Journal of Neuro-Ophthalmology, June 1999, Volume 19, Issue 2

From Cortical Plasticity to Unawareness of Visual Field Defects

It was long held that, following alterations in sensory input, structural changes in the primary visual cortex take place only in early life, during so-called "critical periods." Recently, however, it has been established that, in adults, cortical maps in the brain are not fixed, and the cortex does not perform stereotyped operations. Instead, neuronal receptive fields in the cortex can reorganize following deactivation or an altered pattern of activation. Plasticity is essential for the normal adjustment of the brain to modifications in the sensory environment, and for improving perceptual skills and sensorimotor performances. It also plays a crucial role in recovery from damage to the visual system. Cortical remapping generates a filling-in of visual field defects. Consequently, it alters the image perceived. Cortical rearrangement following lesions in the visual pathways does not restore function to the destroyed tissue, but it helps to compensate for gaps in perception. In this review article, we focus on effects of plasticity in the adult visual cortex which are of major importance in the daily practice of neuroophthalmology. Cortical reorganization, together with resulting filling-in, affects the early recognition and evaluation of visual field defects. The importance of brain remapping in these matters is still largely underestimated by clinicians.

Journal of Neuro- Ophthalmology 19( 2): 84- 88, 1999. © 1999 Lippincott Williams & Wilkins, Inc., Philadelphia From Cortical Plasticity to Unawareness of Visual Field Defects Avinoam B. Safran, M. D., and Theodor Landis, M. D. It was long held that, following alterations in sensory input, structural changes in the primary visual cortex take place only in early life, during so- called " critical periods." Recently, how­ever, it has been established that, in adults, cortical maps in the brain are not fixed, and the cortex does not perform stereotyped operations. Instead, neuronal receptive fields in the cortex can reorganize following deactivation or an altered pattern of acti­vation. Plasticity is essential for the normal adjustment of the brain to modifications in the sensory environment, and for im­proving perceptual skills and sensorimotor performances. It also plays a crucial role in recovery from damage to the visual system. Cortical remapping generates a filling- in of visual field defects. Consequently, it alters the image perceived. Cortical rearrangement following lesions in the visual pathways does not restore function to the destroyed tissue, but it helps to compensate for gaps in perception. In this review article, we focus on effects of plasticity in the adult visual cortex which are of major importance in the daily practice of neur­ophthalmology. Cortical reorganization, together with resulting filling- in, affects the early recognition and evaluation of visual field defects. The importance of brain remapping in these mat­ters is still largely underestimated by clinicians. Key Words: Retina- Visual cortex- Cortical plasticity- Visual field- Filling- in- Troxler's phenomenon. It has long been held that, after alterations in sensory input, structural changes in the primary visual cortex take place only in early life, during so- called " critical peri­ods." In recent years, however, a large body of evidence has shown that, even in adults, some essential character­istics of cortical cells are affected by experience, and sensory and motor maps are not stable. Alteration of input from the periphery results in changes in topography in the cortex, including the primary visual cortex ( 1,2). When lesions occur in the visual pathways, cortical re­arrangement does not restore function to the destroyed tissue, but it helps to compensate for gaps in perception. The clinical implications of these processes are immense, yet widely unrecognized ( 3). Manuscript received January 22, 1998; accepted February 10, 1999. From the Ophthalmology ( A. B. S.) and the Neurology Clinics ( T. L.), Department of Clinical Neuroscience, Geneva University Hospitals, Geneva, Switzerland. Address correspondence and reprint requests to Dr. Avinoam B. Safran, Ophthalmology Clinic, Geneva University Hospitals, 1211 Geneva 14, Switzerland. THE NATURE OF FILLING- IN Filling- in is one of the main mechanisms involved in cortical plasticity occurring after focal visual deafferen-tation. It causes visual stimuli to be perceived as arising from an area of the visual field where there is no actual visual input. The scotoma appears, in some way, to be invaded by the pattern from neighboring areas of the visual field ( 4). Consequently, affected subjects ignore or underestimate their defects. Filling- in occurs in normal and disease conditions. As a result, when viewing a scene with only one eye, the viewer does not perceive the scotomata produced by the optic disc and by blood vessels overlying retinal photo­receptors ( 5). Acquired defects in the visual field can also be filled in and usually are not perceived ( 6- 10). The presence of these defects is often recognized only indirectly, owing to the invisibility of objects located within the altered area of the visual field. When visual field defects are limited and do not involve the foveal area ( i. e., when visual acuity is preserved), recognition of the scotoma depends on careful observation. ARTIFICIAL SCOTOMATA To conduct a systematic analysis of filling- in charac­teristics, Ramachandran and Gregory ( 11) devised visual stimuli to induce in healthy people scotoma- like changes that were temporary and entirely harmless. Such changes, called artificial scotomata, were generated as follows. The subject was presented with a small mask occluding part of the visual field, while, over the sur­rounding region, a random- dot pattern was shown ( simi­lar to the white- noise flicker on a television screen after the end of transmission). The subject was then asked to fixate a spot located several degrees to one side of the masked area. After a few seconds of steady fixation, the masked area completely vanished, having been filled in by the surrounding pattern ( Fig. 1) ( 11). Random texture patterns are particularly effective in generating artificial scotomata when the pixels in the texture are flickering. FILLING- IN EXPERIENCED WITH A PAINTING BY CLAUDE MONET We recently observed that contemplating Impression: Rising Sun by the French impressionist Claude Monet, 84 PERCEPTUAL FILLING- IN 85 Fixation point FIG. 1. Artificial scotoma. Filling- in can be experienced in the small, homogeneous area, when steadily fixating the fixation point for approximately 10 seconds. The small area then gradu­ally vanishes, having been replaced by the texture pattern invad­ing from the surroundings. can produce a remarkable experience of perceptual fill­ing- in ( 12). In this painting, which depicts a sunrise over the harbor of Le Havre, when the observer looks steadily at the head of the sailor standing in the boat, the sun's image gradually vanishes within less than 20 seconds. During this process, the solar disc seems to be invaded by the brightness and color of the surrounding sky, until it can no longer be distinguished from the sky. The condition was recreated experimentally, on a computer screen. The sun was represented as a red disc against a blue background. When filling- in of the solar disc by the surrounding sky had occurred, it was found that, if the blue background was switched to a red one, identical with the sun in color, a bluish color was still perceived for several seconds in the area originally oc­cupied by the solar disc. This observation supported the suggestion that the filling- in phenomenon observed in this picture reflected a process of cortical remapping. COMPETING MECHANISMS More than one mechanism is involved in perceptual filling- in. For example, when a masked area is defined against a random noise by a difference in motion and color, filling- in occurs in successive stages, involving first the color component, then the movement ( 11). Dif­ferent mechanisms involved in visual processing can sometimes compete for filling in the gap. We observed such a competition in subjects with a small right super­ior homonymous paracentral scotoma, resulting from a left occipital lesion ( 16). Amsler grids were used, which have two diagonals superimposed on a regular grid pattern ( Fig. 2). When presented with the chart, the patients experienced filling- in of the regular grid pattern. However, no completion was observed with the diagonal line crossing the scotoma. In contrast, di­agonal lines presented in isolation showed a completion phenomenon. \ _ -- ^ ^ ; ^ \ \ /_^ ^_ A.*^^*-" ^ L £ - ' " * * . / < TZr& 1 A4 • - 3 \ * L* n* iZ^ S I! tL ^ N Perceived defect in the diagonal line Visual target Actual borders of the scotoma, delineated with a target Perceived defect in the grid pattern FILLING- IN AND TROXLER PHENOMENON In the article about Monet's painting, the possible as­sociations between filling- in and Troxler phenomenon were also discussed. Troxler described a perceptual phe­nomenon defined as a fading of stabilized images located in the peripheral visual field ( 13). It has been speculated that Troxler phenomenon either originates in local adap­tation mechanisms ( e. g., in bleaching of photoreceptors at the positions of the stabilized images) or depends on neural mechanisms ( 14,15). We suggested that Troxler phenomenon in fact relies on filling- in mechanisms, be­cause the stabilized, disappearing images acquire the vi­sual attributes of neighboring areas ( 12). We chose how­ever, when describing the phenomenon of the sun's van­ishing, to refer to filling- in rather than to Troxler phenomenon, because the mechanisms involved are bet­ter explained by filling- in than by general understanding of Troxler effect. Moreover, in the visual cortex, neuro-physiologic correlates of perceptual filling- in have been clearly established, as mentioned earlier. B Perceived defect in the diagonal line Actual borders of the scotoma FIG. 2. Dissociation between perceived and unperceived defects. A: In this Amsler grid, a patient with an absolute homogeneous paracentral scotoma experienced partial filling- in of the grid pat­tern. However, he did not observe any filling- in with the diagonal line crossing the gap in the visual field. B: In contrast, when diagonal lines were presented in isolation, they also demon­strated partial filling- in. Modified from Safran and Landis ( 3). J Neuw- Ophthalmol, Vol. 19, No. 2, 1999 86 A. B. SAFRANAND T. LANDIS CORTICAL REMAPPING AFTER FOCAL BINOCULAR RETINAL LESIONS To study experience- dependent plasticity in the visual system, focal retinal lesions at homologous positions were produced by Gilbert et al. ( 1) in eyes of cats and monkeys, thereby removing visual input from a limited area of the visual cortex. Lesions in the retina first si­lenced the area in the primary visual cortex correspond­ing to the part of the retina destroyed. Over the next few months, the silenced area of cortex recovered functional visual input. The areas of cells that recovered visual re­sponses shifted from the damaged part of the retina to surrounding regions ( 1). These processes may involve a differential strength­ening and weakening of subsets of connections within extensive axonal fields, the long- range horizontal con­nections representing a likely substrate for many of the observed effects. This idea is supported by the finding that, after recovery, the pattern of orientation columns was similar to that seen before the lesion was made, although the receptive fields of the cells in this region had shifted considerably ( 17). Approximately a year after the lesion was made, the density of the horizontal pro­jection into the reorganized region had doubled, indicat­ing that the strengthening was mediated by sprouting of axon collaterals and synaptogenesis ( 18). PERCEPTUAL FILLING- IN WITHOUT VI REORGANIZATION Although reorganization clearly occurs in the primary visual cortex, when lesions have been produced in cor­responding parts of both retinas, conflicting results have been reported with unilateral retinal lesions ( 19). To clarify this issue, a study was recently conducted in the monkey by Murakami et al., ( 20) with monocular laser- induced retinal lesions. They found strong evidence for filling- in a few days after making the lesion. How­ever, they were unable to show irregularity in the VI topography in the lesioned- eye- only viewing condition. The maintained input from the intact eye to the affected cortical region may have prevented neuronal reorganiza­tion. A CONTROVERSY ABOUT SHORT- TERM CHANGES IN THE VISUAL CORTEX The form of plasticity described implies alterations that develop over a period of months. To investigate whether faster changes are observable when a visual field defect is produced, Pettet and Gilbert ( 21) con­ducted a study using an artificial scotoma in the primary visual cortex of cats. The results of this experiment demonstrated that, only minutes after the pattern was presented, the receptive fields of cells, originally located within what was later to become the boundary of the scotoma, had expanded, av­eraging a fivefold increase in area and had shifted to positions outside the lesion ( 22). In another study, de Weerd et al. ( 23) showed that the firing rates of cells in which receptive fields were located within the artificial scotoma gradually increased during stimulus presenta­tion, until they reached a plateau at a rate similar to that in the area surrounding the artificial scotoma. In recent observations, DeAngelis et al. ( 24) chal­lenged the findings of receptive field expansion in re­sponse to artificial scotomata. Measurements were made using a conditioning stimulus similar to those used pre­viously, but the mapping technique was different from that used in the earlier study. Chapman and Stone ( 25) analyzed the data presented by the two laboratories and concluded that the major differences between the two studies were not in the re­sults but rather in the interpretation. Pettet and Gilbert ( 21) found a change in cell minimum response fields and concluded that receptive field size had increased. In con­trast, Chapman and Stone ( 25) concluded that it was appropriate to consider that a neuron has undergone an increase in receptive field size when the region of the visual field over which it responds has grown fivefold, and when the neuron then responds to images in places where it did not respond before. FILLING- IN AND BRIGHTNESS PERCEPTION Filling- in percepts can be generated by luminance modulation. It was recently reported by Paradiso and Hahn ( 26) that when the luminance of a homogeneous spot of light gradually increases or decreases, there are conditions in which the brightness of the spot is spatially nonuniform. When luminance is increased, apparent brightness spreads from the edge of the spot toward the center. Conversely, when the luminance is progressively reduced, the center of the spot appears brighter, and darkness seems to spread inward ( Fig. 3). That filling- in is based on the activation of neurons in the visual cortex, rather than in the retina, is supported by the fact that the percept is nearly identical, even when the stimulus spans the blind spot. Although the only part of the stimulus activating photoreceptors is an annulus, brightness or darkness appears to sweep to the center of the disk ( 26). It was concluded that the filling- in component of brightness is responsible for spatial interactions involved in brightness perception. However, this is probably not the only determinant of brightness. A direct neural re­sponse to the onset of light is also likely to occur, and this is modified by filling- in ( 26). UNRECOGNIZED CLINICAL IMPLICATIONS OF FILLING- IN Filling- in has considerable but widely underestimated consequences in clinical practice ( 3). It significantly de­lays the recognition of visual field defects and therefore delays treatment, especially when scotomata do not af­fect the foveal function- that is, when visual acuity is preserved. Although not usually mentioned in textbooks, in our opinion, filling- in is a major cause of failure of J Neuro- Ophthalmol, Vol. 19, No. 2, 1999 PERCEPTUAL FILLING- IN 87 Luminance time ^<& m. Brightness time FIG. 3. Filling- in effect on brightness perception. When the lumi­nance of a spot of light is gradually reduced, the brightness of the spot appears to be nonhomogeneous, with the darkness seeming to spread inward. Modified from Paradiso and Hahn ( 26). patients with simple chronic glaucoma to recognize their visual defects at an early stage. It dramatically affects patients with ring scotomata due to pigmentary retinop­athy, who often remain unaware of the defect until late in the disease process. It causes underestimation of visual dysfunction due to photocoagulation in diabetic retinop­athy and probably in many other ocular conditions. In clinical practice, the major implications of filling-in, and indeed its very occurrence, have rarely been rec­ognized. Unawareness ( or underestimation) of the effects of filling- in on the perception of visual field defects was demonstrated in a recent article about the measurement of visual function and the quality of life in patients with cytomegalovirus retinitis ( 27). The authors reported the results of a questionnaire designed to assess visual symp­toms in affected patients, including perception of blurred vision, floaters, blind spots, seeing to the side, and bump­ing into things. They noted that the question about blind spots or blurry spots was misinterpreted by patients as referring to floaters. We believe that this observation reflected that, as a result of filling- in, affected subjects usually do not perceive their visual field defect ( 28). BIASED VISUAL FIELD EVALUATIONS Visual field assessment can also be markedly affected by filling- in when examination techniques rely on the subject's perception of the defect against a structured background, as with the Amsler grid test and noise- field campimetry ( NFC). By comparing the results of delineating central scoto­mata using either a tangent screen or an Amsler grid, we assessed filling- in in subjects with macular disorders. As a rule, Amsler grid testing showed smaller defects than did the tangent screen and occasionally failed to show any abnormality at all, even in patients with marked dis­orders. Moreover, when two examinations were con­ducted in succession, the size and location of the appar­ent defects varied markedly, probably because of the dynamic nature of filling- in ( 9). We also investigated filling- in with Amsler grids in two patients with a right homonymous paracentral scotoma due to left occipital damage. Both showed filling- in. With NFC, defects are defined against a field of black-and- white dots flickering randomly at high frequency ( 27,29). They appear as areas showing less or no flick­ering, or a brightness that differs from that of the sur­rounding area. Although congenital visual field defects, long- stand­ing scotomata, and old postgeniculate lesions cannot usu­ally be detected using NFC ( 30), owing to filling- in, acquired circumscribed field defects, resulting from damage in the retina, or in the pregeniculate visual path­ways, usually are clearly perceived in the noise field if the patient is capable of steady fixation ( 29). Based on neuro- radiologic findings, it has been suggested that, in postgeniculate lesions, involvement of the primary visual cortex contributes to perception of scotoma by use of NFC ( 30). At least in prechiasmatic lesions, filling- in appears to occur more frequently with Amsler grid testing than with NFC ( 30). We believe this may be related to differences in background textures used with these techniques. The suggestion is supported by findings from studies with artificial scotomata, showing that filling- in of homoge­neous areas by the twinkling of the surrounding area can be prevented by flickering the background at a frequency of 1.9 Hz ( 23). The duration of damage may also be a critical factor when assessing field defects with such techniques ( 3,31). DISSOCIATION BETWEEN PERCEIVED AND UNPERCEIVED FIELD DEFECTS Changes in visual perception resulting from filling- in cause a dissociation between perceived and unperceived defects in the visual field. This may be one of the most important forms of dissociation in visual function. It oc­curs frequently and produces marked effects on daily living activities. In clinical practice, a simple technique can be used to detect the dissociation and also to allow the patient to understand the nature and extent of the phenomenon. Named the double Amsler grid test, its purpose is to plot first the subjective appearance of the defect, then its true extent ( 10,16). The first part of the procedure should be performed in the usual way for an Amsler grid test, by asking the patient to define the subjective appearance of the defect. The second part should be conducted using a small tangent screen- type stimulus, which can be made from a Q- tip. Perceptual filling- in is demonstrated most J Neuro- Ophthalmol, Vol. 19, No. 2, 1999 88 A. B. SAFRAN AND T. LANDIS easily when the same chart is used to define both the actual and the perceived borders of the defects. Acknowledgment: Supported in part by the Swiss National Fund for Scientific Research, grants 3200- 049594.96 and 4035- 0440 ( PNR38). REFERENCES 1. Gilbert CD, Das A, Ito M, Kapadia M, Weitheimer G. Spatial integration and cortical dynamics. Proc Natl Acad Sci USA 1996; 93: 615- 22. 2. Gilbert CD. Plasticity in visual perception and physiology. Curr Opin Neurobiol 1996; 6: 269- T' 4. 3. Safran AB, Landis Th. Plasticity in the adult visual cortex. Impli­cations for the diagnosis of visual field defects and visual reha­bilitation. Curr Opin Ophthalmol 1996; 7: 53- 64. 4. Ramachandran VS. Blind spots. Sci Am 1992; 266: 85- 91. 5. Safran AB, Halfon A, Safran E, Mermoud C. Angioscotomata and morphological features of related vessels in automated perimetry. Br J Ophthalmol 1995; 79: 118- 24. 6. Gassel MM, Williams D. Visual function in patients with hom­onymous hemianopia: III. The completion phenomenon, insight and attitude to the defect, and visual functional efficiency. Brain 1963; 86: 229- 60. 7. Gerrits HJM, Timmerman GJMEN. The filling- in process in pa­tients with retinal scotomata. Vision Res 1969; 9: 439- 42. 8. Schuchard RA. Validity and interpretation of Amsler grid reports. Arch Ophthalmol 1993; 111: 776- 80. 9. Achard OA, Safran AB, Duret FC, Ragama E. Role of the comple­tion phenomenon in the evaluation of Amsler grid results. Am J Ophthalmol 1995; 120: 322- 9. 10. Safran AB, Landis T. A simple technique for routine evaluation of the filling- in phenomenon in clinical practice. Neuro- Ophthalmol 1996; 16( Suppl 149): 306. 11. Ramachandran VS, Gregory RL. Perceptual filling- in of artificially induced scotomas in human vision. Nature 1991; 350: 699- 702. 12. Safran AB, Landis Th. The vanishing of the sun. A manifestation of cortical plasticity. Surv Ophthalmol 1998,42: 449- 52. 13. Troxler D. Uber das Verschwinden gegebener Geganstande inner-halb unseres Gesichtskreise. In: Himly K, Schmidt A, eds. Oph­thalmologic Bibliothek. Jena, Germany: Springer; 1804: 431- 573. 14. Dunlap K. Light spot adaptation. Am J Physiol 1921; 55: 201- 11. 15. Clarke FJJ, Belcher SJ. On the localization of Troxler's effect in the visual pathway. Vision Res 1962; 2: 53- 68. 16. Safran AB, Duret F, Issenhut M, Landis T. Dissociation between perceived and unperceived defects in the visual field, demonstrated by a double Amsler grid test. The " filling- in" phenomenon revis­ited. In: Mills RP, Heijl A, eds. Perimetry update 1996/ 1997. Am­sterdam: Kugler, 1997: 241- 2. 17. Das A, Gilbert CD. Long- range horizontal connections and their role in cortical reorganization revealed by optical recording of cat primary visual cortex. Nature 1995; 375: 780- 4. 18. Darian- Smith C, Gilbert CD. Axonal sprouting accompanies func­tional reorganization in adult cat striate cortex. Nature 1994; 368: 737^ 10. 19. Kaas JH, Krubitzer LA, Chino YM, Langston AL, Polley EH, Blair N. Reorganization of retinotopic cortical maps in adults mammals after lesions of the retina. Science 1990; 248: 229- 31. 20. Murakami I, Komatsu H, Kinita M. Perceptual filling- in at the scotoma following a monocular retinal lesion in the monkey. Vis Neurosci 1997; 14: 89- 101. 21. Pettet MW, Gilbert CD. Dynamic changes in receptive field size in cat primary visual cortex. Proc Natl Acad Sci USA 1992; 89: 8366- 70. 22. Darian- Smith C, Gilbert CD. Topographic reorganization in the striate cortex of the adult cat and monkey is cortically mediated. J Neurosci 1995; 15: 1631^ 17. 23. De Weerd P, Gattass R, Desimone R, Ungerleider LG. Responses of cells in monkey visual cortex during perceptual filling- in of an artificial scotoma. Nature 1995; 377: 731^ 1. 24. DeAngelis GC, Anzai A, Ohzawa I, Freeman RD. Proc Natl Acad USA 1995; 92: 9682- 6. 25. Chapman B, Stone LS. Turning a blind eye to cortical receptive fields. Neuron 1996; 16: 9- 12. 26. Paradiso MA, Hahn S. Filling- in percepts produced by luminance modulation. Vision Res 1996; 36: 2657- 63. 27. Wu WA, Coleson LC, Holbrook J, Jabs DA, for the Studies of Ocular Complications of AIDS Research Group. Measuring visual function and quality of life in patients with cytomegalovirus reti­nitis: Development of a questionnaire. Arch Ophthalmol 1996; 114: 841- 7. 28. Safran AB. Unperceived visual field defect. Arch Ophthalmol 1997; 115: 686- 7. 29. Aulhorn EL, Kost G. Noise- field campimetry: a new perimetric method ( snow campimetry). In: Heijl A, ed. Perimetry update 1988/ 1989. Amsterdam: Kugler & Ghedini, 1989: 331- 6. 30. Kolb M, Petersen D, Schiefer U, Kolb R, Skalej M. Scotoma perception in white- noise- field campimetry and postchiasmal vi­sual pathway lesions. Ger J Ophthalmol 1995; 4: 228- 33. 31. Schiefer U, Kost G, Aulhorn E. Rauschfeld- Untersuchunger gebnisse mit dem Tiibinger Elektronik Kampimeter ( TEC): Ein Vergleich mit herkb'mmlichen perimetrischen Verfahren. Fortschr Ophthalmol 1990; 87: 508- 15. J Neuro- Ophthalmol, Vol. 19, No. 2, 1999