vjsi,ua ~rs.Vol.27,No.4,pp. 517-520, 1987 Printiin Chat Britain. All ti&ts rmcrval

0042-6989/87 S3.00 +O.OO Copyight

Q 1987 hrgamon

Journals Ltd

PREDICTIVE EYE SACCADES ARE DIFFERENT FROM VISUALLY TRIGGERED SACCADES A. M. BRONSTEIN' and C. KENNARD Neurological Department, The London Hospital, Whitechapel, London El IBB, England (Receioed 25 February 1986; in revisedform 16 September 1986) Abstraet-The metrics of eye saccades in response to random and predictable target motion were determined in 8 normal young subjects. The peak velocity of predictive saccades (PS) was significantly reduced when compared with randomly elicited saccades (RS) of equal amplitude. Saccadic gain (initial saccadic amplitude/target amplitude) was also decreased in PS. Saccadic inaccuracy was extremely frequent in PS and, in contrast to RS, it was independent of target amplitude displacement. These stimulus-dependent changes in saccadic characteristics may reflect activation of the saccadic pontine generators by different supranuclear pathways. Saccades

Eye movements

Although the precise mechanism by which saccades are programmed in the paramedian pontine reticular formation (PPRF) is still not certain, there is increasing evidence that the various supranuclear structures, such as the frontal eye field and superior colliculus activate the PPRF to generate saccades under different behavioural conditions (Zee, 1984). Accordingly saccadic parameters may be expected to vary to some degree when made under such different circumstances. In the present experiments we compare the accuracy and velocity of saccades elicted under random or predictive conditions. It was considered that, although a common, task for the ocular motor system, the motor strategy and sensory information available in those conditions differ considerably and, therefore, differences between random (RS) and predictive saccades (PS) of equal amplitude might be expected. The eye movements of eight normal subjects (mean 23 years range 18-37) with no experience in oculomotor experiments, were examined using the infra-red reflection technique (Stark et al., 1962) which had a band width of 300 Hz. The subjects sat in a dimly lit (5 subjects) or a completely dark (3 subjects) room. The target consisted of a white light spot (subtending 30’ of arc) projected onto a curved screen placed *Presentaddress: MRC Neuro-otology Unit, The National Hospital, Queen Square, London WClN 3BG, England. 517

150 cm from the subject’s head. To produce RS the target randomly moved in a step-like manner, with a random time interval of 0.6-1.2 set, between five different positions generating target displacements of 7.5, 1522.5 and 30 deg. To generate PS the same target amplitudes which were used in the random sequence were separately delivered at a regular time interval (0.9 set). In order to avoid any interference by fatigue or a diminished level of arousal experiments of short duration were chosen: two runs of RS, each of 3Osec duration, and then four runs of PS, one for each amplitude, of 20 set duration. Between each run the subjects were given rest periods of 30-40 sec. Before each run the subjects were informed as to whether the target would move randomly or regularly, and instructed to refixate the target as rapidly and accurately as possible without moving their head. Saccade latency, amplitude and velocity were measured by hand from raw position recordings produced at a paper speed of lOOmm/sec on an ink-jet recorder (Elema Mingograph). Saccades were designated PS if their latency was less than 1OOmsec which is generally considered to be insufIicient time to allow adequate visual processing (Becker and Fuchs, 1969). Each subject produced a sufficient number of PS (55 SD 12) to obtain reliable saccadic amplitude-velocity plots. The relationship of amplitude to velocity of one subject for both PS and RS saccades is shown in Fig. 1. It can be seen that the velocities for PS were slower than those of RS and this

A

518

M. BROSS~IN and C

KENVARI)

600 -

;i \

g

E t” ;:: > x : P

500

-

400

-

.

o.,“,

it 300

‘L

80°

0

.

*/’

-

/

:

l / ,*’ 8

200

-

l

Random

.(-I

Predictive

O(------I

I

I

I

5

10

15

Amplitude

I

I

20

2s

I 30

(deg)

Fig. 1.

Amplitude-peak velocity relationship of random and predictive eye saccades in one subject (No. 8 in Table 1). The amplitude IJIU!for the individual meuurrmcntr were of the initial eye saccade and not of the t8rgct diSpbcancn1. The curve fitted was Y-m + n Cxp(-A/p), where V is velocity. A ampiitudcandm,nandpare3 parrmam Utimati by fitting a curve by least squares regression. m is the asymptote, while m + n is the value of velocity at zero amplitude.

Tabkl.

Sxcadicpc&v&city(&&oc)d&odfram6ttcdcurvcat a~amplitmleof25dq

Subject

RuKiom(RS)

: 3 4 5 6 7 8 Mean f SD

535 677 664 412 554 669 532 546 573 f 91

was confirmed for all subjects, as shown in Table 1 (P < 0.01, Wilcoxon Signed Ranks Test). These diffmnces were more marked in larger amplitude saccades and this could account for the fact that Findlay (1981), working with saccades no bigger than 3.2 deg, failed to show slowing of saccades under predictive conditions. No consistent differences were found between the su@ects studied with the room dimly lit or in compktcdarkness. In addition to being slow, P!3 were highly hypometric. Man saccadic gain (initial sauznde amplitude/target amplitude) was 0.92 SD 0.03 for RS and 0.78 SD 0.03 for PS (Table 2). These difkrcuces were highly sign&ant (F = 122; P < 0.001; programme BMDP 1R for multiple linear regression analysis).

Pmdidivc(Ps)

Wruencc(RS-Ps)

430 463 561 321 506 565 so7 482 479 f 79

147 72 104 91 48 104 25 64 82 f 38

Saccadic inaccuracy was extremely frequent during PS, the proportion of normomctric saccadcs being less than 8%. The frequency of occurrence of inaccuracy increased with target amplitude, for RS but not for PS (Fig. 2). This wouId sum that the u8c of uncertain information about trvgst position, common to Ps at all target amptitudes, is the likely source of insocurecy.

7.5

0.97 SD 0.M 0.74 SD 0.09

ii.5

0.92 0.93 SD 0.03 0.78 0.81 SD 0.1 0.09I

30

0.89 SD 0.03 0.81 SD 0.12

519

Predictive eye saccades 100

r Random

*I-l

Predictive

Ol------)

Target amplitude Idagl Fig. 2. Percentage of normome&ic random and predi&ve saccades vs taget amplitude. Normometric saccades were defined as those showing no secondary or corrective saccade up to 1 set after the target had appeared. Vertical bars express I SD.

It has previously been shown that saccades are of reduced velocity and accuracy when made to an accoustic source (Zahn er al., 1978; Zambarbieri et al., 1982) and in darkness (Becker and Fuchs, 1969); to these may now be added predictive saccades. All three situations have one feature in common, the absence of a visible target prior to the onset of the saccade. This suggests that although various supranuclear neural centres are capable of triggering a saccade, unless they have been primed with visual information they are unable to provide an optimized neural input to the PPRF. Two structures are undoubtedly involved in the supranuclear control of saccade generation by the PPRF and ocular motor nuclei, the frontal eye field and the superior colliculus, which receives visual inputs both directly and indirectly via the parietal lobe (SchiIIer et al., 1980; Bruce and Goldberg, 1985). Clinical observations (Holmes, 1938; Luria et al., 1966; Guitton et al., 1985) as we11 as behavioural (Collin et al., 1982) and physiological work (Ingle and Sprague, 1975; Schiller, 1977; Kurt&erg and Vaughan, 1980) tend to attribute to the superior colliculus a more imprint role visually elicited saccades, in spontaneous, whereas internally organised, voluntary eye scanning is thought to be dependent on the frontal lobes. Therefore, we would suggest that the differences we have found between random (visually triggered) saccades and predictive (internally guided) saccades, might well reflect

selective or predominant activation by the superior colliculus and frontal eye fields respectively. In agreement with this view would be the facts that (a) collicuiar lesions in the monkey produce a more profound derangement in accuracy and a deficit in veIocity not seen after frontal lesions @chiller et al., 1980) and (b) collicular ablations abolish only saccades elicited by stimulation of the visual areas (occipital lobe) of the cerebral cortex but not those elicited by frontal eye field stimulation (Schiller, 1977). These results provide an example of stimulusdependent changes in saccadic characteristics and, from a clinical point of view, stress the importance of comparing similar testing conditions when drawing conchzsions about saccades in neurological diseases. For example, Parkinsonian patients were found to have relatively selective abno~aliti~ of predictive (Bronstein and Kennard, 1985) or self paced saccades (De Jong and Jones, 1971). We suggest that normative and clinical data for saccades should be obtained separately for RS and PS. Acknowtedgemenrs-We

are

extremely grateful to Mr S.

Evans and MS M. Fuller for statistic advia and to the Welfcome Foundation and the Mason Medical Research Trust for financial support. REFERENCES Becker W. and Fuchs A. F. (1%9) Further properties of the

human saccadic system: eye movements and comction saccades with and without visual fixation points. V&on Res. 9, 1247-1258.

520

A. M. BRONSTEIN and C.

Bronstein A. M. and Kennard C. (1985) Predictive ocular motor control in Parkinson’s disease. Brain 108,925-940. Bruce C. J. and Goldberg M. E. (1985) Primate frontal eye fields. I. Single neurons discharging before saccades. J. Neurophysiol. 53, 603-635.

Collin N. G., Cowey A., Latto R. and Narzi C. (1982) The role of frontal eye fields and superior colliculi in visual search and non-visual search in Rhesus monkeys. Behaa. Brain Res. 4, 171-193.

DeJong J. D. and Melvill Jones G. (1971) Akinesia, hypokinesia and bradykinesia in the oculomotor system of patients with Parkinson’s disease. Expl Neural. 32,58-68. Findlay J. M. (1981) Spatial and temporal factors in the predictive generation of saccadic eye movements. Qrion Res. 21, 347-354.

Guitton D. H., Buchtel A. and Douglas R. M. (1985) Frontal lobe lesions in man cause difficulties in suppressing reflexive glances and in generating goal-directed saccades. Expl Brain Res. 58, 455-572. Holmes G. (1938) The cerebral integration of the ocular movements. Br. Med. J. 2, 107-112. Ingle D. and Sprague J. M. (1975) Sensory motor function of the midbrain tectum. Neurosci. Res. Prog. Bull. 13, 45-57. Kurtzberg D. and Vaughan H. G. (1980) Differential topog-

KENNARD

raphy of human eye movement potentials precedmg visually triggered and self-initiated saccades. Prop. Bratn Res. 54, 203-208. Luria A. R., Karpov B. A. and Yarbusi A. L. ( 1966) Disturbances of active visual perception with lesions of the frontal lobes. Corerex 2, 202-212. Chiller P. H. (1977) The effect of superior colliculus ablation on saccades elicited by cortical stimulation, Brain Res. 122, 154-156.

Schiller P. H., True S. D. and Conway J. L. (1980) Deficits in eye movements following frontal eye-field and superior colliculus ablations. J. Neurophysiol. 44, 1175-l 194. Stark L., Vossius G. and Young L. R. (1962) Predictive control of eye tracking movements. IRE Trans. Hum. Fact. Electr. HFE3, 52-57. Zahn J. R., Abel L. A. and Dell’Osso L. F. (1978) Audioocular response characteristics, Sea.rorJi Proc. 2, 32-37. Zambarbieri D., S&mid R., Magnees G. and Prablanc C. (1982) Saccadic responses evoked by presentation of visual and auditory targets. Expl Brain Res. 47, 417-427. Zee D. (1984) Ocular motor control: the cerebral control of saccadic eye movements. in Neuro-ophthalmology 1984. 3 (Edited by Lessell S. and van Dalen J. T. W.), pp. 141-156. Elsevier, Amsterdam.