- Email: [email protected]
Binocularity and spatial frequency dependence of calcarine activation in two types of amblyopia
Neuroscience Research 40 (2001) 147– 153 www.elsevier.com/locate/neures
Binocularity and spatial frequency dependence of calcarine activation in two types of amblyopia Kyoung-Min Lee a,e,f,*, Soo-Hwa Lee a, Na-Young Kim e, Chai-Youn Kim e, Jung-Woo Sohn a, Mi Young Choi g, Dong Gyu Choi h, Jung-Min Hwang b, Ki Ho Park b, Dong Soo Lee c,e,f, Young Suk Yu b, Kee Hyun Chang d,f a
Department of Neurology, Seoul National Uni6ersity, Seoul, 110 -744, South Korea Department of Ophthalmology, Seoul National Uni6ersity, Seoul, 110 -744, South Korea c Department of Nuclear Medicine, Seoul National Uni6ersity, Seoul, 110 -744, South Korea d Department of Radiology, Seoul National Uni6ersity, Seoul, 110 -744, South Korea e Interdisciplinary Program for Cogniti6e Science, Seoul National Uni6ersity, Seoul, 110 -744, South Korea f BK21 Human Life Science Di6ision, Seoul National Uni6ersity, Seoul, 110 -744, South Korea g Department of Ophthalmology, Chongbuk National Uni6ersity, Chongju, South Korea h Department of Ophthalmology, Hallym Uni6ersity, Seoul, South Korea b
Received 28 June 2000; accepted 16 February 2001
Abstract Objecti6e and Background: Strabismus and anisometropia early in life frequently causes monocular amblyopia. Activation of the visual cortex is compared between the two types of amblyopia to elucidate differences in the pathogenetic mechanism of the disease. Methods: Using an EPI gradient echo sequence in 1.5T MRI, calcarine activation by monocular viewing of checkerboard patterns with reversal was examined in terms of binocularity of the activation and dependence on the spatial frequency of the stimuli. Results: First, the proportion of voxels activated by both normal and amblyopic eye monocular stimulations is lower in the strabismic group than in the anisometropic group. Second, the activation by higher-spatial-frequency stimuli is reduced in the anisometropic group, but not in the strabismic group. Conclusions: These findings from the human visual cortex are consistent with the view proposed based on animal research that the loss of binocular interaction and the undersampling of high-spatial-frequency components of visual stimuli are each one of the underlying changes in strabismic and anisometropic amblyopia, respectively. © 2001 Elsevier Science Ireland Ltd and the Japan Neuroscience Society. All rights reserved. Keywords: Amblyopia; Strabismus; Anisometropia; Visual cortex; Brain activation; fMRI
1. Introduction Development of the visual function critically depends on visual experience early in life (Daw, 1995). The early visual experience is deranged under such conditions as strabismus, in which the visual axes are misaligned, and anisometropia, in which the refractive state is significantly different between the two eyes. The two conditions often lead to monocular amblyopia without a macroscopic lesion along the visual pathway. The
* Corresponding author. Tel.: + 82-2-7602985; fax: 36727553. E-mail address: [email protected] (K.-M. Lee).
+ 82-2-
pathogenetic mechanism of amblyopia may differ between the two conditions (Blakemore and VitalDurand, 1992), since strabismus prevents concordant stimulation of binocularly corresponding points in the two retinae, whereas anisometropia imposes constant blurring of one retinal image relative to the other. Clinical studies on human amblyopics confirmed that the two conditions lead to different forms of amblyopia (Hess et al., 1978; Levi and Klein, 1982), although this issue remains controversial (McKee et al., 1992; Movshon et al., 1996). From experiments using a number of animal models, it has been suggested that strabismic amblyopia results from the breakdown of binocular interaction (Crawford and von Noorden, 1979; Wiesel, 1982; Harrad et al., 1996; Kiorpes et al., 1998). In
0168-0102/01/$ - see front matter © 2001 Elsevier Science Ireland Ltd and the Japan Neuroscience Society. All rights reserved. PII: S 0 1 6 8 - 0 1 0 2 ( 0 1 ) 0 0 2 2 0 - 6
K.-M. Lee et al. / Neuroscience Research 40 (2001) 147–153
148
contrast, reduction in the spatial resolution of neurons with preferred spatial frequency in the higher range or the loss of such neurons may underlie anisometropic amblyopia (Hess et al., 1978; Kiorpes et al., 1987; Movshon et al., 1987). Thus, in this study, activation of the human visual cortex around the calcarine sulci observed using functional MRI (fMRI) is compared between the two types of amblyopia, in terms of binocularity of the activation and its response characteristics to various spatial frequencies.
projector controlled by a personal computer. The frequency of checkerboard reversal was 8 Hz (i.e. 4 cycles per second, since one cycle contained two reversals). The distance from the inside of the scanner to the screen was 4 m. The extent of the stimulated visual field was 16 and 10 degrees in horizontal and vertical dimensions, respectively, in the middle of which a crosshair was shown for fixation. Patients were asked to view the visual stimuli monocularly under the following three viewing conditions: (1) with the amblyopic eye, (2) with the naked normal eye, and (3) with the normal eye defocused to match the visual acuity with that of the amblyopic eye by a plus lens. The strength of the lens was selected before the MR scanning such that the visual acuity of the normal eye measured with the lens in front was the same as that of the amblyopic eye at a 6-m viewing distance. Monocular viewing during MRI scanning was achieved by placing a black plate in front of one eye onto a custom-made MRI-compatible goggle frame. Defocusing, when needed, was similarly achieved by placing the defocusing lens in front of the viewing eye. For switching the side of the blocked eye, the scanner table was slid out just enough for an experimenter to move the blocking plate to the other side of the goggle frame without touching the patient’s head on the headrest. Each patient was subjected to three MR scan runs in total, one for each viewing condition. The sequence of the viewing conditions was pseudorandomized across patients. An MR scan run for each viewing condition lasted 180 s, divided into six 30-s periods. No visual stimulus except a center crosshair was presented during the first and sixth epochs, regarded as the baseline epochs. During the middle four activation epochs, the checker-
2. Methods
2.1. Patients Eleven amblyopic patients (six anisometropic, aged 8 – 18 years old, and five strabismic patients, aged 5–32 years old) and three normal subjects (aged 18– 22 years old) participated in this study after giving their informed consent. Clinical data on the patients and normal subjects are summarized in Table 1. Fourteen patients were originally recruited and subjected to MR scanning, but in three of them (one strabismic and two anisometropic), the data contained too much movement artifacts for a valid analysis and were therefore excluded.
2.2. Visual stimulus and experimental conditions Visual stimuli were black-and-white checkerboard patterns presented at a spatial frequency of 2, 1, 0.5, or 0.25 cycles per degree of visual angle (i.e. individual check-sizes of 0.25, 0.5, 1, and 2 degrees of visual angle, respectively), back-projected onto a screen by a LCD Table 1 Clinical data on tested amblyopic patients and normal control subjectsa Patient ID
S1 S2 S3 S4 S5 A1 A2 A3 A4 A5 A6 C1 C2 C3 a
Amblyopia
OD OS OS OD OS OD OS OD OD OD OD – – –
Sex
M F M M F M M M M F M M F M
Age (yrs)
5 23 5 8 23 9 17 10 8 16 10 18 22 18
VA OD
OS
0.04 1.50 1.00 0.15 1.00 0.08 1.00 0.10 0.02 0.20 0.10 1.00 1.00 1.50
1.00 0.10 0.10 1.00 0.15 1.00 0.08 1.00 0.90 1.20 1.00 1.00 1.50 1.20
EOM abnormality, prism diopter
Anisometropia
eso, 50 exo, 35 eso, 45 eso, 16 eso, 25 ortho ortho ortho ortho eso, B10 ortho ortho ortho ortho
– – – – – hyperopic hyperopic hyperopic myopic hyperopic hyperopic – – –
VA: visual acuity (best corrected), OD: right eye, OS: left eye, eso: esotropic, exo: exotropic, ortho: orthotropic.
K.-M. Lee et al. / Neuroscience Research 40 (2001) 147–153
Fig. 1. In the left panel, the location of image slices (dotted lines) and the ROI within them (solid portion of the dotted lines) is indicated for a typical case. In the right panel, an example of the ROIs is shown in a slice image with the activation around the calcarine sulci overlayed. See text for details on defining the ROI.
board pattern of four different spatial frequencies was presented. The sequence of the spatial frequencies tested was fixed for each patient, but counter-balanced across patients. The patients were instructed to fixate on the center crosshair at all times during the scanning. Their eye movements were not monitored objectively.
2.3. MR image acquisition T2*-weighted images were obtained using a 1.5T MR scanner equipped with a 5-in. surface coil provided with the scanner (General Electric, Milwaukee). Scan parameters were as follows: gradient echo EPI sequence, TR 3000 ms, TE 60 ms, flip angle 90 degree, 15 3-mm-thick slices without separation, FOV 14 cm, image matrix 128×128 (thus, a voxel size of 1.09× 1.09 ×3 mm). The surface coil was centered at the inion. Image slices were positioned perpendicular to the inter-commissural plane (Fig. 1, left panel). The images were reconstructed off-line, exported to volumes, and aligned for movement correction using the AIR3.0 package (Woods et al., 1992).
2.4. Data analysis 2.4.1. Analysis of binocularity of fMRI response This analysis was carried out within the region of interest (ROI) along the calcarine sulci, which was defined using T1-weighted images that were obtained on the same slice locations as the EPI images (Fig. 1, left panel). The ROIs were drawn on the caudalmost slice that included at least one occipital pole and five consecutive slices rostral to it, i.e. six slices in total and 18 mm in rostro-caudal direction. After identifying the calcarine sulcus on each slice, an intersection between the midline and the line linking the two calcarine sulci on both sides was chosen as the center point. A square ROI was drawn based on the center point; the boundary of the ROI was 1 cm above and below the
149
center point and at 2 cm on both sides from the midline (Fig. 1, right panel). Although we did not attempt functional demarcation among the visual areas, we considered that the ROI most likely included some higher visual areas, such as V2, as well as the primary visual cortex. The voxelwise t-test (PB 0.05, two-tailed, uncorrected) was performed to compare the MR signal between four activation epochs and two baseline epochs. The binocularity of the visual response was then determined using two kinds of analysis. First, whether monocular stimulation through either eye evoked activation was determined for every voxel. A voxel was classified as binocularly activated if the signal intensity significantly increased in both amblyopic-eye- and defocused-normal-eye-viewing conditions. The proportion of such voxels was calculated as the binocular voxel index (BVI) with respect to the total number of voxels that showed activation under at least one of the two conditions, i.e. BVI = No. of voxels in (ODSOS)/No. of voxels in (OD@OS) ×100, where S and @ are the intersection between and the union of OD and OS, or monocular stimulation of the right and left eye, respectively. Second, correlation of activation was assessed by plotting the percent change in the MR signal during amblyopic eye stimulation and that during defocused normal eye stimulation for every voxel. The coefficient of determination (R 2) was then calculated for each patient.
2.4.2. Analysis of spatial-frequency dependence of fMRI response Whether the calcarine activation by a checkerboard pattern of various check sizes is different between the two amblyopic types is examined. SPM96 software (Wellcome Department of Cognitive Neurology, London) was employed to identify the locus of the maximum response within the ROI. The voxel with the maximum activation and 24 neighboring voxels on the same slice were chosen as a region with maximal response. From the maximal activation region, the mean percent change in MR signal was computed between the first baseline epoch and each of four different spatial-frequency activation epochs for each patient. The group average of the results was then compared between the three different viewing conditions using two-tailed t-test.
3. Results
3.1. Difference in binocularity of calcarine acti6ation The total number of voxels activated in any of the three viewing conditions was not different between the two amblyopic groups (453 voxels in the stabismic
150
K.-M. Lee et al. / Neuroscience Research 40 (2001) 147–153
group vs. 513 in the anisometropic group, P \ 0.45, two-tailed t-test). The ratio of the number of activated voxels in contralateral hemisphere to that in hemisphere ipsilateral to the stimulated eye did not reveal a consistent pattern across patients or across patient groups. The proportion of binocularly activated voxels, i.e. BVI was, however, significantly lower in the strabismic amblyopes (mean9standard error, 16.859 2.98%) than in the anisometropic amblyopes (mean9 standard error, 34.9894.60% P B 0.05, two-tailed t-test; Fig. 2). Comparison between the amblyopic eye and the naked normal eye revealed a similar difference (strabismic group BVI mean9 standard error 19.894.86%, and anisometropic group BVI mean9 standard error, 30.2 93.63%) but the difference did not reach statistical significance (P\0.05, two-tailed t-test). For the control group of three normal subjects, BVI was 45.949 3.56%, computed under the two naked-eye-viewing conditions. This value was significantly higher than that of the strabismic group (P B 0.001, two-tailed t-test), but not significantly different from that of the anisometric group (P\0.15, two-tailed t-test). When BVI was plotted as a function of the difference in visual acuity between normal and amblyopic eyes, which is an index of amblyopia severity, only a modest correlation (R 2 = 0.15) was found in addition to the segregation by the type of amblyopia (Fig. 3). In Fig. 4, the cross-correlation of the percent change in the MR signal between amblyopic eye stimulation and defocused normal eye stimulation is presented for each patient. The coefficient of determination (R 2) was higher in the anisometropic amblyopes, as a group, than in the strabismic amblyopes.
Fig. 3. The BVI is plotted as a function of amblyopia severity, namely, the strength in the diopter of a defocusing lens that was required to match the visual acuity of the normal eye with that of the amblyopic eye. Closed circles are for strabismic amblyopes, open circles for anisometropic amblyopes. As shown in Fig. 2, the BVIs were higher in the anisometropic group than in the strabismic group, but there was no clear relationship between the BVI and amblyopia severity (correlation coefficient: 0.15).
3.2. Difference in spatial-frequency dependence of calcarine acti6ation In Fig. 5, the percent change in the MR signal is plotted as a function of spatial frequency of the checkerboard pattern. Defocusing reduced the cortical response to higher-spatial-frequency stimulus as expected in both amblyopic groups. More importantly, the response curves of amblyopic eye stimulation were different between the two groups under the smaller check size or higher-spatial-frequency conditions. Whereas the response curve of the strabismic group was similar to that in the case of naked normal eye stimulation, the response curve of the anisometropic group was similar to that in the case of defocused normal eye stimulation, i.e. a reduced response to high-spatial-frequency stimulation.
4. Discussion
Fig. 2. The proportion of voxels within the calcarine region activated with either amblyopic eye or defocused normal eye viewing, or under both viewing conditions is shown. Since the proportion is relative to the total number of voxels activated in either of the two monocular viewing conditions, the sum of proportions for each patient is equal to one. The mean BVI is 16.85%(S.E. 2.98%) for the strabismic group, which is significantly lower than that for the anisometropic group (mean 34.98%, S.E. 4.60%; PB 0.05, two-tailed t-test) and that for normal subjects (mean 45.94%, S.E. 3.56% in comparison between the two naked-eye-viewing conditions).
In this study, two differences are found between strabismic and anisometropic amblyopes are twofold. First, the proportion of voxels activated under both monocular viewing conditions is lower in the strabismic group than in the anisometropic group. Second, the visual cortical response to higher-spatial-frequency stimulation is reduced in the anisometropic group but not in the strabismic group. These findings seem consistent with those of some electrophysiological studies using animal models of amblyopia. Since Hubel and Wiesel (1965) introduced a classic model of artificial squint, investigators employed various techniques to induce strabismus and anisometropia in kittens and monkeys (Eggers and Blakemore, 1978; Hendrickson et al., 1987; Kiorpes et al.,
K.-M. Lee et al. / Neuroscience Research 40 (2001) 147–153
151
Fig. 4. Scatter plot between the percent change in the MR signal by amblyopic eye stimulation and that by defocused normal eye stimulation is shown for every voxel within the ROI at the calcarine sulci of each patient. As a group, anisometropic amblyopes showed a stronger correlation, i.e. higher R 2, than strabismic amblyopes did.
1987; Movshon et al., 1987; Kiorpes et al., 1998). One generally accepted abnormality in animals with a squint is the reduction of conventional binocularity in the primary visual cortex neurons. Ocular dominance shifts toward the normal eye in V1 of amblyopic animals (Baker et al., 1974; Crawford and von Noorden, 1979), which is likely dependent on the degree of amblyopia (Kiorpes et al., 1998). The proportion of binocular neurons is reduced, resulting in the loss of binocular interaction (Crawford and von Noorden, 1979; Wiesel, 1982; Harrad et al., 1996; Kiorpes et al., 1998). The finding in our study that BVI was reduced in strabismic patients suggests that the less overlap between two ocular dominance columns or reduction of binocular neurons is also true in these patients. Even if the spatial resolution of our experiment was larger than the size of ocular dominance columns, the reduced BVI may still indicate a reduced contribution from presumably smaller binocularly overlapping regions between ocular dominance columns in strabismic patients’ V1. That binocular overlap represented at the voxel level was reduced in strabismic amblyopes was further supported by the finding that voxel-wise correlation between the two monocular stimulations was lower in this group of patients (Fig. 4). However, at the magnetic strength used in the current study, we could not map out the topographical arrangement of the ocular dominance in V1, and admittedly, the finding of reduced BVI should be interpreted with caution. Furthermore, the BVIs of the normal subjects were low when compared to the proportion of binocular cells found in monkey V1 (for example, Hubel and Wiesel, 1968; Schiller et al., 1976). Our speculation is that it had something to do with difference in the signal-to-noise ratio between fMRI and single-unit recording. If the sensitivity of detecting activation is lower for fMRI, then the chance of detecting co-activation by both monocular stimulations is even lower. Thus, we believe that the comparison be-
tween the BVI of our experiment and the binocularity index of previous animal studies is inappropriate, unless the sensitivity of the methods is somehow equated. Amblyopia-like effects of early unilateral blur mimicking anisometropia have been demonstrated in monkeys reared with one eye blurred by daily instillation of atropine (Boothe et al., 1982; Movshon et al., 1987). More specifically, there were changes in the function and architecture of the visual cortex that suggest a selective rearrangement of effective inputs from the amblyopic eye (Hendrickson et al., 1987). Neurons driven by the treated eye tended to have lower optimal spatial frequencies, less spatial resolution, and lower contrast sensitivity than neurons driven by the un-
Fig. 5. The spatial-frequency dependence of the primary visual cortical activation for strabismic and anisometropic groups is shown. Closed circles are for amblyopic eye stimulation, closed diamonds for naked normal eye stimulation, and open diamonds for defocused normal eye stimulation. For both groups, defocusing reduced the cortical response to smaller check size checkerboard stimulation, which contained more high-spatial-frequency components than larger check size checkerboard. For the strabismic group, the response curve under the amblyopic eye viewing condition was similar to that under the naked normal eye viewing condition, while it was similar to that under the defocused normal eye viewing condition for the anisometropic group, showing a reduced response under higher-spatialfrequency conditions. * P B0.05 and ** PB 0.005 with correction for multiple comparison, and P \0.05 for unmarked comparisons.
152
K.-M. Lee et al. / Neuroscience Research 40 (2001) 147–153
treated eye (Movshon et al., 1987). A deleterious effect of image blur on the high-spatial-frequency component of visual function was also shown in an earlier study using kittens raised wearing a high-power lens on one eye (Eggers and Blakemore, 1978). A more recent study using monkeys reared with chronic wearing of a defocusing high-power lens showed similar findings, although the changes in neural properties might be related more to the degree of induced amblyopia than to the causes of amblyopia (Kiorpes et al., 1998). In this study, we made a similar observation of a reduced high-spatial-frequency response in human anisometropic amblyopes using fMRI. Several previous studies compared various aspects of the visual function between human strabismic and anisometropic amblyopes. In strabismic amblyopes, the contrast sensitivity was found to be depressed for only a limited band of high spatial frequencies, while in anisometropic amblyopes, depression of the contrast sensitivity function was found over the entire frequency range (Sjostrand, 1981). Using VEP in correlation with contrast sensitivity, strabismic amblyopes were reported to show an abnormal functioning only in the high-spatial-frequency range, while anisometropic amblyopes showed an abnormal functioning in both the low and high-spatial-frequency ranges (Campos et al., 1984). Reduced activation of the visual cortex by stimulation of the amblyopic eye as compared with the sound eye was observed using other functional imaging tech18 niques, such as H15 2 O and [ F]-2-deoxyglucose PET (Demer et al., 1991; Imamura et al., 1997), 99mTc-HMPAO SPECT (Kabasakal et al., 1995) and MEG (Anderson et al., 1999). These findings are different from the finding of the current study that only anisometropic patients show a reduced fMRI response in the high-spatial-frequency range, while no significant difference between the amblyopic and normal eye was observed in strabismic patients. The reason for the discrepancy is unclear and requires further investigation focusing on the differences in imaging modality, properties of visual stimuli, patient characteristics, and other parameters. In summary, we examined strabismic and anisometropic amblyopes in terms of the binocularity and spatial-frequency dependence of calcarine activation. We found evidence supporting the notion that strabismic amblyopia may result from the breakdown of binocular interaction, while anisometropic amblyopia may be a result of selective undersampling of a visual image at high spatial frequencies.
Acknowledgements Supported by a grant (No. 03-98-054) from the Seoul National University Hospital Research Fund.
References Anderson, S.J., Holliday, I.E., Harding, G.F., 1999. Assessment of cortical dysfunction in human strabismic amblyopia using magnetoencephalography. Vision Res. 39, 1723 – 1738. Baker, F.H., Grigg, P., von Noorden, G.K., 1974. Effects of visual deprivation and strabismus on the response of neurons in the visual cortex of the monkey, including studies of striate and prestriate cortex in the normal animal. Brain Res. 66, 185 –208. Blakemore, C., Vital-Durand, F., 1992. Different neural origins for ‘blur’ amblyopia and strabismic amblyopia. Ophthal. Physiol. Opt. 12, 83. Boothe, R.G., Kiorpes, L., Hendrickson, A.E., 1982. Anisometropic amblyopia in Macaca nemestrina monkeys produced by stropinization of one eye during development. Invest. Ophthal. Vis. Sci. 22, 228 – 233. Campos, E.C., Prampolini, M.L., Gulli, R., 1984. Contrast sensitivity differences between strabismic and anisometropic amblyopia: objective correlate by means of visual evoked responses. Doc. Ophthalmol. 58, 45 – 50. Crawford, M.L., von Noorden, G.K., 1979. The effects of short-term experimental strabismus on the visual system in Macaca mulatta. Invest. Ophthal. Vis. Sci. 18, 496 – 505. Daw, N.W., 1995. Visual Development. Plenum, New York. Demer, J.L., von Noorden, G.K., Volkow, N.D., Gould, K.L., 1991. Brain activity in amblyopia. Am. Orthoptic. J. 41, 56 – 66. Eggers, H.M., Blakemore, C., 1978. Physiological basis of anisometropic amblyopia. Science 201, 264 – 267. Harrad, R., Sengpiel, F., Blakemore, C., 1996. Physiology of suppression in strabismic amblyopia. Brit. J. Ophthal. 80, 373 –377. Hendrickson, A.E., Movshon, J.A., Eggers, H.M., Gizzi, M.S., Boothe, R.G., Kiorpes, L., 1987. Effects of early unilateral blur on the macaque’s visual system: II. Anatomical observations. J. Neurosci. 7, 1327 – 1339. Hess, R.F., Campbell, F.W., Greenhalgh, T., 1978. On the nature of the neural abnormality in human amblyopia; neural aberrations and neural sensitivity loss. Pflugers Arch. 377, 201 – 207. Hubel, D.H., Wiesel, T.N., 1965. Binocular interaction in striate cortex of kittens reared with artificial squint. J. Neurophysiol. 28, 1041 – 1059. Hubel, D.H., Wiesel, T.N., 1968. Receptive fields and functional architecture of monkey striate cortex. J. Physiol. (Lond.) 195, 215 – 243. Imamura, K., Richter, H., Fischer, H., Lennerstrand, G., Franzen, O., Rydberg, A., et al., 1997. Reduced activity in the extrastriate visual cortex of individuals with strabismic amblyopia. Neurosci. Lett. 225, 173 – 176. Kabasakal, L., Devranoglu, K., Arslan, O., Erdil, T.Y., Sonmezoglu, K., Uslu, I., et al., 1995. Brain SPECT evaluation of the visual cortex in amblyopia. J. Nucl. Med. 36, 1170 – 1174. Kiorpes, L., Boothe, R.G., Hendrickson, A.E., Movshon, J.A., Eggers, H.M., Gizzi, M.S., 1987. Effects of early unilateral blur on the macaque’s visual system: I. Behavioral observations. J. Neurosci. 7, 1318 – 1326. Kiorpes, L., Kiper, D.C., O’Keefe, L.P., Cavanaugh, J.R., Movshon, J.A., 1998. Neuronal correlates of amblyopia in the visual cortex of macaque monkeys with experimental strabismus and anisometropia. J. Neurosci. 18, 6411 – 6424. Levi, D.M., Klein, S., 1982. Hyperacuity and amblyopia. Nature 298, 268 – 270. McKee, S.P., Schor, C.M., Steinman, S.B., Wilson, N., Koch, G.G., Davis, S.M., et al., 1992. The classification of amblyopia on the basis of visual and oculomotor performance. Trans. Am. Ophthalmol. Soc. 90, 123 – 144. Movshon, J.A., Eggers, H.M., Gizzi, M.S., Hendrickson, A.E., Kiorpes, L., Boothe, R.G., 1987. Effects of early unilateral blur on the
K.-M. Lee et al. / Neuroscience Research 40 (2001) 147–153 macaque’s visual system: III. Physiological observations. J. Neurosci. 7, 1327 – 1339. Movshon, J.A., McKee, S.P., Levi, D.M., 1996. Visual acuity in a large population of normal, strabismic, and anisometropic observers. Invest. Ophthalmol. Vis. Sci. 37, S670. Schiller, P.H., Finlay, B.L., Volman, S.L., 1976. Quantitative studies of single cell properties in monkey striate cortex. II. Orientation specificity and ocular dominance. J. Neurophysiol. 39, 1320 – 1333.
.
153
Sjostrand, J., 1981. Contrast sensitivity in children with strabismic and anisometropic amblyopia. A study of the effect of treatment. Acta Ophthalmol. (Copenhagen) 59, 25 – 34. Wiesel, T.N., 1982. Postnatal development of the visual cortex and the influence of environment. Nature 299, 583 – 591. Woods, R.P., Cherry, S.R., Mazziotta, J.C., 1992. Rapid automated algorithm for aligning and reslicing PET images. J. Comp. Assist. Tomogr. 16, 620 – 633.
Binocularity and spatial frequency dependence of calcarine activation in two types of amblyopia Kyoung-Min Lee a,e,f,*, Soo-Hwa Lee a, Na-Young Kim e, Chai-Youn Kim e, Jung-Woo Sohn a, Mi Young Choi g, Dong Gyu Choi h, Jung-Min Hwang b, Ki Ho Park b, Dong Soo Lee c,e,f, Young Suk Yu b, Kee Hyun Chang d,f a
Department of Neurology, Seoul National Uni6ersity, Seoul, 110 -744, South Korea Department of Ophthalmology, Seoul National Uni6ersity, Seoul, 110 -744, South Korea c Department of Nuclear Medicine, Seoul National Uni6ersity, Seoul, 110 -744, South Korea d Department of Radiology, Seoul National Uni6ersity, Seoul, 110 -744, South Korea e Interdisciplinary Program for Cogniti6e Science, Seoul National Uni6ersity, Seoul, 110 -744, South Korea f BK21 Human Life Science Di6ision, Seoul National Uni6ersity, Seoul, 110 -744, South Korea g Department of Ophthalmology, Chongbuk National Uni6ersity, Chongju, South Korea h Department of Ophthalmology, Hallym Uni6ersity, Seoul, South Korea b
Received 28 June 2000; accepted 16 February 2001
Abstract Objecti6e and Background: Strabismus and anisometropia early in life frequently causes monocular amblyopia. Activation of the visual cortex is compared between the two types of amblyopia to elucidate differences in the pathogenetic mechanism of the disease. Methods: Using an EPI gradient echo sequence in 1.5T MRI, calcarine activation by monocular viewing of checkerboard patterns with reversal was examined in terms of binocularity of the activation and dependence on the spatial frequency of the stimuli. Results: First, the proportion of voxels activated by both normal and amblyopic eye monocular stimulations is lower in the strabismic group than in the anisometropic group. Second, the activation by higher-spatial-frequency stimuli is reduced in the anisometropic group, but not in the strabismic group. Conclusions: These findings from the human visual cortex are consistent with the view proposed based on animal research that the loss of binocular interaction and the undersampling of high-spatial-frequency components of visual stimuli are each one of the underlying changes in strabismic and anisometropic amblyopia, respectively. © 2001 Elsevier Science Ireland Ltd and the Japan Neuroscience Society. All rights reserved. Keywords: Amblyopia; Strabismus; Anisometropia; Visual cortex; Brain activation; fMRI
1. Introduction Development of the visual function critically depends on visual experience early in life (Daw, 1995). The early visual experience is deranged under such conditions as strabismus, in which the visual axes are misaligned, and anisometropia, in which the refractive state is significantly different between the two eyes. The two conditions often lead to monocular amblyopia without a macroscopic lesion along the visual pathway. The
* Corresponding author. Tel.: + 82-2-7602985; fax: 36727553. E-mail address: [email protected] (K.-M. Lee).
+ 82-2-
pathogenetic mechanism of amblyopia may differ between the two conditions (Blakemore and VitalDurand, 1992), since strabismus prevents concordant stimulation of binocularly corresponding points in the two retinae, whereas anisometropia imposes constant blurring of one retinal image relative to the other. Clinical studies on human amblyopics confirmed that the two conditions lead to different forms of amblyopia (Hess et al., 1978; Levi and Klein, 1982), although this issue remains controversial (McKee et al., 1992; Movshon et al., 1996). From experiments using a number of animal models, it has been suggested that strabismic amblyopia results from the breakdown of binocular interaction (Crawford and von Noorden, 1979; Wiesel, 1982; Harrad et al., 1996; Kiorpes et al., 1998). In
0168-0102/01/$ - see front matter © 2001 Elsevier Science Ireland Ltd and the Japan Neuroscience Society. All rights reserved. PII: S 0 1 6 8 - 0 1 0 2 ( 0 1 ) 0 0 2 2 0 - 6
K.-M. Lee et al. / Neuroscience Research 40 (2001) 147–153
148
contrast, reduction in the spatial resolution of neurons with preferred spatial frequency in the higher range or the loss of such neurons may underlie anisometropic amblyopia (Hess et al., 1978; Kiorpes et al., 1987; Movshon et al., 1987). Thus, in this study, activation of the human visual cortex around the calcarine sulci observed using functional MRI (fMRI) is compared between the two types of amblyopia, in terms of binocularity of the activation and its response characteristics to various spatial frequencies.
projector controlled by a personal computer. The frequency of checkerboard reversal was 8 Hz (i.e. 4 cycles per second, since one cycle contained two reversals). The distance from the inside of the scanner to the screen was 4 m. The extent of the stimulated visual field was 16 and 10 degrees in horizontal and vertical dimensions, respectively, in the middle of which a crosshair was shown for fixation. Patients were asked to view the visual stimuli monocularly under the following three viewing conditions: (1) with the amblyopic eye, (2) with the naked normal eye, and (3) with the normal eye defocused to match the visual acuity with that of the amblyopic eye by a plus lens. The strength of the lens was selected before the MR scanning such that the visual acuity of the normal eye measured with the lens in front was the same as that of the amblyopic eye at a 6-m viewing distance. Monocular viewing during MRI scanning was achieved by placing a black plate in front of one eye onto a custom-made MRI-compatible goggle frame. Defocusing, when needed, was similarly achieved by placing the defocusing lens in front of the viewing eye. For switching the side of the blocked eye, the scanner table was slid out just enough for an experimenter to move the blocking plate to the other side of the goggle frame without touching the patient’s head on the headrest. Each patient was subjected to three MR scan runs in total, one for each viewing condition. The sequence of the viewing conditions was pseudorandomized across patients. An MR scan run for each viewing condition lasted 180 s, divided into six 30-s periods. No visual stimulus except a center crosshair was presented during the first and sixth epochs, regarded as the baseline epochs. During the middle four activation epochs, the checker-
2. Methods
2.1. Patients Eleven amblyopic patients (six anisometropic, aged 8 – 18 years old, and five strabismic patients, aged 5–32 years old) and three normal subjects (aged 18– 22 years old) participated in this study after giving their informed consent. Clinical data on the patients and normal subjects are summarized in Table 1. Fourteen patients were originally recruited and subjected to MR scanning, but in three of them (one strabismic and two anisometropic), the data contained too much movement artifacts for a valid analysis and were therefore excluded.
2.2. Visual stimulus and experimental conditions Visual stimuli were black-and-white checkerboard patterns presented at a spatial frequency of 2, 1, 0.5, or 0.25 cycles per degree of visual angle (i.e. individual check-sizes of 0.25, 0.5, 1, and 2 degrees of visual angle, respectively), back-projected onto a screen by a LCD Table 1 Clinical data on tested amblyopic patients and normal control subjectsa Patient ID
S1 S2 S3 S4 S5 A1 A2 A3 A4 A5 A6 C1 C2 C3 a
Amblyopia
OD OS OS OD OS OD OS OD OD OD OD – – –
Sex
M F M M F M M M M F M M F M
Age (yrs)
5 23 5 8 23 9 17 10 8 16 10 18 22 18
VA OD
OS
0.04 1.50 1.00 0.15 1.00 0.08 1.00 0.10 0.02 0.20 0.10 1.00 1.00 1.50
1.00 0.10 0.10 1.00 0.15 1.00 0.08 1.00 0.90 1.20 1.00 1.00 1.50 1.20
EOM abnormality, prism diopter
Anisometropia
eso, 50 exo, 35 eso, 45 eso, 16 eso, 25 ortho ortho ortho ortho eso, B10 ortho ortho ortho ortho
– – – – – hyperopic hyperopic hyperopic myopic hyperopic hyperopic – – –
VA: visual acuity (best corrected), OD: right eye, OS: left eye, eso: esotropic, exo: exotropic, ortho: orthotropic.
K.-M. Lee et al. / Neuroscience Research 40 (2001) 147–153
Fig. 1. In the left panel, the location of image slices (dotted lines) and the ROI within them (solid portion of the dotted lines) is indicated for a typical case. In the right panel, an example of the ROIs is shown in a slice image with the activation around the calcarine sulci overlayed. See text for details on defining the ROI.
board pattern of four different spatial frequencies was presented. The sequence of the spatial frequencies tested was fixed for each patient, but counter-balanced across patients. The patients were instructed to fixate on the center crosshair at all times during the scanning. Their eye movements were not monitored objectively.
2.3. MR image acquisition T2*-weighted images were obtained using a 1.5T MR scanner equipped with a 5-in. surface coil provided with the scanner (General Electric, Milwaukee). Scan parameters were as follows: gradient echo EPI sequence, TR 3000 ms, TE 60 ms, flip angle 90 degree, 15 3-mm-thick slices without separation, FOV 14 cm, image matrix 128×128 (thus, a voxel size of 1.09× 1.09 ×3 mm). The surface coil was centered at the inion. Image slices were positioned perpendicular to the inter-commissural plane (Fig. 1, left panel). The images were reconstructed off-line, exported to volumes, and aligned for movement correction using the AIR3.0 package (Woods et al., 1992).
2.4. Data analysis 2.4.1. Analysis of binocularity of fMRI response This analysis was carried out within the region of interest (ROI) along the calcarine sulci, which was defined using T1-weighted images that were obtained on the same slice locations as the EPI images (Fig. 1, left panel). The ROIs were drawn on the caudalmost slice that included at least one occipital pole and five consecutive slices rostral to it, i.e. six slices in total and 18 mm in rostro-caudal direction. After identifying the calcarine sulcus on each slice, an intersection between the midline and the line linking the two calcarine sulci on both sides was chosen as the center point. A square ROI was drawn based on the center point; the boundary of the ROI was 1 cm above and below the
149
center point and at 2 cm on both sides from the midline (Fig. 1, right panel). Although we did not attempt functional demarcation among the visual areas, we considered that the ROI most likely included some higher visual areas, such as V2, as well as the primary visual cortex. The voxelwise t-test (PB 0.05, two-tailed, uncorrected) was performed to compare the MR signal between four activation epochs and two baseline epochs. The binocularity of the visual response was then determined using two kinds of analysis. First, whether monocular stimulation through either eye evoked activation was determined for every voxel. A voxel was classified as binocularly activated if the signal intensity significantly increased in both amblyopic-eye- and defocused-normal-eye-viewing conditions. The proportion of such voxels was calculated as the binocular voxel index (BVI) with respect to the total number of voxels that showed activation under at least one of the two conditions, i.e. BVI = No. of voxels in (ODSOS)/No. of voxels in (OD@OS) ×100, where S and @ are the intersection between and the union of OD and OS, or monocular stimulation of the right and left eye, respectively. Second, correlation of activation was assessed by plotting the percent change in the MR signal during amblyopic eye stimulation and that during defocused normal eye stimulation for every voxel. The coefficient of determination (R 2) was then calculated for each patient.
2.4.2. Analysis of spatial-frequency dependence of fMRI response Whether the calcarine activation by a checkerboard pattern of various check sizes is different between the two amblyopic types is examined. SPM96 software (Wellcome Department of Cognitive Neurology, London) was employed to identify the locus of the maximum response within the ROI. The voxel with the maximum activation and 24 neighboring voxels on the same slice were chosen as a region with maximal response. From the maximal activation region, the mean percent change in MR signal was computed between the first baseline epoch and each of four different spatial-frequency activation epochs for each patient. The group average of the results was then compared between the three different viewing conditions using two-tailed t-test.
3. Results
3.1. Difference in binocularity of calcarine acti6ation The total number of voxels activated in any of the three viewing conditions was not different between the two amblyopic groups (453 voxels in the stabismic
150
K.-M. Lee et al. / Neuroscience Research 40 (2001) 147–153
group vs. 513 in the anisometropic group, P \ 0.45, two-tailed t-test). The ratio of the number of activated voxels in contralateral hemisphere to that in hemisphere ipsilateral to the stimulated eye did not reveal a consistent pattern across patients or across patient groups. The proportion of binocularly activated voxels, i.e. BVI was, however, significantly lower in the strabismic amblyopes (mean9standard error, 16.859 2.98%) than in the anisometropic amblyopes (mean9 standard error, 34.9894.60% P B 0.05, two-tailed t-test; Fig. 2). Comparison between the amblyopic eye and the naked normal eye revealed a similar difference (strabismic group BVI mean9 standard error 19.894.86%, and anisometropic group BVI mean9 standard error, 30.2 93.63%) but the difference did not reach statistical significance (P\0.05, two-tailed t-test). For the control group of three normal subjects, BVI was 45.949 3.56%, computed under the two naked-eye-viewing conditions. This value was significantly higher than that of the strabismic group (P B 0.001, two-tailed t-test), but not significantly different from that of the anisometric group (P\0.15, two-tailed t-test). When BVI was plotted as a function of the difference in visual acuity between normal and amblyopic eyes, which is an index of amblyopia severity, only a modest correlation (R 2 = 0.15) was found in addition to the segregation by the type of amblyopia (Fig. 3). In Fig. 4, the cross-correlation of the percent change in the MR signal between amblyopic eye stimulation and defocused normal eye stimulation is presented for each patient. The coefficient of determination (R 2) was higher in the anisometropic amblyopes, as a group, than in the strabismic amblyopes.
Fig. 3. The BVI is plotted as a function of amblyopia severity, namely, the strength in the diopter of a defocusing lens that was required to match the visual acuity of the normal eye with that of the amblyopic eye. Closed circles are for strabismic amblyopes, open circles for anisometropic amblyopes. As shown in Fig. 2, the BVIs were higher in the anisometropic group than in the strabismic group, but there was no clear relationship between the BVI and amblyopia severity (correlation coefficient: 0.15).
3.2. Difference in spatial-frequency dependence of calcarine acti6ation In Fig. 5, the percent change in the MR signal is plotted as a function of spatial frequency of the checkerboard pattern. Defocusing reduced the cortical response to higher-spatial-frequency stimulus as expected in both amblyopic groups. More importantly, the response curves of amblyopic eye stimulation were different between the two groups under the smaller check size or higher-spatial-frequency conditions. Whereas the response curve of the strabismic group was similar to that in the case of naked normal eye stimulation, the response curve of the anisometropic group was similar to that in the case of defocused normal eye stimulation, i.e. a reduced response to high-spatial-frequency stimulation.
4. Discussion
Fig. 2. The proportion of voxels within the calcarine region activated with either amblyopic eye or defocused normal eye viewing, or under both viewing conditions is shown. Since the proportion is relative to the total number of voxels activated in either of the two monocular viewing conditions, the sum of proportions for each patient is equal to one. The mean BVI is 16.85%(S.E. 2.98%) for the strabismic group, which is significantly lower than that for the anisometropic group (mean 34.98%, S.E. 4.60%; PB 0.05, two-tailed t-test) and that for normal subjects (mean 45.94%, S.E. 3.56% in comparison between the two naked-eye-viewing conditions).
In this study, two differences are found between strabismic and anisometropic amblyopes are twofold. First, the proportion of voxels activated under both monocular viewing conditions is lower in the strabismic group than in the anisometropic group. Second, the visual cortical response to higher-spatial-frequency stimulation is reduced in the anisometropic group but not in the strabismic group. These findings seem consistent with those of some electrophysiological studies using animal models of amblyopia. Since Hubel and Wiesel (1965) introduced a classic model of artificial squint, investigators employed various techniques to induce strabismus and anisometropia in kittens and monkeys (Eggers and Blakemore, 1978; Hendrickson et al., 1987; Kiorpes et al.,
K.-M. Lee et al. / Neuroscience Research 40 (2001) 147–153
151
Fig. 4. Scatter plot between the percent change in the MR signal by amblyopic eye stimulation and that by defocused normal eye stimulation is shown for every voxel within the ROI at the calcarine sulci of each patient. As a group, anisometropic amblyopes showed a stronger correlation, i.e. higher R 2, than strabismic amblyopes did.
1987; Movshon et al., 1987; Kiorpes et al., 1998). One generally accepted abnormality in animals with a squint is the reduction of conventional binocularity in the primary visual cortex neurons. Ocular dominance shifts toward the normal eye in V1 of amblyopic animals (Baker et al., 1974; Crawford and von Noorden, 1979), which is likely dependent on the degree of amblyopia (Kiorpes et al., 1998). The proportion of binocular neurons is reduced, resulting in the loss of binocular interaction (Crawford and von Noorden, 1979; Wiesel, 1982; Harrad et al., 1996; Kiorpes et al., 1998). The finding in our study that BVI was reduced in strabismic patients suggests that the less overlap between two ocular dominance columns or reduction of binocular neurons is also true in these patients. Even if the spatial resolution of our experiment was larger than the size of ocular dominance columns, the reduced BVI may still indicate a reduced contribution from presumably smaller binocularly overlapping regions between ocular dominance columns in strabismic patients’ V1. That binocular overlap represented at the voxel level was reduced in strabismic amblyopes was further supported by the finding that voxel-wise correlation between the two monocular stimulations was lower in this group of patients (Fig. 4). However, at the magnetic strength used in the current study, we could not map out the topographical arrangement of the ocular dominance in V1, and admittedly, the finding of reduced BVI should be interpreted with caution. Furthermore, the BVIs of the normal subjects were low when compared to the proportion of binocular cells found in monkey V1 (for example, Hubel and Wiesel, 1968; Schiller et al., 1976). Our speculation is that it had something to do with difference in the signal-to-noise ratio between fMRI and single-unit recording. If the sensitivity of detecting activation is lower for fMRI, then the chance of detecting co-activation by both monocular stimulations is even lower. Thus, we believe that the comparison be-
tween the BVI of our experiment and the binocularity index of previous animal studies is inappropriate, unless the sensitivity of the methods is somehow equated. Amblyopia-like effects of early unilateral blur mimicking anisometropia have been demonstrated in monkeys reared with one eye blurred by daily instillation of atropine (Boothe et al., 1982; Movshon et al., 1987). More specifically, there were changes in the function and architecture of the visual cortex that suggest a selective rearrangement of effective inputs from the amblyopic eye (Hendrickson et al., 1987). Neurons driven by the treated eye tended to have lower optimal spatial frequencies, less spatial resolution, and lower contrast sensitivity than neurons driven by the un-
Fig. 5. The spatial-frequency dependence of the primary visual cortical activation for strabismic and anisometropic groups is shown. Closed circles are for amblyopic eye stimulation, closed diamonds for naked normal eye stimulation, and open diamonds for defocused normal eye stimulation. For both groups, defocusing reduced the cortical response to smaller check size checkerboard stimulation, which contained more high-spatial-frequency components than larger check size checkerboard. For the strabismic group, the response curve under the amblyopic eye viewing condition was similar to that under the naked normal eye viewing condition, while it was similar to that under the defocused normal eye viewing condition for the anisometropic group, showing a reduced response under higher-spatialfrequency conditions. * P B0.05 and ** PB 0.005 with correction for multiple comparison, and P \0.05 for unmarked comparisons.
152
K.-M. Lee et al. / Neuroscience Research 40 (2001) 147–153
treated eye (Movshon et al., 1987). A deleterious effect of image blur on the high-spatial-frequency component of visual function was also shown in an earlier study using kittens raised wearing a high-power lens on one eye (Eggers and Blakemore, 1978). A more recent study using monkeys reared with chronic wearing of a defocusing high-power lens showed similar findings, although the changes in neural properties might be related more to the degree of induced amblyopia than to the causes of amblyopia (Kiorpes et al., 1998). In this study, we made a similar observation of a reduced high-spatial-frequency response in human anisometropic amblyopes using fMRI. Several previous studies compared various aspects of the visual function between human strabismic and anisometropic amblyopes. In strabismic amblyopes, the contrast sensitivity was found to be depressed for only a limited band of high spatial frequencies, while in anisometropic amblyopes, depression of the contrast sensitivity function was found over the entire frequency range (Sjostrand, 1981). Using VEP in correlation with contrast sensitivity, strabismic amblyopes were reported to show an abnormal functioning only in the high-spatial-frequency range, while anisometropic amblyopes showed an abnormal functioning in both the low and high-spatial-frequency ranges (Campos et al., 1984). Reduced activation of the visual cortex by stimulation of the amblyopic eye as compared with the sound eye was observed using other functional imaging tech18 niques, such as H15 2 O and [ F]-2-deoxyglucose PET (Demer et al., 1991; Imamura et al., 1997), 99mTc-HMPAO SPECT (Kabasakal et al., 1995) and MEG (Anderson et al., 1999). These findings are different from the finding of the current study that only anisometropic patients show a reduced fMRI response in the high-spatial-frequency range, while no significant difference between the amblyopic and normal eye was observed in strabismic patients. The reason for the discrepancy is unclear and requires further investigation focusing on the differences in imaging modality, properties of visual stimuli, patient characteristics, and other parameters. In summary, we examined strabismic and anisometropic amblyopes in terms of the binocularity and spatial-frequency dependence of calcarine activation. We found evidence supporting the notion that strabismic amblyopia may result from the breakdown of binocular interaction, while anisometropic amblyopia may be a result of selective undersampling of a visual image at high spatial frequencies.
Acknowledgements Supported by a grant (No. 03-98-054) from the Seoul National University Hospital Research Fund.
References Anderson, S.J., Holliday, I.E., Harding, G.F., 1999. Assessment of cortical dysfunction in human strabismic amblyopia using magnetoencephalography. Vision Res. 39, 1723 – 1738. Baker, F.H., Grigg, P., von Noorden, G.K., 1974. Effects of visual deprivation and strabismus on the response of neurons in the visual cortex of the monkey, including studies of striate and prestriate cortex in the normal animal. Brain Res. 66, 185 –208. Blakemore, C., Vital-Durand, F., 1992. Different neural origins for ‘blur’ amblyopia and strabismic amblyopia. Ophthal. Physiol. Opt. 12, 83. Boothe, R.G., Kiorpes, L., Hendrickson, A.E., 1982. Anisometropic amblyopia in Macaca nemestrina monkeys produced by stropinization of one eye during development. Invest. Ophthal. Vis. Sci. 22, 228 – 233. Campos, E.C., Prampolini, M.L., Gulli, R., 1984. Contrast sensitivity differences between strabismic and anisometropic amblyopia: objective correlate by means of visual evoked responses. Doc. Ophthalmol. 58, 45 – 50. Crawford, M.L., von Noorden, G.K., 1979. The effects of short-term experimental strabismus on the visual system in Macaca mulatta. Invest. Ophthal. Vis. Sci. 18, 496 – 505. Daw, N.W., 1995. Visual Development. Plenum, New York. Demer, J.L., von Noorden, G.K., Volkow, N.D., Gould, K.L., 1991. Brain activity in amblyopia. Am. Orthoptic. J. 41, 56 – 66. Eggers, H.M., Blakemore, C., 1978. Physiological basis of anisometropic amblyopia. Science 201, 264 – 267. Harrad, R., Sengpiel, F., Blakemore, C., 1996. Physiology of suppression in strabismic amblyopia. Brit. J. Ophthal. 80, 373 –377. Hendrickson, A.E., Movshon, J.A., Eggers, H.M., Gizzi, M.S., Boothe, R.G., Kiorpes, L., 1987. Effects of early unilateral blur on the macaque’s visual system: II. Anatomical observations. J. Neurosci. 7, 1327 – 1339. Hess, R.F., Campbell, F.W., Greenhalgh, T., 1978. On the nature of the neural abnormality in human amblyopia; neural aberrations and neural sensitivity loss. Pflugers Arch. 377, 201 – 207. Hubel, D.H., Wiesel, T.N., 1965. Binocular interaction in striate cortex of kittens reared with artificial squint. J. Neurophysiol. 28, 1041 – 1059. Hubel, D.H., Wiesel, T.N., 1968. Receptive fields and functional architecture of monkey striate cortex. J. Physiol. (Lond.) 195, 215 – 243. Imamura, K., Richter, H., Fischer, H., Lennerstrand, G., Franzen, O., Rydberg, A., et al., 1997. Reduced activity in the extrastriate visual cortex of individuals with strabismic amblyopia. Neurosci. Lett. 225, 173 – 176. Kabasakal, L., Devranoglu, K., Arslan, O., Erdil, T.Y., Sonmezoglu, K., Uslu, I., et al., 1995. Brain SPECT evaluation of the visual cortex in amblyopia. J. Nucl. Med. 36, 1170 – 1174. Kiorpes, L., Boothe, R.G., Hendrickson, A.E., Movshon, J.A., Eggers, H.M., Gizzi, M.S., 1987. Effects of early unilateral blur on the macaque’s visual system: I. Behavioral observations. J. Neurosci. 7, 1318 – 1326. Kiorpes, L., Kiper, D.C., O’Keefe, L.P., Cavanaugh, J.R., Movshon, J.A., 1998. Neuronal correlates of amblyopia in the visual cortex of macaque monkeys with experimental strabismus and anisometropia. J. Neurosci. 18, 6411 – 6424. Levi, D.M., Klein, S., 1982. Hyperacuity and amblyopia. Nature 298, 268 – 270. McKee, S.P., Schor, C.M., Steinman, S.B., Wilson, N., Koch, G.G., Davis, S.M., et al., 1992. The classification of amblyopia on the basis of visual and oculomotor performance. Trans. Am. Ophthalmol. Soc. 90, 123 – 144. Movshon, J.A., Eggers, H.M., Gizzi, M.S., Hendrickson, A.E., Kiorpes, L., Boothe, R.G., 1987. Effects of early unilateral blur on the
K.-M. Lee et al. / Neuroscience Research 40 (2001) 147–153 macaque’s visual system: III. Physiological observations. J. Neurosci. 7, 1327 – 1339. Movshon, J.A., McKee, S.P., Levi, D.M., 1996. Visual acuity in a large population of normal, strabismic, and anisometropic observers. Invest. Ophthalmol. Vis. Sci. 37, S670. Schiller, P.H., Finlay, B.L., Volman, S.L., 1976. Quantitative studies of single cell properties in monkey striate cortex. II. Orientation specificity and ocular dominance. J. Neurophysiol. 39, 1320 – 1333.
.
153
Sjostrand, J., 1981. Contrast sensitivity in children with strabismic and anisometropic amblyopia. A study of the effect of treatment. Acta Ophthalmol. (Copenhagen) 59, 25 – 34. Wiesel, T.N., 1982. Postnatal development of the visual cortex and the influence of environment. Nature 299, 583 – 591. Woods, R.P., Cherry, S.R., Mazziotta, J.C., 1992. Rapid automated algorithm for aligning and reslicing PET images. J. Comp. Assist. Tomogr. 16, 620 – 633.