Exp Brain Res (1992) 90:589-598
9 Springer-Verlag 1992
Interaction of active and passive slow eye movement systems Ralph Worfolk and Graham R. Barnes MRC Human Movement and Balance Unit, Institute of Neurology,23 Queen Square, London, WC1N 3BG, UK Received November 11, 1991/Accepted April 8, 1992
Summary. Independent target and background motions have been used to generate conflicting activity within the pursuit and optokinetic systems. Subjects were required to pursue a small target against a structured background which moved independently. Selective enhancement of the response to the target generated high-gain active pursuit which dominated the eye movements. Passive eye movements induced during relative target and background motion are not normally directly quantifiable due to their low gain. By reducing the gain of the active pursuit optokinetically induced eye movements were enhanced and quantified. Three techniques are described for degrading active pursuit: tachistoscopic, eccentric and pseudorandom methods of target presentation. Our results demonstrate the synchronous input of active and passive eye movement drives to the oculomotor system and illustrate their interaction. Key words: Ocular pursuit
OKN
Introduction It has long been recognised that nystagmic eye movements may arise as contours sweep over the retina, even when the subject remains passive to the stimulus both behaviourally and perceptually. Under these conditions low gain slow eye movements with little phase lag are generated, with fast phases in the opposite direction (Cheng and Outerbridge 1975; Barnes and Hill 1984, Pola and Wyatt 1985). The involuntary slow phases induced in this manner may be considered as genuine, wholly reflexive, optokinetic responses. In addition to these passive responses humans are able deliberately to attend to any object within the visual field, and this introduces an active (and usually dominant) component to the eye movement response. A Correspondence to. R. Worfolk
stimulus does not have to fall on the central retina in order to provoke active smooth pursuit responses; smooth eye movement gain to a moving retinal image dramatically increases when subjects actively attend to image movement presented either centrally or peripherally (Cheng and Outerbridge 1975; Dubois and Collewijn 1979, Barnes and Hill 1984). It is possible to attend to any selected detail within the visual field and thereby introduce active fixation or pursuit into our response. Selective enhancement of a response can be demonstrated using conflicting motion stimuli. For example, Murasugi et al. (1986) showed that when stationary contours were superimposed on an optokinetic field then gains were dependent on whether subjects attended to the moving or stationary contours. During pursuit over a structured background active pursuit is in conflict with passive optokinetic drives which respond to the global retinal image slip. Consequently, in order to avoid disruption from the image slip of background contours, the pursuit response to signals derived from the target motion must be enhanced relative to the optokinetic eye movement drives. Kowler and colleagues (1984) illustrated the effectiveness of selective enhancement using two superimposed full field, fine structured, high density, random dot displays. One display was stationary whilst the other drifted at low velocity (70.2 minarc/s). Their subjects were able to fixate or pursue whichever field they chose with virtually no detrimental effects from the conflicting motion of the non-selected background. The capacity to selectively attend to, and pursue, specific visual stimuli is self-evident and not disputed however there is some divergence in the literature as to the magnitude of pursuit degradation caused by relative movement between a target and its background. Although early observations suggested that pursuit over a stationary structured background may not be degraded (Hood 1975; Guedry et al. 1979), more recent, quantitative, studies have recorded small deteriorations in pursuit gain attributable to the background structure (Yee et al. 1983, Collewijn and Tamminga 1984; Kowler et al. 1984; Barnes and Crombie 1985; Kaufman and Abel 1986). In general, structured backgrounds reduce smooth pursuit gain by about 10%.
590 In h u m a n s the ability to selectively e n h a n c e a p u r s u i t r e s p o n s e to a specific target a p p e a r s to be sufficiently evolved to p r a c t i c a l l y o v e r r i d e passive drives a s s o c i a t e d with b a c k g r o u n d detail. If the active p u r s u i t gain were to be c o m p r o m i s e d then one w o u l d p r e d i c t t h a t the passive effects of conflicting retinal i m a g e m o t i o n w o u l d b e c o m e m o r e p r o m i n e n t . T a r g e t p r o p e r t i e s can dictate the r o b u s t n e s s of p u r s u i t a n d hence the i n t e r a c t i o n b e t w e e n active a n d passive eye m o v e m e n t s . Barnes a n d C r o m b i e (1985) i n s t r u c t e d their subjects to t r a c k either a single central t a r g e t or an i m a g i n a r y p o i n t m i d w a y between two targets which were s e p a r a t e d vertically b y 7 deg. S m o o t h p u r s u i t gain over a h o m o g e n e o u s b a c k g r o u n d was high in b o t h target conditions. I n t r o d u c i n g a s t r u c t u r e d b a c k g r o u n d , however, r a d i c a l l y r e d u c e d p e r f o r m a n c e to the vertically s e p a r a t e d targets; similar results were o b t a i n e d by Collewijn a n d T a m m i n g a (1986). T h u s it a p p e a r s t h a t the effects of passive drives on slow eye m o v e m e n t c o n t r o l m a y be m o r e in evidence when active p u r s u i t has a r e d u c e d i n t e r n a l f e e d b a c k gain. This effect m a y e x p l a i n the r e p o r t s t h a t p u r s u i t w e a k e n e d by cerebellar lesions, P a r k i n s o n ' s disease o r p a r i e t a l lesions m a y be m a r k e d l y d i s r u p t e d b y a visible b a c k g r o u n d ( H o o d 1975; B a l o h et al. 1982; Yee et al. 1983; H o o d a n d W a n i e w s k i 1984). In a d d i t i o n to q u a n t i f y i n g the effects of the b a c k g r o u n d on pursuit, it is possible to e x a m i n e the reciprocal effect of the p u r s u i t task on the o p t o k i n e t i c drive from the b a c k g r o u n d . If the b a c k g r o u n d itself oscillates at a different frequency to the p u r s u i t target, c o r r e l a t i o n techniques can be used to identify the c o m p o n e n t s of the s m o o t h eye m o v e m e n t a t t r i b u t a b l e to the target a n d the b a c k g r o u n d (Collewijn a n d T a m m i n g a 1986). H i t h e r t o no research has e x a m i n e d h o w passive eye m o v e m e n t s v a r y with the active p u r s u i t gain. It was the i n t e n t i o n of these e x p e r i m e n t s to progressively d e g r a d e the gain of active p u r s u i t a n d thereb y d e t e r m i n e the c o n t r i b u t i o n a n d i n t e r p l a y of active a n d passive drives in the final slow eye m o v e m e n t s . T h r e e different techniques were used to give a c o n t r o l l e d interference with pursuit: t a c h i s t o s c o p i c t a r g e t p r e s e n t a t i o n , eccentric target p r e s e n t a t i o n a n d p s e u d o - r a n d o m target m o t i o n . W e d e m o n s t r a t e t h a t passively i n d u c e d eye m o v e m e n t s a s s o c i a t e d with the b a c k g r o u n d are present in the m o t o r o u t p u t a n d t h a t their m a g n i t u d e is inversely related to the s i m u l t a n e o u s gain of the active pursuit.
converter. Saccades were removed from the raw eye movement data using a method similar to that described by Barnes (1982). The smooth pursuit gain and phase were computed at each of the target and background frequencies by independently determining the best (least squares error) fit of these frequencies to the residual slow eye movements and the stimulus velocities. Each experiment was performed on eight subjects, but the subject populations were not identical.
Tachistoscopic presentation The target was projected onto the screen via a mirror galvanometer; an electro-mechanical shutter system mounted in front of the projector (Gerbrands G1166) allowed precise tachistoscopic control of the target. The target was exposed in a series of 20 ms pulses and intervals between pulse onset were set to 20 (continuous presentation), 80, 160, 240, 320, 400, 640 and 960 ms. The stimulus presentation order was randomised across subjects. Two trials were run at each pulse interval, one with a structured background, and one with the background absent. A computer-generated graphics image was projected onto the screen to provide a structured background consisting of 8 cycles of a 0.25 cpd square wave red and black grating (luminance of light and dark bars was 1.4 and 0.02 Cd/m2). The grating subtended 32 x 9.2~ and had a central horizontal 2.2~ band blanked out along its length, within which the target moved (Fig. la). The structured background, when present, was continuously visible and moved sinusoidally in the horizontal plane at 0.66 Hz, peak velocity 8~ The target moved horizontally with a sinusoidal waveform at 0.2 Hz, peak velocity 8~ for 35 s, data were digitised at 50 Hz.
Eccentric targets In this experiment the frequencies and peak velocities of the target and background motions were set to 0.45 Hz, 20~ and 0.2 Hz, 8~ respectively. The background was identical in spatial parameters to the previous experiment. The white 0.8 ~ targets were generated by computer graphics and either a single central target or pairs of targets separated horizontally by 10, 20or 30~ were presented to each subject (Fig. lb). Subjects were instructed to follow the single target or, if separated targets were present, to track an imaginary point midway between the targets as smoothly as possible. Each target condition was presented under background present and absent conditions in randomised order. Eight subjects participated in this experiment. Eye movements were digitised at 100 Hz.
Pseudo-random target movement Methods Eye movements were recorded using an IR limbal reflection technique (Skalar Iris recorders, linear range 4- 20 ~ sensitivity 1/6 degree). The recording sensors were mounted in a tightly fitting helmet which was clamped to the subject's head and chair. A chin rest was also utilised to prevent head movements. The pursuit target, cross hairs within a 1~ circle, was projected onto a white tangent screen subtending 90 by 54 degrees, sited 1.5 m from the subject. A Barco 400 video projector superimposed computer-generated graphics images on to the screen. The room was light-tight and in the absence of the background stimulus other contours were not visible. Subjects were instructed not to be distracted by the background but to pursue the target as smoothly and as accurately as possible. Target movements and data collection were controlled by a 12 bit D/A & A/D
The third type of stimulus used to induce a reduction in pursuit gain was a pseudo-random target motion composed of harmonically unrelated sine waves (Stark et al. 1962; Collewijn and Tamminga 1984; Barnes et al. 1987). Four component frequencies of 0.11, 0.24, 0.37 and 1.52 Hz were combined. A peak velocity of 8~ was assigned to the three lower frequencies whilst the peak velocity of the highest frequency was set to 0, 8 or 16~ in different trials. Thus, in this experiment, the ratio between the peak velocity of the highest frequency component and the lower frequency components (velocity ratio, 'VR') was set to 0, 1 or 2. Each of the three waveforms was presented under background present and background absent conditions, in random order, to 8 subjects. A green and black 0.25 cpd vertical square wave grating subtending 40 x 6~ was generated by the computer to act as the background; this moved sinusoidally at 0.64Hz, 8~ (Fig. lc). Eye movements were digitised at 66.7 Hz.
591 A
I I I I I I I I IIIIIIII
Background 0.66 Hz Peak Velocity 8~ Target 0.2 Hz Peak Velocity 8~ PD 20 msec IPI 20-960 msec
B
IIIIIIII IIIIIIII 0
C
IIII
0
Background 0.2 Hz Peak Velocity 8 ~
Fig. 1A-C. The backgrounds and conditions used in the three experiTarget 0.45 Hz o ments. No other contours were visible Peak Velocity 20 /s Separation 0,10,20 or 30 ~ other than the detail shown. IPI = inter-pulse interval, PD = pulse-duration. A Background and target used during tachistoscopic target presentation in which inter-pulse interval was varied. B Background and targets used during pursuit of the midpoint of eccentric targets. C Background and tarPeak Velocity 8 % get used during pseudo-random target Target motion. The target intensity was greater than that of the background so 0.11Hz, 8~ that the target was always clearly vis0.24Hz, 8 ~ ible above the background structure 0.37Hz, 8 ~
] ] ii ii ono4"z
Results
Tachistoscopic presentation As the inter-pulse interval between target presentations was increased the quality of the visual feedback fell and this was accompanied by a fall in active smooth pursuit gain as described by Barnes and Asselman (1992) (Fig. 2b, c). In the absence of the background, the velocity gain of the active pursuit fell from 0.94 to 0.30 as the condition changed from continuous target presentation to one 20 ms pulse in every 960 ms (Fig. 3a). When the background was present the active gain to the target was significantly reduced at all inter-pulse intervals tested (paired t-test, P < 0.05), falling progressively from 0.84 to 0.19 as the inter-pulse interval rose. The reduction of active smooth pursuit gain caused by the background reached a maximum of 48% at the inter-pulse interval of 640 ms. The eye displacement gain (composite smooth pursuit and saccadic movements) remained close to unity in all conditions, confirming that the subjects were pursuing the target. After averaging over the inter-pulse interval conditions the mean (_+SD) displacement gain at the target frequency was 1.01 _+0.08 in the background absent condition and 0.96__ 0.08 in the background present condition. Thus the background produced a slight, but significant (F(1,7) =27, P < 0.01), reduction in the displacement gain. The velocity gain and phase of eye movements passively induced by the background at 0.66 Hz were also quantified (Fig. 3b). As the active gain fell, the passive gain
1.52Hz, 0,8or 16~
rose from a negligible value (0.08) to 0.47. At intermediate inter-pulse intervals the individual frequency components associated with the target and background were visible within the slow eye movement traces (Fig. 2d). Another notable effect of the background was a change in phase associated with the active pursuit gain. In the background absent condition phase lag increased with increasing inter-pulse interval (Fig. 3a). However, with the background present the trend of phase lag was not observed; instead the mean phase remained close to zero. Barnes and Asselman (1992) recorded small phase leads at the lowest frequency component during tachistoscopic presentation of a mixed frequency stimulus. It seems that when more than one frequency of retinal image motion is present that the phase of the lowest frequency may be advanced, possibly as a result of the filtering characteristics of the pursuit system (Barnes and Ruddock 1989). The phase of the passively induced eye movements was close to zero except at the two shortest inter-pulse intervals when the mean phase was near to 180 degrees (Fig. 3b). However the diminutive gain under the latter conditions renders this phase measurement unreliable.
Eccentric targets Figure 4a shows the gain at the target frequency as subjects pursued the eccentric targets in background absent and present conditions. In the background absent condition the smooth pursuit gain fell from 0.97 to 0.89 as
592
A Target Position Background Position
12 deg
~
12 deg 5s
B Eye Position Target Strobe Eye Velocity
12 deg on , off
9J
~
16deg/s Background Absent
5s
C Eye Position Target Strobe
/
~
12 deg
IlllllllllllllllllllllllllllllllllllllllllllllllllllllflJ off on
Eye Velocity
16 deg/s BackgroundAbsent 5s
D Eye Position Target Strobe
12deg ~
~
~
~
~
~
on off
16 deg/s
Eye Velocity Background Velocity ~
#
5s V V
~
v
~
A
A A A A A f ~ %
the central target was replaced by the _+ 15 ~ targets (paired t-test P=0.005). The addition of the structured background severely degraded the pursuit so that the smooth pursuit gain was significantly lower under all target conditions and fell progressively as the target separation increased. Active pursuit showed a very slight phase lag under both conditions which was slightly greater when the background was present. Eye displacement gains remained close to unity and were not dependent on the background condition (F(1,7)=2.46, P=0.16). Passive slow eye movements at the background frequency were dependent on the separation of the targets, showing a progressive rise in gain from 0.07 to 0.22 (Fig. 4b). As the target separation increased the mean phase lag of passive smooth eye movements fell from - 1 0 7 ~ to - 2 7 ~
I
]16 deg/-~
Fig. 2A-D. Effects of tachistoscopic target presentation and a structured background on pursuit. Eye position traces show the raw eye movement data collected during the experiment. Saccadic components have been removed from the eye velocity traces. A Target (0.2 Hz, peak velocity 8~ and background (0.66 Hz peak velocity 8~ displacements. B Continuous target presentation at 0.2 Hz, 8~ background absent. C Tachistoscopic target presentation, background absent, inter-pulse interval = 640 ms, pulse duration=20 ms. Note the broken pursuit and reduced smooth pursuit gain. D Tachistoscopic target presentation, background present, inter-pulse interval = 320 ms, pulse duration = 20 ms. Note components of smooth eye velocity at both the target and background frequencies. Components present at the background frequency are in phase with the background velocity
Pseudo-random target movement
Increasing the ratio between the peak velocity of the highest frequency component and that of the low frequency components of the stimulus decreased the smooth eye movement gain at the low frequencies in the manner described by Barnes and Ruddock (1989). This effect is illustrated in Fig. 5 which shows the gain at each of the frequency components at the different velocity ratios (VR). As VR assumed values of 0, 1 and 2 the mean gain averaged over the three low frequencies was 0.95, 0.74, and 0.64 respectively. Analysis of variance confirmed that the gain at the three low frequencies was dependent on the velocity ratio (F(2,14)=84, P
9 Springer-Verlag 1992
Interaction of active and passive slow eye movement systems Ralph Worfolk and Graham R. Barnes MRC Human Movement and Balance Unit, Institute of Neurology,23 Queen Square, London, WC1N 3BG, UK Received November 11, 1991/Accepted April 8, 1992
Summary. Independent target and background motions have been used to generate conflicting activity within the pursuit and optokinetic systems. Subjects were required to pursue a small target against a structured background which moved independently. Selective enhancement of the response to the target generated high-gain active pursuit which dominated the eye movements. Passive eye movements induced during relative target and background motion are not normally directly quantifiable due to their low gain. By reducing the gain of the active pursuit optokinetically induced eye movements were enhanced and quantified. Three techniques are described for degrading active pursuit: tachistoscopic, eccentric and pseudorandom methods of target presentation. Our results demonstrate the synchronous input of active and passive eye movement drives to the oculomotor system and illustrate their interaction. Key words: Ocular pursuit
OKN
Introduction It has long been recognised that nystagmic eye movements may arise as contours sweep over the retina, even when the subject remains passive to the stimulus both behaviourally and perceptually. Under these conditions low gain slow eye movements with little phase lag are generated, with fast phases in the opposite direction (Cheng and Outerbridge 1975; Barnes and Hill 1984, Pola and Wyatt 1985). The involuntary slow phases induced in this manner may be considered as genuine, wholly reflexive, optokinetic responses. In addition to these passive responses humans are able deliberately to attend to any object within the visual field, and this introduces an active (and usually dominant) component to the eye movement response. A Correspondence to. R. Worfolk
stimulus does not have to fall on the central retina in order to provoke active smooth pursuit responses; smooth eye movement gain to a moving retinal image dramatically increases when subjects actively attend to image movement presented either centrally or peripherally (Cheng and Outerbridge 1975; Dubois and Collewijn 1979, Barnes and Hill 1984). It is possible to attend to any selected detail within the visual field and thereby introduce active fixation or pursuit into our response. Selective enhancement of a response can be demonstrated using conflicting motion stimuli. For example, Murasugi et al. (1986) showed that when stationary contours were superimposed on an optokinetic field then gains were dependent on whether subjects attended to the moving or stationary contours. During pursuit over a structured background active pursuit is in conflict with passive optokinetic drives which respond to the global retinal image slip. Consequently, in order to avoid disruption from the image slip of background contours, the pursuit response to signals derived from the target motion must be enhanced relative to the optokinetic eye movement drives. Kowler and colleagues (1984) illustrated the effectiveness of selective enhancement using two superimposed full field, fine structured, high density, random dot displays. One display was stationary whilst the other drifted at low velocity (70.2 minarc/s). Their subjects were able to fixate or pursue whichever field they chose with virtually no detrimental effects from the conflicting motion of the non-selected background. The capacity to selectively attend to, and pursue, specific visual stimuli is self-evident and not disputed however there is some divergence in the literature as to the magnitude of pursuit degradation caused by relative movement between a target and its background. Although early observations suggested that pursuit over a stationary structured background may not be degraded (Hood 1975; Guedry et al. 1979), more recent, quantitative, studies have recorded small deteriorations in pursuit gain attributable to the background structure (Yee et al. 1983, Collewijn and Tamminga 1984; Kowler et al. 1984; Barnes and Crombie 1985; Kaufman and Abel 1986). In general, structured backgrounds reduce smooth pursuit gain by about 10%.
590 In h u m a n s the ability to selectively e n h a n c e a p u r s u i t r e s p o n s e to a specific target a p p e a r s to be sufficiently evolved to p r a c t i c a l l y o v e r r i d e passive drives a s s o c i a t e d with b a c k g r o u n d detail. If the active p u r s u i t gain were to be c o m p r o m i s e d then one w o u l d p r e d i c t t h a t the passive effects of conflicting retinal i m a g e m o t i o n w o u l d b e c o m e m o r e p r o m i n e n t . T a r g e t p r o p e r t i e s can dictate the r o b u s t n e s s of p u r s u i t a n d hence the i n t e r a c t i o n b e t w e e n active a n d passive eye m o v e m e n t s . Barnes a n d C r o m b i e (1985) i n s t r u c t e d their subjects to t r a c k either a single central t a r g e t or an i m a g i n a r y p o i n t m i d w a y between two targets which were s e p a r a t e d vertically b y 7 deg. S m o o t h p u r s u i t gain over a h o m o g e n e o u s b a c k g r o u n d was high in b o t h target conditions. I n t r o d u c i n g a s t r u c t u r e d b a c k g r o u n d , however, r a d i c a l l y r e d u c e d p e r f o r m a n c e to the vertically s e p a r a t e d targets; similar results were o b t a i n e d by Collewijn a n d T a m m i n g a (1986). T h u s it a p p e a r s t h a t the effects of passive drives on slow eye m o v e m e n t c o n t r o l m a y be m o r e in evidence when active p u r s u i t has a r e d u c e d i n t e r n a l f e e d b a c k gain. This effect m a y e x p l a i n the r e p o r t s t h a t p u r s u i t w e a k e n e d by cerebellar lesions, P a r k i n s o n ' s disease o r p a r i e t a l lesions m a y be m a r k e d l y d i s r u p t e d b y a visible b a c k g r o u n d ( H o o d 1975; B a l o h et al. 1982; Yee et al. 1983; H o o d a n d W a n i e w s k i 1984). In a d d i t i o n to q u a n t i f y i n g the effects of the b a c k g r o u n d on pursuit, it is possible to e x a m i n e the reciprocal effect of the p u r s u i t task on the o p t o k i n e t i c drive from the b a c k g r o u n d . If the b a c k g r o u n d itself oscillates at a different frequency to the p u r s u i t target, c o r r e l a t i o n techniques can be used to identify the c o m p o n e n t s of the s m o o t h eye m o v e m e n t a t t r i b u t a b l e to the target a n d the b a c k g r o u n d (Collewijn a n d T a m m i n g a 1986). H i t h e r t o no research has e x a m i n e d h o w passive eye m o v e m e n t s v a r y with the active p u r s u i t gain. It was the i n t e n t i o n of these e x p e r i m e n t s to progressively d e g r a d e the gain of active p u r s u i t a n d thereb y d e t e r m i n e the c o n t r i b u t i o n a n d i n t e r p l a y of active a n d passive drives in the final slow eye m o v e m e n t s . T h r e e different techniques were used to give a c o n t r o l l e d interference with pursuit: t a c h i s t o s c o p i c t a r g e t p r e s e n t a t i o n , eccentric target p r e s e n t a t i o n a n d p s e u d o - r a n d o m target m o t i o n . W e d e m o n s t r a t e t h a t passively i n d u c e d eye m o v e m e n t s a s s o c i a t e d with the b a c k g r o u n d are present in the m o t o r o u t p u t a n d t h a t their m a g n i t u d e is inversely related to the s i m u l t a n e o u s gain of the active pursuit.
converter. Saccades were removed from the raw eye movement data using a method similar to that described by Barnes (1982). The smooth pursuit gain and phase were computed at each of the target and background frequencies by independently determining the best (least squares error) fit of these frequencies to the residual slow eye movements and the stimulus velocities. Each experiment was performed on eight subjects, but the subject populations were not identical.
Tachistoscopic presentation The target was projected onto the screen via a mirror galvanometer; an electro-mechanical shutter system mounted in front of the projector (Gerbrands G1166) allowed precise tachistoscopic control of the target. The target was exposed in a series of 20 ms pulses and intervals between pulse onset were set to 20 (continuous presentation), 80, 160, 240, 320, 400, 640 and 960 ms. The stimulus presentation order was randomised across subjects. Two trials were run at each pulse interval, one with a structured background, and one with the background absent. A computer-generated graphics image was projected onto the screen to provide a structured background consisting of 8 cycles of a 0.25 cpd square wave red and black grating (luminance of light and dark bars was 1.4 and 0.02 Cd/m2). The grating subtended 32 x 9.2~ and had a central horizontal 2.2~ band blanked out along its length, within which the target moved (Fig. la). The structured background, when present, was continuously visible and moved sinusoidally in the horizontal plane at 0.66 Hz, peak velocity 8~ The target moved horizontally with a sinusoidal waveform at 0.2 Hz, peak velocity 8~ for 35 s, data were digitised at 50 Hz.
Eccentric targets In this experiment the frequencies and peak velocities of the target and background motions were set to 0.45 Hz, 20~ and 0.2 Hz, 8~ respectively. The background was identical in spatial parameters to the previous experiment. The white 0.8 ~ targets were generated by computer graphics and either a single central target or pairs of targets separated horizontally by 10, 20or 30~ were presented to each subject (Fig. lb). Subjects were instructed to follow the single target or, if separated targets were present, to track an imaginary point midway between the targets as smoothly as possible. Each target condition was presented under background present and absent conditions in randomised order. Eight subjects participated in this experiment. Eye movements were digitised at 100 Hz.
Pseudo-random target movement Methods Eye movements were recorded using an IR limbal reflection technique (Skalar Iris recorders, linear range 4- 20 ~ sensitivity 1/6 degree). The recording sensors were mounted in a tightly fitting helmet which was clamped to the subject's head and chair. A chin rest was also utilised to prevent head movements. The pursuit target, cross hairs within a 1~ circle, was projected onto a white tangent screen subtending 90 by 54 degrees, sited 1.5 m from the subject. A Barco 400 video projector superimposed computer-generated graphics images on to the screen. The room was light-tight and in the absence of the background stimulus other contours were not visible. Subjects were instructed not to be distracted by the background but to pursue the target as smoothly and as accurately as possible. Target movements and data collection were controlled by a 12 bit D/A & A/D
The third type of stimulus used to induce a reduction in pursuit gain was a pseudo-random target motion composed of harmonically unrelated sine waves (Stark et al. 1962; Collewijn and Tamminga 1984; Barnes et al. 1987). Four component frequencies of 0.11, 0.24, 0.37 and 1.52 Hz were combined. A peak velocity of 8~ was assigned to the three lower frequencies whilst the peak velocity of the highest frequency was set to 0, 8 or 16~ in different trials. Thus, in this experiment, the ratio between the peak velocity of the highest frequency component and the lower frequency components (velocity ratio, 'VR') was set to 0, 1 or 2. Each of the three waveforms was presented under background present and background absent conditions, in random order, to 8 subjects. A green and black 0.25 cpd vertical square wave grating subtending 40 x 6~ was generated by the computer to act as the background; this moved sinusoidally at 0.64Hz, 8~ (Fig. lc). Eye movements were digitised at 66.7 Hz.
591 A
I I I I I I I I IIIIIIII
Background 0.66 Hz Peak Velocity 8~ Target 0.2 Hz Peak Velocity 8~ PD 20 msec IPI 20-960 msec
B
IIIIIIII IIIIIIII 0
C
IIII
0
Background 0.2 Hz Peak Velocity 8 ~
Fig. 1A-C. The backgrounds and conditions used in the three experiTarget 0.45 Hz o ments. No other contours were visible Peak Velocity 20 /s Separation 0,10,20 or 30 ~ other than the detail shown. IPI = inter-pulse interval, PD = pulse-duration. A Background and target used during tachistoscopic target presentation in which inter-pulse interval was varied. B Background and targets used during pursuit of the midpoint of eccentric targets. C Background and tarPeak Velocity 8 % get used during pseudo-random target Target motion. The target intensity was greater than that of the background so 0.11Hz, 8~ that the target was always clearly vis0.24Hz, 8 ~ ible above the background structure 0.37Hz, 8 ~
] ] ii ii ono4"z
Results
Tachistoscopic presentation As the inter-pulse interval between target presentations was increased the quality of the visual feedback fell and this was accompanied by a fall in active smooth pursuit gain as described by Barnes and Asselman (1992) (Fig. 2b, c). In the absence of the background, the velocity gain of the active pursuit fell from 0.94 to 0.30 as the condition changed from continuous target presentation to one 20 ms pulse in every 960 ms (Fig. 3a). When the background was present the active gain to the target was significantly reduced at all inter-pulse intervals tested (paired t-test, P < 0.05), falling progressively from 0.84 to 0.19 as the inter-pulse interval rose. The reduction of active smooth pursuit gain caused by the background reached a maximum of 48% at the inter-pulse interval of 640 ms. The eye displacement gain (composite smooth pursuit and saccadic movements) remained close to unity in all conditions, confirming that the subjects were pursuing the target. After averaging over the inter-pulse interval conditions the mean (_+SD) displacement gain at the target frequency was 1.01 _+0.08 in the background absent condition and 0.96__ 0.08 in the background present condition. Thus the background produced a slight, but significant (F(1,7) =27, P < 0.01), reduction in the displacement gain. The velocity gain and phase of eye movements passively induced by the background at 0.66 Hz were also quantified (Fig. 3b). As the active gain fell, the passive gain
1.52Hz, 0,8or 16~
rose from a negligible value (0.08) to 0.47. At intermediate inter-pulse intervals the individual frequency components associated with the target and background were visible within the slow eye movement traces (Fig. 2d). Another notable effect of the background was a change in phase associated with the active pursuit gain. In the background absent condition phase lag increased with increasing inter-pulse interval (Fig. 3a). However, with the background present the trend of phase lag was not observed; instead the mean phase remained close to zero. Barnes and Asselman (1992) recorded small phase leads at the lowest frequency component during tachistoscopic presentation of a mixed frequency stimulus. It seems that when more than one frequency of retinal image motion is present that the phase of the lowest frequency may be advanced, possibly as a result of the filtering characteristics of the pursuit system (Barnes and Ruddock 1989). The phase of the passively induced eye movements was close to zero except at the two shortest inter-pulse intervals when the mean phase was near to 180 degrees (Fig. 3b). However the diminutive gain under the latter conditions renders this phase measurement unreliable.
Eccentric targets Figure 4a shows the gain at the target frequency as subjects pursued the eccentric targets in background absent and present conditions. In the background absent condition the smooth pursuit gain fell from 0.97 to 0.89 as
592
A Target Position Background Position
12 deg
~
12 deg 5s
B Eye Position Target Strobe Eye Velocity
12 deg on , off
9J
~
16deg/s Background Absent
5s
C Eye Position Target Strobe
/
~
12 deg
IlllllllllllllllllllllllllllllllllllllllllllllllllllllflJ off on
Eye Velocity
16 deg/s BackgroundAbsent 5s
D Eye Position Target Strobe
12deg ~
~
~
~
~
~
on off
16 deg/s
Eye Velocity Background Velocity ~
#
5s V V
~
v
~
A
A A A A A f ~ %
the central target was replaced by the _+ 15 ~ targets (paired t-test P=0.005). The addition of the structured background severely degraded the pursuit so that the smooth pursuit gain was significantly lower under all target conditions and fell progressively as the target separation increased. Active pursuit showed a very slight phase lag under both conditions which was slightly greater when the background was present. Eye displacement gains remained close to unity and were not dependent on the background condition (F(1,7)=2.46, P=0.16). Passive slow eye movements at the background frequency were dependent on the separation of the targets, showing a progressive rise in gain from 0.07 to 0.22 (Fig. 4b). As the target separation increased the mean phase lag of passive smooth eye movements fell from - 1 0 7 ~ to - 2 7 ~
I
]16 deg/-~
Fig. 2A-D. Effects of tachistoscopic target presentation and a structured background on pursuit. Eye position traces show the raw eye movement data collected during the experiment. Saccadic components have been removed from the eye velocity traces. A Target (0.2 Hz, peak velocity 8~ and background (0.66 Hz peak velocity 8~ displacements. B Continuous target presentation at 0.2 Hz, 8~ background absent. C Tachistoscopic target presentation, background absent, inter-pulse interval = 640 ms, pulse duration=20 ms. Note the broken pursuit and reduced smooth pursuit gain. D Tachistoscopic target presentation, background present, inter-pulse interval = 320 ms, pulse duration = 20 ms. Note components of smooth eye velocity at both the target and background frequencies. Components present at the background frequency are in phase with the background velocity
Pseudo-random target movement
Increasing the ratio between the peak velocity of the highest frequency component and that of the low frequency components of the stimulus decreased the smooth eye movement gain at the low frequencies in the manner described by Barnes and Ruddock (1989). This effect is illustrated in Fig. 5 which shows the gain at each of the frequency components at the different velocity ratios (VR). As VR assumed values of 0, 1 and 2 the mean gain averaged over the three low frequencies was 0.95, 0.74, and 0.64 respectively. Analysis of variance confirmed that the gain at the three low frequencies was dependent on the velocity ratio (F(2,14)=84, P
