ogy 1979; 29: 115^ 4. 12. Blair CJ, Goldberg MF, von Noorden GK. Voluntary nystagmus: electro- oculographic findings in four cases. Arch Ophthalmol 1967; 77: 349- 54. 13. Hain TC, Zee DS, Mordes M. Blink- induced saccadic oscil­lations. Ann Neurol 1986; 19: 299- 300. 14. Vignaendra V, Lim CL. Electro- oculographic analysis of op­soclonus: its relationship to saccadic and non- saccadic eye movements. Neurology 1977; 27: 1129- 33. 15. Bergenius J. Saccade abnormalities in patients with ocular flutter. Acta Otolaryngol 1986; 102: 228- 33. 16. Gresty MA, Findley LJ, Wade P. Mechanism of rotatory eye movements in opsoclonus. Br ] Ophthalmol 1980; 64: 923- 5. 17. Robins A, Vallat M, Tapie P, et al. Opsoclonus et troubles neurologiques accocies. Arch Ophtalmol 1976; 36: 645- 56. 18. Turazzi S, Alexandre A, Bricolo A, et al. Opsoclonus and palatal myoclonus during prolonged post- traumatic coma. Eur Neurol 1977; 15: 257- 63. 19. Zee DS, Fitzgibbon EJ, Optican LM. Saccade- vergence in­teractions in humans. / Neurophysiol 1992; 68: 1624- 41. / Neuro- Ophthalmol, Vol. 14, No. 2, 1994