Functional compensation in the lateral suprasylvian visual area following bilateral visual cortex damage in kittens

Summary Previous studies have shown that functional compensation is present in the cat's posteromedial lateral suprasylvian (PMLS) area of cortex after damage to areas 17, 18, and 19 (visual cortex) early in life but not after damage in adults. These studies all have investigated animals with a unilateral visual cortex lesion, whereas all behavioral studies of compensation for early visual cortex damage have investigated animals with a bilateral lesion. In the present experiment, we investigated whether functional compensation also is present in PMLS cortex after a bilateral visual cortex lesion early in life. We recorded from single neurons in the PMLS cortex of adult cats that had received a bilateral lesion of areas 17, 18, and 19 on the day of birth or at 8 weeks of age. We found that PMLS cells in both groups of cats had functional compensation (normal direction selectivity and ocular dominance) similar to that seen after a unilateral lesion at the same ages. These results are consistent with the hypothesis that PMLS cortex is involved in the behavioral compensation seen after early visual cortex damage. In addition, the results indicate that inputs from contralateral visual cortex are not necessary for the development of functional compensation seen in PMLS cortex.     

Effects of age and visual experience on [3H] MK801 binding to NMDA receptors in the kitten visual cortex

Summary We have investigated the postnatal development of N-methyl-D-aspartate (NMDA) receptors in kitten neocortex by measuring the density of [³H] MK801 binding sites. NMDA receptor density in area 17 markedly increases between postnatal day 7 and 35, and then remains at a high level into adulthood. Similar temporal changes were observed in area 6. A brief period of binocular deprivation did not alter the density of [³H] MK801 binding sites in either brain region. However, dark rearing kittens from birth resulted in a significant decrease after 35 and 120 days in both area 17 and area 6. These results demonstrate that the density of NMDA receptors does not effectively predict the capacity for visual cortical plasticity in kitten neocortex. In addition, these results show that dark rearing produces significant effects in areas of the kitten neocortex not normally associated with visual function.     

Ocular dominance in kitten cortex: Induced changes of single cells while they are recorded

Summary We have monitored extracellularly individual neurons in the striate cortex of 4-week-old unparalyzed kittens with the aim of changing the ocular dominance of these cells during recording. To do this, we elicited conjugate eye movements using a bipolar stimulating electrode positioned in the internal medullary lamina (IML) of the thalamus. During electrical stimulation of this region, one eye was occluded and the other was visually activated with optimal stimuli. Receptive fields were studied subjectively and objectively and relative response strengths were assessed. Of 42 cells studied in detail, 62% underwent changes in ocular dominance following conditioning periods of, generally, 15–20 min. Control experiments suggest that this plasticity is: agerelated; requires both visual stimulation and activation of pathways associated with eye movement; and does not appear to be caused solely by increased arousal levels.Supported by grant EY01175 and Research Career Development Award EY00092 from the US National Eye InstituteOn leave from Department of Physiology Kanazawa University Medical School, Kanazawa 920, Japan     

Emergence of callosally projecting neurons with stellate morphology in the visual cortex of the kitten

Summary Callosally projecting neurons in areas 17 and 18 of the adult cat can be classified into two types on the basis of their dendritic morphology: pyramidal and stellate cells. The latter are nearly exclusively of the spinous type and are predominantly located in upper layer IV. Retrograde transport of the carbocyanine dye DiI, applied to the corpus callosum, showed that, up to P6, all callosally projecting neurons resemble pyramids in the possession of an apical dendrite reaching layer I. At P10, however, callosally projecting neurons with stellate morphology were found. A study was designed to distinguish whether these neurons are late in extending their axons to the corpus callosum or, alternatively, have transient apical dendrites. To this end, callosally projecting neurons were retrogradely labeled by fluorescent beads injected in areas 17 and 18 at P1–P3 and then either relabeled with DiI applied to the corpus callosum at P10 or intracellularly injected with Lucifer Yellow at P57. Double-labeled stellate and pyramidal cells were found in similar proportions to those found for the total, single-labeled population of callosally projecting neurons. It is therefore concluded that callosally projecting spiny stellate cells initially possess an apical dendrite and a pyramidal morphology. At P6, i.e. close to the time when stellate cells appear, layer IV neurons with an atrophic apical dendrite were found, suggestive of an apical dendrite in the process of being eliminated.     

Effects of luminance and flicker on ocular dominance shift in kitten visual cortex

Summary We raised monocularly deprived kittens in visual environments with low level illumination that was either steady or flickering. With steady scotopic luminance ocular dominance shifted as it does in normal photopic lighting. In flickering light with an average frequency of 2 Hz there was virtually no ocular dominance shift, while in flickering light averaging 0.1 Hz there was a significant shift. Recordings from the 2 Hz flicker-reared were similar to the dark-reared recordings. The flickering illumination was produced in one case by a high contrast-low brightness TV near the cage, and in another case, by a low voltage incandescent bulb driven by a pseudo-random sequence generator. This circuit delivered either a maximum ON time of 1.7 s or a maximum of 40 s, for the 2 Hz and 0.1 Hz respectively. Both the TV and flickering bulb produced average illumination comparable to the dim (0.01 cd/ m²) steady scotopic illumination. We conclude that dim flickering light is not a sufficient stimulus for promoting ocular dominance shift in kittens in the critical period unless the flicker rate approaches 0.1 Hz. Furthermore results from the TV rearing suggest that flicker may be capable of preventing an ocular dominance shift expected from a concurrent steady low light level background.This work was supported by ONR Contract N00014-81-K-0136. Flaxedil was supplied by Lederle Laboratories. We thank K. Cullen, M. Sutter, J. Kape and M. Motuz for technical assistance. L.N. Cooper and J.P. Rauschecker provided helpful comments.     

Instructive changes in the kitten's visual cortex and their limitation

Summary Kittens were reared wearing masks that contained strong cylindrical lenses, which allowed them to see only contours of one orientation. Selective exposure was alternated between the two eyes on successive days, while each time the other eye was covered by the mask. The total duration of exposure was different in the two eyes, amounting to 50 h for one eye and at least 100 h for the other. This resulted in asymmetric distributions of ocular dominance: neurones preferring precisely the experienced orientation favoured the eye with longer exposure, whereas neurones preferring oblique orientations adjacent to the experienced one surprisingly were dominated by the eye with shorter exposure. Thus neurones originally belonging to this group and dominated by the longer exposed eye must have tuned their orientation preference to the experienced orientation as a result of the longer exposure. Such instructive changes seem to be limited by the original response borders of the cortical neurones, as predicted by Hebb's rules for synaptic plasticity.Supported by the Deutsche Forschungsgemeinschaft, SFB 50 (Kybernetik)A part of this study was done at the Max-Planck-Institute for Psychiatry, Dept. of Neurophysiology, D-8000 Munich, Federal Republic of Germany     

How complete is physiological compensation in extrastriate cortex after visual cortex damage in kittens?

Summary Previous studies indicate that neurons in the cat's posteromedial lateral suprasylvian (PMLS) visual area of cortex show physiological compensation after neonatal but not adult damage to areas 17, 18, and 19 of the visual cortex (collectively, VC). Thus, VC damage in adults produces a loss of direction selectivity and a decrease in response to the ipsilateral eye among PMLS cells, but these changes are not seen in adult cats that received VC damage as kittens. This represents compensation for early VC damage in the sense that PMLS neurons develop properties they would have had if there had been no brain damage. However, this is only a partial compensation for the effects of VC damage. A full compensation would involve development of properties of the VC cells that were removed in the damage. The present study investigated whether this type of compensation occurs for detailed spatial- and temporal-frequency processing. Single-cell recordings were made in PMLS cortex of adult cats that had received a VC lesion on the day of birth or at 8 weeks of age. Responses to sine-wave gratings that varied in spatial frequency, contrast, and temporal frequency were assessed quantitatively. We found that the spatial- and temporal-frequency processing of PMLS cells in adult cats that had neonatal VC damage were not significantly different from PMLS cells in normal cats. Therefore, there was no evidence that PMLS cells can compensate for VC damage by developing properties that are better than normal and like those of the striate cortex cells that were damaged.We also assessed the effects of long-term VC damage in adult cats to determine whether the normal properties seen in cats with neonatal VC damage represent a compensation for abnormalities in PMLS cortex present after adult damage. In a previous study, we found that acute VC damage in adult cats has small but reliable effects on maximal response amplitude, maximal contrast sensitivity, and spatial resolution (Guido et al. 1990b). In the present study, we found that long-term VC damage in adult cats does not increase these abnormalities as a result of secondary degenerative changes. In fact, the minor abnormalities that were present after an acute VC lesion were virtually absent following a long-term adult lesion, perhaps because they were due to transient traumatic effects. Therefore, there was little evidence for abnormalities in spatial- or temporal-frequency processing following long-term adult VC damage for which PMLS cells might show compensation following long-term neonatal damage.Our results thus indicate that there is little or no difference in the spatial- or temporal-frequency processing of PMLS cells in normal cats and cats with long-term VC damage received early in life or as adults. These findings are discussed in relation to the inputs to PMLS cortex and to the behavioral abilities of cats with VC damage at different ages. The implications for under-standing the role of lateral suprasylvian visual cortex in behavioral recovery from VC damage is considered.     

Development of the kitten visual cortex depends on the relationship between the plane of eye movements and visual inputs

Summary 1. Previous experiments have demonstrated that eye movements, acting through the extraocular muscle (EOM) proprioceptive afferents, are necessary for the development of orientation selectivity in the cells of the kitten visual cortex. New experiments were carried out to study the effect of the plane of eye movements on the preferred orientation acquired by the visual cortical cells. 2. Darkreared (DR) kittens were operated on at 5–6 weeks of age. In the first series of experiments, 4 out of the 6 EOMs were removed bilaterally in such a way that both eyes could only move in a single plane, either vertical or horizontal. In the second series of experiments, the same operation was performed on one eye which was also sutured shut and, on the other side, the EOM were deafferented by intracranial section of the ophthalmic branch of Vth nerve and the eye left open. 3. 1–4 days after surgery the kittens were given 6 h of visual experience and 12 h later were prepared for visual cell recording in Area 17. 4. In kittens of the first series: orientation selectivity developed in the majority (60–65%) of visual cells, most of which encoded horizontal orientations when the eyes had moved in the vertical plane and vertical orientations when the eyes had moved in the horizontal plane. These results show that the plane of eye movements during early visual experience influences the distribution of preferred orientations with an orthogonal relation. Ocular dominance histograms were strabismic like. 5. In kittens of the second series: orientation selectivity developed in 40–50% of cells, about half of which were tuned for the orientation orthogonal to the direction of movement of the occluded eye, as in experiment I. The seeing, deafferented eye, presumably would have sent normal visual inputs centrally, corresponding to displacements on the retina in every direction since the ocular motility of that eye had not been disturbed. However, proprioceptive information about its movements was suppressed. As only some of the EOMs of the occluded eye were still present and connected, the conclusion is that the observed influence of the plane of eye movements acts through the proprioceptive afferents. The ocular dominance histograms showed: 1) a powerful change in favour of the seeing eye after only 6 h of monocular visual experience; 2) a larger effect of monocular visual experience in the hemisphere contralateral to the seeing eye; 3) a linkage between acquisition of orientation selectivity and shift in ocular dominance. 6. Our results suggest that normal development of orientation selectivity in visual cortical cells results from the close association of visual and EOM afferent inputs. It is suggested that these two signals must occur with a precise temporal relationship.     

Blockade of serotonin-2C receptors by mesulergine reduces ocular dominance plasticity in kitten visual cortex

 We have investigated the role of serotonin-2C (5-HT_2C) receptors in modulation of ocular dominance plasticity in kitten visual cortex. A small quantity of the 5-HT_2C receptor blocker, mesulergine, was infused into the visual cortex of one hemisphere of 5- to 7-week-old kittens using osmotic minipumps, while the control hemisphere received vehicle solution. At the same time, one eyelid of the experimental animals was sutured shut. The ocular dominance distributions in the visual cortex (area 17) were assessed using extracellular recording methods after 1 week of combined mesulergine infusion and monocular deprivation. We found that the majority of the neurons remained binocularly responsive in the mesulergine-treated hemisphere, while most of the neurons recorded were either unresponsive or only weakly responsive to the deprived eye in the control hemisphere. Local infusion of mesulergine into the kitten visual cortex thus reduced the shift of ocular dominance that normally occurs in animals of these ages following monocular deprivation. The blocking effect seems to be distance-dependent and therefore dose-dependent: the farther away the recording sites were from the injection site, the fewer binocularly responsive cells were found. These results are relevant to previous findings indicating transient overexpression of 5-HT_2C receptor in visual cortex of kittens at these ages. The data suggest that the 5-HT_2C receptor system may be involved in the formation and modification of ocular dominance columns in the developing visual cortex. Received: 11 December 1995 / Accepted: 9 June 1996     

Central core control of developmental plasticity in the kitten visual cortex: I. Diencephalic lesions

Summary In five, dark-reared, 4-week-old kittens the posterior two thirds of the corpus callosum were split, and a lesion comprising the intralaminar nuclei was made of the left medial thalamic complex. In addition, the right eye was closed by suture. Postoperatively, the kittens showed abnormal orienting responses, neglecting visual stimuli presented in the hemifield contralateral to the side of the lesion. Sudden changes in light, sound, or somatosensory stimulation elicited orienting responses that all tended toward the side of the lesion. These massive symptoms faded within a few weeks but the kittens continued to neglect visual stimuli in the hemifield contralateral to the lesion when a second stimulus was presented simultaneously in the other hemifield. Electrophysiologic analysis of the visual cortex, performed after the end of the critical period, revealed marked interhemispheric differences. In the visual cortex of the normal hemisphere most neurons were monocular and responded exclusively to stimulation of the open eye, but otherwise had normal receptive field properties. In the visual cortex of the hemisphere containing the thalamic lesion, the majority of the neurons remained binocular. In addition, the selectivity for stimulus orientation and the vigor of responses to optimally aligned stimuli were subnormal on this side. Thus, the same retinal signals, which in the control hemisphere suppressed the pathways from the deprived eye and supported the development of normal receptive fields, failed to do either in the hemisphere containing the thalamic lesion. Apparently, experience-dependent changes in the visual cortex require both retinal stimulation and the functioning of diencephalic structures which modulate cortical excitability and control selective attention.     

Experience-dependent modifications of kitten striate cortex are not prevented by thalamic lesions that include the intralaminar nuclei

Summary It has been shown previously that surgical lesions of the antero-medial thalamus interfere with ocular dominance modifications that normally result from monocular deprivation in young kittens (Singer 1982). The aim of the present study was to determine whether this effect was due specifically to the destruction of the visual cortical projections of the anterior intralaminar nuclei. We report here that large excitotoxin lesions of the anterior dorsal thalamus have no effect on the cortical response to monocular deprivation. These data indicate that the intralaminar projection is not essential for ocular dominance plasticity.MFB was supported in part by U. S. Office of Naval Research contract N000-14-81-K-0041.     

Grating detection and visual acuity after lesions of striate cortex in hooded rats

Summary The ability of rats to detect high-contrast square-wave gratings over a range of spatial frequencies was measured before and after ablation of striate cortex. The animals relearnt to detect low-frequency gratings very quickly after operation, and their acuity was reduced from 1.0 c/deg to about 0.7 c/deg. These effects were in striking contrast to those produced by larger posterior cortical ablations, which included both striate and prestriate cortex (Dean 1978); after the larger lesions, rats required many weeks of retraining to detect even low-frequency gratings and their acuity was reduced to 0.3 c/deg. The difference in the effects of the two lesions suggested that the rats with striate ablation were using information about spatial contrast that was relayed either by spared remnants of the geniculo-cortical pathway, or by the pathway from superior colliculus to prestriate cortex via the lateral posterior nucleus. To try and distinguish between these possibilities, the destriate rats were given a further lesion of the superior colliculus. This second lesion severely disrupted contrast detection: the animals made about as many errors as rats with large posterior cortical removal in relearning to detect a low-frequency grating, which is about 20 to 30 times as many as after either striate cortex or superior colliculus lesions alone. This result suggests that rats, like other mammals, can use spatial information conveyed in the tectocortical path when striate cortex has been destroyed.This work was supported by MRC Grant G987/429/N     

Modification of single neurons in the kitten's visual cortex after brief periods of monocular visual experience

Summary Kittens were deprived of form vision by suturing the lids of both eyes, except for a brief period (1, 6 or 20 hours) on the 29th day when the right eye was opened. 6 space and 20 hours of monocular vision produced a distinct shift in the ocular dominance of visual cortical neurons towards the experienced eye, and an increase in the proportion of cells with obvious orientation selectivity. These modifications in the visual cortex were enhanced by a period of consolidation: they were somewhat less obvious if recordings were taken immediately after the exposure but were complete 2 days later. Although remarkably little visual experience was needed for these changes, the results contrast with the effects of rearing in an environment of vertical stripes, where only 1 hour of exposure produces much more striking effects. A normal visual environment may have a less powerful organizing influence on cortical neurons than such an environment containing only one orientation.On leave from Department of Psychology, Pomona College, Claremont, California 91711 (USA).     

A quantitative study of the classification and stability of ocular dominance in the cat's visual cortex

Summary We recorded from single cells in the cat's visual cortex to quantitatively evaluate (1) the reliability of subjective assessments of ocular dominance (101 cells) and (2) the stability of ocular dominance over time (25 cells). We found that the correlation between subjective and objective measures of this variable was poorer than expected, and was worst for cells with low overall response strengths. This result appears to reflect variability in the subjective assessment procedure. For the second part of the study, we recorded from single cortical cells of 5-week-old kittens, and made repeated objective measurements of ocular dominance over time. Twenty-four of the twenty-five cells examined were quite stable in ocular dominance for periods so long as 8 h. One unit was encountered which showed substantial progressive shifts in ocular dominance over time.Supported by grant EY01175 and Research Career Development Award EY00029 from the US National Eye Institute to R.D. Freeman     

Reemergence of ocular dominance plasticity during recovery from the effects of propranolol infused in kitten visual cortex

Summary We wanted to know whether ocular dominance plasticity can increase under the condition in which the number of available adrenoreceptors is expected to increase within kitten visual cortex. We adopted a paradigm in which monocular lid suture was carried out some time after the termination of direct infusion of the cortex with a adrenoreceptor antagonist. A significant change in ocular dominance was obtained as shown by a decrease in binocular cortical neurons, when time interval between the end of the d,l-propranolol infusion and the start of monocular deprivation was one week. With a 3-week interval (the longest tested), an even greater change in ocular dominance was evident. This consisted of a marked decrease in binocular neurons and a shift in ocular dominance toward the nondeprived eye. In a control study an inert stereoisomer, d-propranolol, did not block the ocular dominance shift. These results were interpreted as suggesting that the level of ocular dominance plasticity becomes high in parallel to an expected increase in availability of adrenoreceptors for endogenous noradrenaline (NA). We next asked whether it is possible to accelerate or decelerate the naturally occurring recovery of ocular dominance plasticity. When either NA or tunicamycin (an inhibitor of protein glycosylation) was infused into the same cortical area immediately after the end of the propranolol infusion, opposite effects were observed: exogenous NA accelerated the recovery of the shift in ocular dominance and tunicamycin suppressed it. When tunicamycin infusion was delayed by one week, however, its suppressive effect was negligible. Thus, the restoration of ocular dominance plasticity seems to occur in parallel to an increase in the availability of adrenoreceptors for endogenous as well as exogenous NA.     

TMS modulation of visual and auditory processing in the posterior parietal cortex

The activity patterns of a neuronal network originate from the intrinsic properties and synaptic interactions of the constituent neurons. Our recent studies support this view, showing that the discharge of a single frog retina ganglion cell brings an elementary neuronal network of the tectum (tectum column) to a suprathreshold activity of two distinct levels that are related to the activation of the slow L-type calcium current in dendrites of the recurrent pear-shaped neurons (lower level) and the NMDA receptors in neurons (higher level) of the tectum column. We show in the present study that the dendritic slow L-type calcium current is necessary for the NMDA receptor activation in the tectum column. A small decrease of this current prevents the NMDA receptor activation and, hence, the transition of the network to the higher activity level, at which the efferent neuron of the network fires. So, the activity of the frog tectum column can be effectively controlled through the intrinsic properties of the recurrent pear-shaped neurons of the column.     

Substantial reduction of noradrenaline in kitten visual cortex by intraventricular injections of 6-hydroxydopamine does not always prevent ocular dominance shifts after monocular deprivation

Summary Ten kittens had cannulas inserted into their lateral ventricles for daily injections of 6-hydroxydopamine (6-OHDA). At 5–6 weeks of age one eye was sutured shut, and one week later recordings were made from the visual cortex to assay the ocular dominance of a sample of cells. In six kittens the injections of 6-OHDA were continued until the day before recording, while in four kittens the injections were stopped around the time of eye suture, on the assumption that continued injections of 6-OHDA over several days has effects that are not specific to the noradrenaline (NA) system and that the two procedures might show different results. In all animals the concentration of NA in the visual cortex near the site of recording was reduced by approximately 90%. In all animals the ocular dominance histograms recorded from the visual cortex were shifted so that the majority of cells (83 ± 13%) were dominated by the open eye. There were no substantial differences between the two groups of experimental animals or between the experimental animals and two control animals that had cannulas implanted and ascorbate alone injected without 6-OHDA. We conclude that the concentration of NA in the visual cortex can be reduced substantially by injections of 6-OHDA into the lateral ventricle without preventing the shift in ocular dominance that usually occurs after suturing shut the eyelids of one eye.     

Receptor subtypes involved in callosally-induced postsynaptic potentials in rat frontal agranular cortex in vitro

Summary A slice preparation of rat frontal agranular cortex preserving commissural inputs has been used for intracellular recording from layer V pyramidal cells, in order to characterize the synaptic potentials induced by stimulation of the corpus callosum and to reveal the subtypes of amino acid receptors involved. Stimulation of the corpus callosum induced EPSPs followed by early IPSPs with a peak latency of 30 ± 2 ms and late IPSPs with a peak latency of 185 ± 18 ms. Reversal potentials for early and late IPSPs were –75 ± 5 mV (early) and –96 ± 5 mV (late). Late IPSPs were more dependent on extracellular K⁺ concentration. The early IPSPs were blocked by GABA_A antagonists, bicuculline and picrotoxin, whereas the late IPSPs were reduced by the GABA_B antagonist, phaclofen. CNQX (6-cyano-7-nitroquinoxaline-2,3-dione), an antagonist of non-NMDA (N-methyl-D-aspartate) receptors, suppressed both EPSPs and late IPSPs at 5 µM. Early IPSPs remained at this concentration but were suppressed by 20 µM CNQX. In Mg²⁺-free solution, EPSPs were larger and more prolonged than in control solution. These enhanced EPSPs persisted after 5 to 20 µM CNQX, but were reduced in amplitude, and their onset was delayed by 3.6 ± 0.8 ms. The remaining EPSPs were suppressed by 50 µM APV (DL-2-amino-5-phosphono-valeric acid), an antagonist of NMDA receptors. In Mg²⁺-free solution containing 5 to 20 µM CNQX, the late IPSPs were not diminished. The remaining late IPSPs were suppressed by APV or by phaclofen. By contrast, the amplitude of early IPSPs was not affected by APV in Mg²⁺-free solution containing 5 µM CNQX. These results show that stimulation of the corpus callosum can induce GABA_A and GABA_B dependent IPSPs and NMDA and non-NMDA dependent excitation. It is suggested that these four types of amino acid-based transmission are conveyed by intracortical pathways with different characteristics.     

Local cortical lesions abolish lateral inhibition at direction selective cells in cat visual cortex

Summary Many cells in the cat visual cortex display a strong selectivity for the direction of motion of an optimally oriented stimulus. Postsynaptic inhibition has been suggested to generate this direction selectivity in simple cells, but the intracortical pathways involved have not been identified. While continuously recording from simple cells in layers 4 and 6, we have inactivated the superficial cortical layers in small regions 0.4–2.5 mm from the cortical column under study by using heat lesions, localized cooling or -aminobutyric acid (GABA) microiontophoresis. When inactivation affected cortical regions retinotopically representing motion in the non-preferred direction towards the receptive field, the responses to movement in this direction increased, and the recorded cells lost direction selectivity due to loss of inhibition. Our results indicate that direction selectivity of simple cells involves asymmetric inhibition of predictable cortical topography.     

Effect of contrast on spatial frequency tuning of neurones in area 17 of cat's visual cortex

Summary Previous investigations have revealed that perceived spatial frequency of gratings rises as contrast is lowered. In order to account for this finding it has been postulated that the spatial frequency which produces the maximum response from cortical neurones decreases with contrast. We have examined this hypothesis by determining optimal spatial frequency at 3–5 different contrast levels for 37 neurones in the cat striate cortex. For the complete sample no systematic changes in optimal spatial frequency was observed. However, a shift in the predicted direction was found for cells tuned to high spatial frequencies.