You are here

Article Text

Article menu

Download PDFPDF

Clinical science
Perceptual learning treatment in patients with anisometropic amblyopia: a neuroimaging study
  1. Jingjing Zhai1,2,
  2. Min Chen1,
  3. Lijuan Liu3,
  4. Xuna Zhao4,
  5. Hong Zhang5,
  6. Xiaojie Luo1,
  7. Jiahong Gao4
  1. 1Department of Radiology, Beijing Hospital, Beijing, China
  2. 2Department of Radiology, Tianjin Medical University Cancer Institute and Hospital, Tianjin, China
  3. 3Beijing Institute of Ophthalmology, Beijing Tongren Eye Center, Beijing Tongren Hospital, Capital Medical University, Beijing, China
  4. 4Center for MRI Research, Beijing City Key Lab for Medical Physics and Engineering, Peking University, Beijing, China
  5. 5Imaging Center, Beijing Children's Hospital, Capital Medical University, Beijing, China
  1. Correspondence to Dr Min Chen, Department of Radiology, Beijing Hospital, Beijing 100730, China; cjr.chenmin{at}vip.163.com

Abstract

Aims To investigate the neuromechanisms of perceptual learning treatment in patients with anisometropic amblyopia using functional MRI (fMRI) and diffusion tensor imaging (DTI) techniques.

Methods 20 patients with monocular anisometropic amblyopia participated in the study. Both fMRI and DTI data were acquired for each patient twice: before and after 30 days’ perceptual learning treatment for the amblyopic eye. During fMRI scanning, patients viewed the stimuli with either the sound or amblyopic eye. Changes of cortical activation after treatment were evaluated. In the DTI exams, the fractional anisotropy (FA) values, apparent diffusion coefficient (ADC) values, the voxel numbers of optic radiations (ORs), and the number of tracks were compared between the ipsilateral and the contralateral ORs and also between the previous and posterior scans.

Results Remarkable increased activation via the amblyopic eyes was found in Brodmann Area (BA) 17–19, bilateral temporal lobes, and right cingulate gyrus after the perceptual learning treatment. No significant changes were found in the FA values, ADC values, voxel numbers, and the number of tracks after the treatment.

Conclusions These results indicate that perceptual learning treatment for amblyopia had a positive effect on the visual cortex and temporal lobe visual areas in patients with anisometropic amblyopia.

https://doi.org/10.1136/bjophthalmol-2013-303778

Statistics from Altmetric.com

    Request Permissions

    If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

    Introduction

    Amblyopia is a reduction of visual acuity that cannot be attributed directly to the effect of any structural abnormality of the eye or the posterior visual pathway.1 It is thought to be caused by an abnormal visual experience early in life, but its neural mechanisms are still not fully understood. According to the traditional clinical theory, the visual system in humans beyond the sensitive period (generally older than 12 years) has no ‘plasticity’. This traditional view has been challenged by recent clinical studies, in which researchers found that perceptual learning could significantly improve the visual acuity and visual spatial ability in older patients with amblyopia.2–5 The neural rehabilitation mechanism based on these new clinical findings needs to be explored.

    It is well known that blood oxygen level dependent functional MRI (BOLD-fMRI) can detect the activation of the cortex non-invasively and graphically. Although BOLD-fMRI has been used previously on patients with amblyopia, to the best of our knowledge no fMRI study has been reported on the effects of perceptual learning treatment on patients with amblyopia. It is recognised that diffusion tensor imaging (DTI) is an effective technique which can provide both structural and functional information about the white matter quantitatively in vivo,6 but there are few DTI studies on amblyopia. There are also no DTI studies on the effects of perceptual learning treatment on patients with amblyopia.

    In this report, we describe our work on the study of neuromechanisms of perceptual learning treatment in patients with amblyopia, combining the techniques of BOLD-fMRI and DTI.

    Subject and methods

    Subjects

    The study was approved by the Ethics Committees of Beijing Hospital and followed the tenets of the Declaration of Helsinki. Written informed consent was obtained from all participants enrolled in the study or their legal guardians, and all participants received detailed eye examinations before the start of the study. All subjects who participated in the study were randomly selected from patients referred to the department of ophthalmology of the Beijing Tong-ren Hospital. In total, 20 patients (11 males, nine females, age range 11–32 years, amblyopic eyes’ vision 0.2–0.7) with monocular anisometropic amblyopia were enrolled in the study and all of the patients were right-handed. Patients with a known organic brain disorder or with specific clinical evidence of neurologic dysfunction were excluded. After the patients completed their ophthalmologic exam and their type of amblyopia was determined, they were entered into the study. MR scanning was performed twice: before and after 30 days’ perceptual learning treatment.

    Perceptual learning treatment

    All of the patients received the professional grating stimuli, which was based on the software packages of Psychotoolbox and MATLAB. The ‘cut-off’ frequency of the stimuli for the amblyopic eye was found by drawing the visual contrast sensitivity curve of each patient. The grating stimulus with the ‘cut-off’ frequency was used to treat the amblyopic eye for 40 min per day and treatment course was 30 days.

    Stimuli

    A block design fMRI paradigm was used in this study. A full field reversible black and white circular checkerboard (spatial frequency=40°/cycle, visual angle=17°) flashing at 8 Hz was designed as the visual stimuli. The control task consisted of a static white cross centred on a black background. The stimulus was back-projected onto a translucent screen on the scanner bed. An angled mirror, positioned above the subjects’ eyes, provided a full view of the screen to the subjects. Both eyes were optimally focused to best-corrected visual acuity using lenses.

    Data acquisition

    All the MRIs were performed with a Philips Intera Achieva 3 T MR scanner. The subjects were in a supine position, and head motion was minimised with restraining sponge pads.

    Conventional axial T1-weighted images were obtained with the following acquisition parameters: repetition time (TR)=2 s, echo time (TE)=15 ms, field of view=230×230 mm, matrix=512×512. Twenty-four contiguous axial slices were acquired with 4 mm thickness and no gap covering the whole brain.

    Two separate fMRI scans were performed with visual stimulation delivered through the sound eye and the amblyopic eye, respectively. Each scan contained three resting-state blocks, interleaved by three stimulation-state blocks. Each of the six blocks lasted for 30 s. The visual stimulation was presented on the computer screen and the subjects could receive the visual stimuli through a reflector over their forehead. A single-shot T2*-weighted gradient-echo echo-planar imaging sequence was used for BOLD imaging. The slice thickness and location of the BOLD images were identical to those used in the T1 images. The scanning parameters for BOLD imaging were: TR=2 s, TE=35 ms, matrix=128×128. During each of the two scans, 90 images per slice were obtained with a total time of 180 s.

    DTI was acquired with a spin-echo EPI (echo planar imaging) sequence. The acquisition parameters were as follows: diffusion sensitising gradient directions =31, b=800 s/mm2; TR=6 s; TE=60 ms; matrix=144×144; field of view (FOV)=230×230 mm2. Forty-eight contiguous axial slices were acquired with 2 mm thickness and no gap covering the whole brain.

    Data analysis

    The fMRI data were analysed with MATLAB V.7.8 (The MathWorks, Inc, Natick, Massachusetts, USA) using the statistical parametric mapping software SPM8 (http://www.fil.ion.ucl.ac.uk/spm/software/spm8). In order to minimise the transient effects of the haemodynamic responses, the images from the first 8 s of each block were excluded. The image preprocessing included realignment, co-registration, segment, normalisation and smoothing. If the head motion was found to be more than 2.0 mm in translation or 2.0° in rotation, the data of the subject were excluded. Based on these restrictions for head motion, the data of three subjects were excluded from further data processing. After the preprocessing, fMRI models were set up to obtain the primary individual fMRI maps of each subject. Finally we undertook group analysis and statistical results were exported.

    We used the one-sample t test to obtain the location of the cortical activation of the sound eye group and the amblyopic eye group separately. Paired t tests were used to compare the cortical activity induced by visual stimulation through the amblyopic eye before and after the treatment. In order to check the variation of experimental conditions before and after the treatment, and also to explore more possible underlying neuromechanisms, we compared the cortical activity induced by visual stimulation through the sound eye of the two scans at the different times. The difference between the sound and amblyopic eyes after the treatment was also observed by paired t tests. For the fMRI data analysis, we used a significance level of p<0.001, uncorrected, extent threshold=10 voxels.

    Each subject's DTI dataset was transferred to a personal computer running Diffusion Toolkit+TrackVis (V.0.6.2.1; Martinos Center for Biomedical Imaging Massachusetts General Hospital, Boston, Massachusetts, USA). The propagation was continued until they turned at an angle >70° using the auto mask threshold. To extract the optic radiation (OR) from all tracts, we manually defined the lateral geniculate body on the basis of anatomical knowledge as the first target regions in the coronal colour-tensor slice. The second target region containing the white matter of the occipital lobe was prescribed at the level of the middle calcarine sulcus in the coronal colour-tensor slice. All tracts passing these two regions were saved as the results of ORs (figure 1). The fractional anisotropy (FA) values, the apparent diffusion coefficient (ADC) values, the voxel numbers, and the tracks’ number of ORs were calculated. SPSS13.0 was used to complete the statistical analysis of those values above and we used the significance level of p<0.05. Paired t tests were used to compare those values between the right ORs and the left ORs and also between the previous and posterior scans.

    Figure 1
    Figure 1

    Right optic radiation (OR) of one patient.

    Results

    It was found that the main active areas were the Brodmann Area (BA) 17, 18, and 19 for both the sound eye and the amblyopic eye group. The area was smaller and the intensity was lower for the cortical activation result from the visual stimulation on the amblyopic eye than that on the sound eye in BA 17 and 18 (p<0.001, uncorrected, extent threshold=10 voxels) before treatment (table 1).

    View this table:
    Table 1

    Areas of increased cortex activation with stimulation delivered through the sound eye and through the amblyopic eye

    With visual stimulation delivered through the amblyopic eye, significant increases of cortical activation were found in BA 17–19, the bilateral temporal lobes, and the right cingulate gyrus after 30 days’ perceptual learning treatment (p<0.001, uncorrected, extent threshold=10 voxels) (figure 2, table 2).

    View this table:
    Table 2

    Areas of increased cortex activation with stimulation delivered through the amblyopic eye after 30 days’ perceptual learning treatment

    Figure 2
    Figure 2

    Increased cortex activation with stimulation delivered through the amblyopic eye after perceptual learning treatment compared with before treatment.

    In addition, we compared the activation areas with the visual stimulation delivered through the sound eye before and after treatment, but found no significant difference. There were no significant differences between the cortical activation result from the visual stimulation on the amblyopic eye and that on the sound eye after treatment.

    There were no significant differences in FA values, ADC values, the voxel numbers, and the number of tracks between the previous and posterior scans either in the right ORs or left ORs (p>0.05).

    The voxel numbers of the left ORs were significantly more than those of the right ORs (p<0.05) both before and after treatment, while the FA values, ADC values, and number of tracks showed no differences between the left and the right ORs (p>0.05); the details of these results are presented in tables 3 and 4.

    View this table:
    Table 3

    Mean (±SD) values of FA, ADC, and voxel numbers of the left ORs and right ORs in patients with anisometropic amblyopia before perceptual learning

    View this table:
    Table 4

    Mean (±SD) values of FA, ADC, and voxel numbers of the left ORs and right ORs in patients with anisometropic amblyopia after perceptual learning

    Discussion

    There are unilateral or bilateral reductions of visual acuity in patients with amblyopia, but little structural abnormality of the visual pathway could be found using various clinical examinations. The main cause of amblyopia is considered to be an ocular misalignment or uncorrected refractive error. Studies in animals made artificially amblyopic7 ,8 and electrophysiological studies in humans9 ,10 have suggested that the deficit is not in the retina. Further research has concluded that the main site of the amblyopic deficit is located in the visual cortex.11

    The visual systems in adolescents and adults beyond the critical period of visual development were thought to have no ‘plasticity’ in the past, but recently some clinical research has indicated that perceptual learning, which is a new treatment for amblyopia, could significantly improve the visual acuity and the visual spatial ability in adolescents and adults with amblyopia. In our study, improvements of visual acuity to varying degrees were detected in all of the patients after treatment.

    In recent years there has been much fMRI research on the dysfunction of the visual cortex in amblyopia. The main cortical activation areas were proved to be located in BA 17, 18 and 19, and the cortical activation during amblyopic eye stimulation was shown to be decreased compared with normal eye stimulation,12–15 which is consistent with our results.

    As far as we know there has been no previous fMRI study on the perceptual learning treatment of amblyopia, and we are the first to investigate this therapeutic method in patients with amblyopia using fMRI. According to our results, the main areas of increased cortical activation during amblyopic eye stimulation after 30 days’ perceptual learning treatment are in BA 17, 18 and 19. We can see that the improvement of BA 18 and 19 are more remarkable to some extent. BA 17 is also called the striate cortex, which is the primary visual cortex. BA 18 and BA 19 belong to the extrastriate cortex, where the visual information is further processed after primary processing in BA 17. As the higher functional visual cortex, BA18 and BA19 are responsible for synthesising complex visual information to acquire normal visual perception. These areas have more complex structures and functions, and could be easier to injure during the development of amblyopia. On the other hand, they might be more sensitive to the treatment of perceptual learning.

    We also found increased activation in the temporal lobes bilaterally. In humans, the temporal lobe has a role in hearing, vision, memory, and emotion. It has been proved that the middle temporal area (MT/V5) plays an important role in visual motion16 and that eye gaze is regulated by the superior temporal sulci.17 Our results suggest that there might be some compensation and remodelling in the temporal lobe after treatment for amblyopia which needs further study to be confirmed. In addition, we noticed increased activation delivered by the amblyopic eye in the right cingulate gyrus. The role the cingulate gyrus plays in the visual pathway is not clear and we are not sure about what exactly the increased activation indicates. Maybe it is related to the regulation of eye gaze or some other functions indirectly connected to visual ability. Although there has been some research reporting the dysfunction of the lateral geniculate nucleus (LGN) in amblyopia,18 ,19 we did not find activation of the LGN in our patients. This may be attributed to the deep location and small volume of the LGN, making the activation signal difficult to image even in a normal person.

    DTI is a useful method which can provide information on the white matter in vivo. It is believed that FA can provide microstructural information on white matter development,20 which reflects the integrity and directivity of the white matter. The ORs are myelinated fibres, which have a high degree of anisotropy,21 and this makes it possible to undertake tractography of the ORs based on DTI. However, the results from previous studies of amblyopia using the DTI technique are inconsistent.22–24 In our study, we successfully tracked the ORs of each subject. We found that the voxel numbers of left ORs were significantly larger than that of the right ORs, which agrees with some previous studies.22 ,23 This might prove the role of the left-higher-than-right asymmetry in the development of the bilateral ORs.

    We also made efforts to detect whether there are changes of the ORs after perceptual learning treatment in patients with this cause of amblyopia, which can help to explore more about the neural rehabilitation mechanism of this method for amblyopia. We found that there were no significant changes in average FA values, ADC values, voxel numbers or tracks’ numbers of ORs after 30 days’ perceptual learning. However, we noticed that the voxel numbers and tracks’ numbers were increased in both left and right ORs of the patients after treatment, although they did not reach statistical significance. It is our expectation that changes of the white matter might be more remarkable if a longer treatment period of perceptual learning is applied. We will test this hypothesis in our future experiments.

    Our study has some limitations, such as the small sample size and the lack of long term information about whether the visual changes were maintained. Moreover, further areas need to be explored, such as children who remain in the critical period, patients with strabismic amblyopia, etc.

    In the present study we investigated the effect of perceptual learning treatment on patients with amblyopia using BOLD-fMRI and DTI techniques. Our results indicated that perceptual learning treatment for amblyopia had a positive effect on the visual cortex and temporal lobe visual areas.

    References

    1. ↵
      1. Kanonidou E
      . Amblyopia: a mini review of the literature. Int Ophthalmol 2011;31:249–56.
      OpenUrlCrossRefPubMed
    2. ↵
      1. Chung ST,
      2. Li RW,
      3. Levi DM
      . Identification of contrast-defined letters benefits from perceptual learning in adults with amblyopia. Vision Res 2006;46:3853–61.
      OpenUrlCrossRefPubMedWeb of Science
    3. ↵
      1. Huang CB,
      2. Zhou YF,
      3. Lu ZL
      . Broad bandwidth of perceptual learning in the visual system of adults with anisometropic amblyopia. Proc Natl Acad Sci USA 2008;105:4068–73.
      OpenUrlAbstract/FREE Full Text
    4. ↵
      1. Levi DM,
      2. Li RW
      . Perceptual learning as a potential treatment for amblyopia: a mini-review. Vision Res 2009;49:2535–49.
      OpenUrlCrossRefPubMedWeb of Science
    5. ↵
      1. Astle AT,
      2. Webb BS,
      3. McGraw PV
      . Can perceptual learning be used to treat amblyopia beyond the critical period of visual development? Ophthalm Physiol Opt 2011;31:564–73.
      OpenUrlCrossRef
    6. ↵
      1. Alexander AL,
      2. Lee JE,
      3. Lazar M,
      4. et al
      . Diffusion tensor imaging of the brain. Neurotherapeutics 2007;4:316–29.
      OpenUrlCrossRefPubMedWeb of Science
    7. ↵
      1. Cleland BG,
      2. Crewther DP,
      3. Crewther SG,
      4. et al
      . Normality of spatial resolution of retinal ganglion cells in cats with strabismic amblyopia. J Physiol (Lond) 1982;326:235–49.
      OpenUrlPubMedWeb of Science
    8. ↵
      1. Crewther DP,
      2. Crewther SG,
      3. Cleland BG
      . Is the retina sensitive to the effects of prolonged blur? Exp Brain Res 1985;58:427–34.
      OpenUrlPubMedWeb of Science
    9. ↵
      1. Hess RF,
      2. Baker CL
      . Assessment of retinal function in severely amblyopic individuals. Vision Res 1984;24:1367–76.
      OpenUrlCrossRefPubMedWeb of Science
    10. ↵
      1. Hess RF,
      2. Baker CL
      . Human pattern-evoked electroretinogram. J Neurophysiol 1984;51:939–51.
      OpenUrlAbstract/FREE Full Text
    11. ↵
      1. Hess RF
      . Amblyopia: site unseen. Clin Exp Optom 2001;84:321–36.
      OpenUrlPubMed
    12. ↵
      1. Muckli L,
      2. Kiess S,
      3. Tonhausen N,
      4. et al
      . Cerebral correlates of impaired grating perception in individual, psychophysically assessed human amblyopes. Vision Res 2006;46:506–26.
      OpenUrlCrossRefPubMedWeb of Science
    13. ↵
      1. Conner IP,
      2. Odom JV,
      3. Schwartz TL,
      4. et al
      . Monocular activation of V1 and V2 in amblyopic adults measured with functional magnetic resonance imaging. J AAPOS 2007;11:341–50.
      OpenUrlCrossRefPubMedWeb of Science
    14. ↵
      1. Ho CS,
      2. Giaschi DE
      . Low- and high-level motion perception deficits in anisometropic and strabismic amblyopia: evidence from fMRI. Vision Res 2009;49:2891–901.
      OpenUrlCrossRefPubMedWeb of Science
    15. ↵
      1. Li C,
      2. Cheng L,
      3. Yu Q,
      4. et al
      . Relationship of visual cortex function and visual acuity in anisometropic amblyopic children. Int J Med Sci 2012;9:115–20.
      OpenUrlCrossRefPubMed
    16. ↵
      1. Watson JD,
      2. Myers R,
      3. Frackowiak RS,
      4. et al
      . Area of V5 of the human brain: evidence from a combined study using positron emission tomography and magnetic resonance imaging. Cereb Cortex 1993;3:79–94.
      OpenUrlAbstract/FREE Full Text
    17. ↵
      1. Hoffman EA,
      2. Haxby JV
      . Distinct representations of eye gaze and identity in the distributed human neural system for face perception. Nat Neurosci 2000;3:80–4.
      OpenUrlCrossRefPubMedWeb of Science
    18. ↵
      1. Miki A,
      2. Liu GT,
      3. Goldsmith ZG,
      4. et al
      . Decreased activation of the lateral geniculate nucleus in a patient with anisometropic amblyopia demonstrated by functional magnetic resonance imaging. Ophthalmologica 2003;217:365–9.
      OpenUrlCrossRefPubMedWeb of Science
    19. ↵
      1. Hess RF,
      2. Thompson B,
      3. Gole G,
      4. et al
      . Deficient responses from the lateral geniculate nucleus in humans with amblyopia. Eur J Neurosci 2009;29:1064–70.
      OpenUrlCrossRefPubMedWeb of Science
    20. ↵
      1. Hermoye L,
      2. Saint-Martin C,
      3. Cosnard G,
      4. et al
      . Pediatric diffusion tensor imaging: normal database and observation of the white matter maturation in early childhood. Neuroimage 2006;29:493–504.
      OpenUrlCrossRefPubMedWeb of Science
    21. ↵
      1. Mamata H,
      2. Mamata Y,
      3. Westin CF,
      4. et al
      . High resolution line scan diffusion tensor MR imaging of white matter fiber tract anatomy. Am J Neuroradiol 2002;23:67–75.
      OpenUrlAbstract/FREE Full Text
    22. ↵
      1. Xiao MY,
      2. Xu J,
      3. Li YJ,
      4. et al
      . Study on the structure of visual pathway white matter in children with anisometropic amblyopia by diffusion tensor imaging. Ophthalmol CHN 2009;18:315–20. (in Chinese)
      OpenUrl
    23. ↵
      1. Xie S,
      2. Gong GL,
      3. Xiao JX,
      4. et al
      . Underdevelopment of optic radiation in children with amblyopia: a tractography study. Am J Ophthalmol 2007;143:642–6.
      OpenUrlCrossRefPubMedWeb of Science
    24. ↵
      1. Song HY,
      2. Qi S,
      3. Tang HH,
      4. et al
      . MR DTI and DTT study on the development of optic radiation in patients with anisometropia amblyopia. J Sichuan Univ (Med Sci Edi) 2010;41:648–51. (in Chinese)
      OpenUrl

    Footnotes

    • Contributors JZ performed the experiments, analysed the data, wrote the paper and collected clinical information. MC and LL conceived and designed the experiments. HZ, XZ and XL also took part in performing the experiments and analysing the data. MC and JG also contributed in writing the paper. LL was mainly responsible for obtaining informed consent, collecting clinical information, and treating the patients.

    • Funding This study is supported by the National Natural Science Foundation of China (Grant No.81071138).

    • Competing interests None.

    • Patient consent Obtained.

    • Ethics approval The study was approved by the Ethics Committees of Beijing Hospital.

    • Provenance and peer review Not commissioned; externally peer reviewed.