doi: 10.1053/ejpn.2001.0480 available online at http://www.idealibrary.com on European Journal of Paediatric Neurology 2001; 5: 107±114
Visual disorders in children with brain lesions: 1. Maturation of visual function in infants with neonatal brain lesions: correlation with neuroimaging ANDREA GUZZETTA,1 GIOVANNI CIONI,1 FRANCES COWAN,2 EUGENIO MERCURI2 1
Stella Maris Scientific Institute ± Division of Child Neurology and Psychiatry, University of Pisa, Italy; 2Department of Paediatrics and Neonatal Medicine, Imperial College School of Medicine, Hammersmith Hospital, London, UK
In the last two decades there has been a considerable increase in the understanding of visual maturation in the young infant and it has been possible to develop methods for testing visual function that are applicable in the neonatal period and early infancy. Several studies have reported that various aspects of visual function, such as visual acuity, visual fields and optokinetic nystagmus are often impaired in infants with brain lesions of antenatal and perinatal onset. We report our experience with such infants and a more general review of the literature on the maturation of visual function in the first years after birth in normal and brain damaged children. Keywords: Visual impairment. High-risk newborns. Brain damage. Periventricular leukomalacia. Visual acuity. Visual fields
Introduction Visual impairment, due to damage of central visual pathways, is known as cerebral visual impairment, and is frequent in children with brain lesions of antenatal and perinatal onset. Most of the studies reporting the incidence of visual disorders in infants and children with such brain lesions have been published in the last two decades. The literature can be divided in two main groups. The first includes papers describing the incidence of severe visual problems in children with cerebral palsy. The second group includes papers describing visual function in infants with brain lesions of onset in the first years of life. Until recently, most studies reporting visual abnormalities in children with brain damage only used standard ophthalmological evaluation to provide a gross estimate of the more severe visual abnormalities found in these children. This was mainly due to a lack of appropriate and specific
methods for assessing the maturing visual system in early childhood. In the last two decades, however, there have been considerable advances in our understanding of the maturation of visual function and in ways to assess it. It is now well accepted that the visual system functions mainly at a subcortical level in the newborn and in the first months after birth, and becomes progressively integrated with and dominated by cortical processes during the first year. A major breakthrough occurred when it became possible to assess acuity in the newborn by using acuity cards. Now it is possible to assess aspects of visual function such as, visual fields, attention at distance, discrimination of colour or orientation from the early post-neonatal period. Longitudinal studies have followed the onset and the maturation of different aspects of visual function in normal infants thus providing age-dependent normative data. These tests do not require verbal skills or a great degree of collaboration from the child and can be used not only in small infants but also in
Correspondence: Dr Eugenio Mercuri, Department of Paediatrics, Hammersmith Hospital, Du Cane Road, London W12 0HN, UK e-mail: [email protected]
& 2001 European Paediatric Neurology Society
108 children with severe mental retardation and behavioural problems. The development of these tests has made it possible to study visual development in young infants with brain lesions of antenatal and perinatal onset and to compare the findings with the results of other clinical and neuroradiological investigations. We report our experience and a more general review of the literature on the assessment of visual function in the first years after birth in prematurely born and full-term infants with specific patterns of brain lesions which have been characterized using cranial ultrasound and magnetic resonance imaging (MRI) scans. A review of the incidence of visual abnormalities in children with cerebral palsy and a more general review on the correlation between visual abnormalities and other aspects of development will be reported separately.
Assessment of visual function in the first years of life The assessment of visual function in young infants includes behavioural and electrophysiological techniques. Application of these tests should be preceded by a standard ophthalmological examination to ascertain the presence of eye abnormalities such as retinopathy, cataract or optic atrophy. . Blink reflex, in response to an approaching object, can already be observed in preterm infants,1 and is mediated by a tactile sensorial input processed subcortically. The visual component of the reflex can be tested by holding a transparent glass between the object and the infant's eyes. The blink reflex is cortically mediated and usually develops 17 weeks after birth.2 . Oculomotor behaviour can be assessed by testing fixation and following reactions. A short period of fixation on a target can be observed by 30 weeks post-menstrual age. At term age, newborns are generally able to follow a target, such as a red ball, in a full arc. The presence of abnormal eye movements, such as spontaneous nystagmus, can also be noted. Strabismus and eye alignment can be tested by commonly used orthoptic techniques, such as the cover test. . Acuity can be tested by using forced choice preferential looking. The infant is presented at eye level on one side of the midline with a target consisting of black and white stripes paired with a uniform grey background on the other side. The level of acuity is measured as the finest
Review article: A Guzzetta et al. grating (i.e. width of black and white stripes) for which the infant shows a consistent preference (in cycles/degree), and compared with age-specific normative data.3 As a rule of thumb, visual acuity shows a maturation of one cycle/ degree/month in the first months of life. The acuity cards commercially available, are a simplified adaptation of the forced choice preferential looking technique (Fig. 1). . Visual fields can be assessed using kinetic perimetry. The apparatus consists of two perpendicular black metal strips bent to form two arcs, each with a radius of 40 cm. The infant is held in the centre of the arc perimeter. During central fixation of a white ball, an identical target is moved from the periphery towards the fixation point along one of the arcs of the perimeter (Fig. 2). Eye and head movements towards the peripheral ball are used to estimate the outline of the visual fields. Normative data for full-term and preterm infants are available.4,5 The fields are quite narrow in the first months after birth (approximately 30 degrees) and become progressively wider (approximately 60 degrees at 6 months and 80±90 degrees at 1 year of age). . Optokinetic nystagmus can be elicited by using a large piece of paper or a computer-generated random dot pattern in front of the infant's face. The examiner observes the infant's eye movements, recording the presence and the symmetry in ease of generation of the optokinetic nystagmus in response to the movement of the pattern in either direction. Normally, binocular optokinetic nystagmus is symmetrical from birth onwards,
Fig. 1. Example of testing visual acuity by means of Teller acuity cards.
Review article: Visual defects and neonatal brain lesions
109 degrees. The phase reversal response is already present at term, while the orientation reversal response is only consistently elicited at 10 weeks post-term for slow changes (4 rev/s) and after 12 weeks for faster changes (8 rev/s)12 (Fig. 4).
Figure 5 shows details of the maturation of various aspects of visual function in the first year of life.
Fig. 2. Example of testing visual fields by means of the kinetic perimeter.
whereas monocular optokinetic nystagmus shows a better response to stimulation in a temporonasal direction up to about 3±6 months of corrected age.6 . Fixation shift is a test of visual attention evaluating the direction and the latency of saccadic eye movements in response to a peripheral target in the lateral field. A central target is used as a fixation stimulus before the appearance of the peripheral target. While in some trials the central target disappears simultaneously with the appearance of the peripheral target (noncompetition) in others the central target remains visible generating a situation of competition between the two stimuli (Fig. 3). Normal children can reliably shift their attention in a situation of non-competition during the first weeks after birth, but brisk refixations in a situation of competition is only found after 6±8 weeks post-term and reliably by 12±18 weeks post-term. Absent or delayed (a latency of more than 1.2 seconds) refixation at 5 months of age is considered abnormal.7 . Visual evoked potential can be recorded by using flash or orientation-reversal and phase-reversal stimuli. Using flash stimuli it is possible to follow the normal or abnormal maturation of the visual pathway.8±11 Steady-state flash visual evoked potentials have also been used to assess the maturation of cortical and subcortical processing of the visual pathway.12,13 For phasereversal visual evoked potentials the orientation of the black and white stripes is fixed but the contrast reversed periodically. For orientation reversal visual evoked potentials, stimuli periodically change orientation between 45 and 135
Incidence and type of visual impairment in preterm and fullterm infants with specific patterns of brain lesions Periventricular leukomalacia Several studies have assessed visual acuity in infants with periventricular leukomalacia.14±16 The incidence of low acuity in these children is above 60%. Acuity is generally normal in the infants with 'prolonged flares' or periventricular leukomalacia type 1 and 2 (according to the classification of deVries et al.17) and is progressively more severely affected in infants with periventricular leukomalacia grades 3 and 4.14,18±21 The degree of visual impairment in children with periventricular leukomalacia is significantly associated with lesions in the peritrigonal white matter, with the involvement of optic radiation (Fig. 6), as well as with the extent of occipital cortex involvement.14±16 Visual impairment is common and generally severe in infants with subcortical cystic leukomalacia,17,19,20,22 while it is less common in infants with cystic periventricular leukomalacia. Eken et al.19
Stimuli used for testing fixation shift in infants.
Review article: A Guzzetta et al.
Stimuli used for testing phase shift and orientation reversal visual evoked potentials in infants.
Graph illustrating the maturation of some visual functions in the first year of life.
have reported that in infants with cystic periventricular leukomalacia visual abnormalities are more frequent in infants with gestational ages of 35±37 weeks than in those with gestational age below 32 weeks. The difference may be explained by the distribution of the lesion, the mature infants having cystic lesions that extend further into the subcortical white matter with subsequent higher risk of abnormal visual outcome. Other studies have also demonstrated that, in addition to visual acuity, other aspects of visual function, such as visual fields and eye movements are also frequently affected in these children.23,24
Intraventricular haemorrhage Small germinal layer or intraventricular haemorrhages (grade I and IIa25) are generally associated with normal acuity.19,26 Infants with large haemorrhages show poor acuity at term age but this tends
to improve after a few months.19 Similar findings have been reported by Morante et al.27 and Dubowitz et al.28 This may be due to the effect of large intraventricular haemorrhage on the thalamus or inferior colliculi, or to the bleeding of the germinal matrix at the origin of the optic radiations and the posterior thalami. Therefore these effects may be transient, as there is no major tissue involvement. Permanent effects are not so common as the parenchymal involvement with grade IV lesions is more often in mid-anterior parietal lobe and not so often in sites affecting the primary visual pathway i.e. posterior parietal and occipital lobes.
Hypoxic±ischaemic encephalopathy Visual abnormalities are very common in infants with hypoxic±ischaemic encephalopathy.29±31 The presence of visual impairment is not always
Review article: Visual defects and neonatal brain lesions
Fig. 6. Male infant aged 16 months with cerebral visual impairment. T2 weighted spin echo sequence (TR 2000/ 100) image showing high signal intensity at both optic radiations extending to surrounding white matter.
Fig. 7. Male infant aged 2 weeks with delayed visual maturation. T1 weighted spin echo sequence (TE 860/20) image showing bilateral abnormal signal in the lentiform nuclei and thalami.
correlated with the severity of hypoxic±ischaemic encephalopathy at birth. While infants with stage 1 hypoxic±ischaemic encephalopathy, according to Sarnat and Sarnat grading,32 have normal visual function and those with stage III hypoxic±ischaemic encephalopathy always have severe visual impairment, the visual outcome in stage II hypoxic±ischaemic encephalopathy is less predictable. In hypoxic±ischaemic encephalopathy the presence and the severity of visual impairment is significantly related to the extent of brain lesions seen on MRI, particularly those affecting the basal ganglia and thalami. It is important to recognize that not all lesions involving the occipital lobes are associated with impaired visual function but in our experience infants who have lesions involving both the hemispheres and the basal ganglia (putamen and caudate) invariably have persistent and severe abnormalities of various aspects of visual function. Visual abnormalities may also be found in the first months after birth in infants with isolated basal ganglia lesions but these tend to recover by the end of the first year.29
such as visual fields and fixation shift, can be abnormal.33±35 The presence and severity of the impairment, however, cannot always be predicted from the site and the extent of the lesions as seen on
Focal cerebral infarction (stroke) Although visual acuity is generally normal in infants with focal lesions and in particular unilateral lesions, other aspects of visual function,
Fig. 8. Female infant age 15 months with cerebral visual impairment. Inversion recovery (TR 3500/30) image showing severe atrophy of the basal ganglia and thalami and diffuse abnormality in the white matter.
112 MRI. Unlike adults with similar lesions who show a reliable association between the involvement of the occipital cortex and the contralateral visual field, visual fields can be normal in up to 50% of the infants with occipital lobe infarction.36
Visual impairment in infants with brain lesions: rules and exceptions From our experience and from the review of the literature investigating visual function in infants with brain lesions of antenatal and perinatal onset it is clear that visual abnormalities are frequent in infants with brain lesions but that the association between lesions in the visual pathway and visual impairment does not always follow the rules observed in adults with lesions involving the visual pathway. The early use of neonatal brain MRI has given us considerable insight into how visual function is related to the site and extent of brain lesions that occur in the perinatal period. The following points have become clear: . Visual abnormalities tend to be more frequent in full-term infants with hypoxic±ischaemic encephalopathy than in preterm infants. This is probably due to the low incidence of visual abnormalities in the preterm infants with mild leukomalacia or with haemorrhages. Whilst in preterm infants lesions of the occipital visual cortex are generally associated with impaired visual function, in full-term infants both unilateral and bilateral occipital lesions are often associated with normal vision. The different visual behaviour might be explained by the difference in type and site of lesions, as in severe periventricular leukomalacia the lesions are generally quite extensive and bilateral, but also by other risk factors related to prematurity. Extreme prematurity is generally associated with many other more subtle problems (from poor environment, poor nutrition, poor long-term oxygenation, infection etc) and hence no part of the brain is entirely normal and this might affect the process of functional reorganization of the brain after the insult. . In full-term infants the severity of visual impairment seems to be more related to the concomitant involvement of basal ganglia and thalami. The role played by basal ganglia and thalami in visual maturation is still not fully understood. The neuroanatomical basis for these findings may result from the extensive
Review article: A Guzzetta et al. reciprocal connections between visual cortical areas and basal ganglia.37±39 Any interruption to these connections may reduce the transfer of information to other parts of the brain and hence reduce the possibility that other cortical areas may take over function from a damaged occipital region. In other words damage to the subcortical structures may, by inhibiting a general exchange of information within the brain, preclude the possibility of functional reorganization of the damaged developing brain. . Visual function can be abnormal in children with a normal ophthalmological examination and normal optic radiations and visual cortex. This may be explained by the involvement of parts of the brain other than geniculostriate pathways, such as the frontal or the temporal lobes, which are known to be associated with visual attention or with other aspects of visual function. In some cases visual attention, and more generally visual function, may also be disturbed by other clinical problems often occurring in children with brain lesions, such as severe oculomotor impairment or severe epilepsy.33 . A proportion of infants show transient visual impairment with a gradual recovery40 which, in some cases, occurs in the first months after birth. These infants are described as having `delayed visual maturation'. This term is used to describe infants with reduced vision at birth, which subsequently improves by the end of the first year of life. Delayed visual maturation can be an isolated finding, or associated with ocular and neurodevelopmental abnormalities.41 Several mechanisms have been proposed to account for visual recovery in these infants, such as the use of extrageniculo-striate pathways,42,43 the restoration of normal excitability to the surviving neurons, or the recruitment of neurons adjacent to the lesion.44 Although many infants with delayed visual maturation have associated neurodevelopmental abnormalities, brain imaging has only been performed in two studies.45,46 Mercuri et al.13,46 have studied a cohort of full-term infants with neonatal brain lesions, and found that delayed visual maturation was mainly present in those with isolated basal ganglia lesions (Fig. 7). One possible explanation is that the isolated involvement of these subcortical structures causes delayed visual maturation in the first months when visual function is mainly subcortical, and that vision improves in parallel with the onset of more mature aspects of cortical visual functioning. Further imaging studies in infants without obvious perinatal problems are needed to exclude the possibility that minor
Review article: Visual defects and neonatal brain lesions lesions might also occur in that group and be responsible for the delayed visual maturation.
Conclusion The development of new tools for the assessment of visual function in young infants has provided new and more objective information about the early disturbance of visual development in children with brain lesions of perinatal onset. Although neuroimaging techniques are useful in predicting the occurrence and extent of visual impairment, the correlation between imaging findings and function is not always consistent especially in the infant born at term. This is probably due to the involvement of the complex extrastriatal visual pathways and also to the plasticity of the developing brain. Future studies with larger cohorts of infants with a wider variety of lesions and longer follow-up will provide a better comprehension of the different mechanisms underlying visual development.
Acknowledgements The authors gratefully acknowledge the contribution from the colleagues of the Visual Development Unit (University College, London) and the Paediatric and MRI departments (Hammersmith Hospital, London) and of the Division of Child Neurology and Psychiatry and the MRI department (University of Pisa and Stella Maris Scientific Institute). GC and AG were supported by grant RC 1/2000 of the Italian Ministry of Health.
References 1 Hack M, Muszynski SY, Miranda SB. State of awakeness during visual fixation in preterm infants. Pediatrics 1981; 68: 87±92. 2 van Hof-van Duin J, Mohn G. Visual field measurements, optokinetic nystagmus and the visual threatening response: normal and abnormal development. Doc Ophthalm 1986; 45: 305±316. 3 Teller DY, McDonald MA, Preston K et al. Assessment of visual acuity in infants and children: the acuity card procedure. Dev Med Child Neurol 1986; 26: 779±789. 4 Mohn G, van Hof-van Duin J. Development of the binocular and monocular visual field during the first year of life. Clin Vis Sci 1986; 1: 51±64. 5 van Hof-van Duin J, Heersema DJ, Groenendaal F et al. Visual field and grating acuity development in low-risk preterm infants during the first 21¤2 years after term. Behav Brain Res 1992; 49: 115±122.
6 Atkinson J, Braddick OJ. Development of optokinetic nystagmus in infants: an indicator of cortical binocularity. In: Fisher DF, Monty RA, Sender JW (eds) Eye Movements: Cognition and Visual Perception. Hillsdale NJ: Lawrence Erlbaum Associates, 1981: 53±64. 7 Atkinson J, Hood B, Wattam-Bell J, Braddick O. Changes in infants' ability to switch visual attention in the first three months of life. Perception. 1992; 21: 643±653. 8 Taylor MJ, McCulloch DL. Visual evoked potentials in infants and children. J Clin Neurophysiol 1992; 9: 357±372. 9 Eken P, de Vries LS, van Nieuwenhuizen O et al. Early predictors of cerebral visual impairment in infants with cystic leukomalacia. Neuropediatrics 1996; 27: 16±25. 10 Tsuneishi S, Casaer P. Stepwise decrease in VEP latencies and the process of myelination in the human visual pathway. Brain Dev 1997; 19: 547±551. 11 Tsuneishi S, Casaer P. Effects of preterm extrauterine visual experience on the development of the human visual system: a flash VEP study. Dev Med Child Neurol 2000; 42: 663±668. 12 Braddick OJ, Wattam-Bell J, Atkinson J. Orientationspecific cortical responses develop in early infancy. Nature 1986; 320: 617±619. 13 Mercuri E, Braddick O, Atkinson J, et al. Orientationreversal and phase-reversal visual evoked potentials in full-term infants with brain lesions: a longitudinal study. Neuropediatrics 1998; 29: 169±174. 14 Cioni G, Fazzi B, Coluccini M et al. Cerebral visual impairment in preterm infants with periventricular leukomalacia. Pediatr Neurol 1997; 17: 331±338. 15 Uggetti C, Egitto MG, Fazzi E et al. Cerebral visual impairment in periventricular leukomalacia: MR correlation. Am J Neuroradiol 1996; 17: 979±985. 16 Eken P, de Vries L, van der Graaf Y et al. Haemorrhagic±ischaemic lesions of the neonatal brain: correlation between cerebral visual impairment, neurodevelopmental outcome and MRI in infancy. Dev Med Child Neurol 1995; 37: 41±55. 17 de Vries LS, Eken P, Dubowitz LMS. The spectrum of leukomalacia using cranial ultrasound. Behav Brain Res 1992; 49: 1±6. 18 Gibson NA, Fielder AR, Trounce JQ, Levine MI. Ophthalmic findings in infants of very low birthweight. Dev Med Child Neurol 1990; 32: 7±13. 19 Eken P, van Nieuwenhuizen O, van der Graaf Y et al. Relation between neonatal cranial ultrasound abnormalities and cerebral visual impairment in infancy. Dev Med Child Neurol 1994; 36: 3±15 20 Scher MS, Dobson V, Carpenter NA, Guthrie RD. Visual and neurological outcome of infants with periventricular leukomalacia. Dev Med Child Neurol 1989; 31: 353±365. 21 Lanzi G, Fazzi E, Uggetti C et al. Cerebral visual impairment in periventricular leukomalacia. Neuropediatrics 1998; 29: 145±150. 22 Calvert SA, Hoskins EM, Fong KW, Forsyth SC. Periventricular leukomalacia: ultrasonic diagnosis and neurological outcome. Acta Paediat Scand 1986; 75: 489±496.
Review article: A Guzzetta et al. Cioni G, Bertuccelli B, Boldrini A et al. Correlation between visual function, neurodevelopmental outcome, and magnetic resonance imaging findings in infants with periventricular leukomalacia. Arch Dis Child Fetal Neonatal Ed 2000; 82: F134±140. van den Hout BM, Stiers P, Haers M et al. Relation between visual perceptual impairment and neonatal ultrasound diagnosis of haemorrhagic±ischaemic lesions in 5-year-old children. Dev Med Child Neurol 2000; 42: 376±386. Levene MI, de Crespigny LC. Classification of intraventricular hemorrhage. Lancet 1983; i: 643. van Hof-van Duin J, Mohn G. Vision in the preterm infant. In Prechtl HFR (ed) Continuity of Neural Functions from Prenatal to Postnatal Life. Clinics in Developmental Medicine. No. 94. London: SIMP, 1984: 93±114. Morante A, Dubowitz LMS, Levene MI, Dubowitz V. The development of visual function in normal and neurologically abnormal preterm and fullterm infants. Dev Med Child Neurol 1982; 24: 771±784. Dubowitz LMS, Mushin J, Morante A, Placzek M. The maturation of visual acuity in neurologically normal and abnormal newborn infants. Behav Brain Res 1983; 10: 39±45. Mercuri E, Atkinson J, Braddick O et al. Basal ganglia damage and impaired visual function in the newborn infant. Arch Dis Child Fetal Neonatal Ed 1997; 77: F111±114. Mercuri E, Atkinson J, Braddick O et al. Visual function in full-term infants with hypoxic±ischaemic encephalopathy. Neuropediatrics 1997; 28: 155±161. Groenendaal F, van Hof-van Duin J, Baerts W, Fetter WP. Effects of perinatal hypoxia on visual development during the first year of (corrected) age. Early Hum Dev 1989; 20: 267±279. Sarnat HB, Sarnat MS. Neonatal encephalopathy following neonatal distress. A clinical and electroencephalographic study. Arch Neurol 1976; 33: 696±705. Cioni G, Fazzi B, Ipata AE et al. Correlation between cerebral visual impairment and magnetic resonance imaging in children with neonatal encephalopathy. Dev Med Child Neurol 1996; 38: 120±132.
34 35 36 37 38
40 41 42 43 44
Mercuri E, Atkinson J, Braddick O et al. Visual function and perinatal focal cerebral infarction. Arch Dis Child Fetal Neonatal Ed 1996; 75: F76±81. Mercuri E, SpanoÁ M, Bruccini G et al. Visual outcome in children with congenital hemiplegia: correlation with MRI findings. Neuropediatrics 1996; 27: 184±188. Guzzetta A, Fazzi B, Mercuri E et al. Visual function in hemiplegic children in the first years of life. Dev Med Child Neurol (in press). Ungerleider LG, Desimone R, Galwin TW, Mishkin M. Subcortical proiections of area MT in the macaque. J Comp Neurol 1984; 223: 368±386. Updyke BV. Organisation of visual corticostriatal projections in the cat, with observations on visual projections to claustrum and amygdala. J Comp Neurol 1993; 327: 159±193. Serizawa M, McHaffie JG, Hoshino K, Norita M. Corticostriatal and corticotectal projections from visual cortical areas 17, 18 and 18a in the pigmented rat. Arch Histol Cytol 1994; 57: 493±507. Groenendaal F, van Hof-van Duin J. Visual deficits and improvements in children after perinatal hypoxia. J Vis Impair Blind 1992; 86: 215±218. Tresidder J, Fielder AR, Nicholson J. Delayed visual maturation: ophthalmic and neurodevelopmental aspects. Dev Med Child Neurol 1990; 32: 872±881. Atkinson J. Early visual development: differential functioning of parvocellular and magnocellular pathways. Eye 1992; 6: 129±135. Payne BF, Cornwell P. System-wide repercussions of damage to the immature visual cortex. TINS 1994; 17: 126±130. Lambert SR, Hoyt CS, Jan JE et al. Visual recovery from hypoxic cortical blindness during childhood. Computed tomographic and magnetic resonance imaging predictors. Arch Ophthal 1987; 105: 1371±1377. Hoyt CS, Good WV. Visual factors in developmental delay and neurological disorders in infants. In: Simons K (ed) Early Visual Development, Normal and Abnormal. Oxford: Oxford University Press, 1993: 505±512. Mercuri E, Atkinson J, Braddick O et al. The aetiology of delayed visual maturation: short review and personal findings in relation to magnetic resonance imaging. Eur J Paediatr Neurol 1997; 1: 31±4.