Because the visual system is organized in a highly specific way, in adulthood, i.e. beyond the critical period of early development, loss of visual functions due to stroke or brain trauma is generally considered to be irreversible. Perceptual dysfunctions, such as visual field defects in hemianopia following visual cortex lesions or rather diffuse visual field defects after optic neuropathy, are believed to be permanent, without real hope for improvement. Fortunately, several discoveries made during the last decade have unveiled previously unrecognized potentials of residual, visual functions which can be found even deep in the blind field or especially at the border regions of the blind zones with considerable restoration potential.
The traditional view of an “inflexible” visual system is giving way to a view stressing the potential for plasticity and self-repair.1 In fact, the visual system can adapt to change rather well, both in the normal (during learning) and in the damaged brain. This flexibility is termed “neuroplasticity” and it can be observed at all levels of information processing (retina, lateral geniculate nucleus of the thalamus and primary visual cortex), even in old age.1,2
One example of plasticity is the ability of the visual system to recover from damage. The potential for spontaneous recovery of visual functions after damage is considerable (though still somewhat limited) in patients with pre- or postchiasmatic visual field defects that have sustained stroke, traumatic brain injury, tumors or vascular lesions. Zhang et al.3 investigated spontaneous recovery of visual function in postchiasmatic lesioned patients and found that the greatest spontaneous functional improvement (recovery) occur within the first weeks up to six months after the damage explaining much of the variability in perimetry results within this period.3
Looking at a hemianopia case (Figure 1), the visual field may be subdivided into regions of absolute blindness (black) or those of “normal” vision (white). In addition, there are “areas of residual vision (ARV)” which have been known before as “relative defects”. These are typically located at the border of the scotoma, i.e. in a “transition zones”. ARV are the functional representation of regions of the visual pathway (e.g. the primary visual cortex) which are only partially injured, thus being subject to different influences in their functional activation state. Whether these areas are active (functional) or not at a given point in time, i.e. their variability, is influenced by different factors that determine either “favorable” or “unfavorable” conditions.
One factor contributing to this variability is diurnal variation, i.e. systematic or unsystematic fluctuations in the psychophysiological state of the visual system or retinal noise. Zihl et al.4 observed diurnal variations of visual field defect size in patients with post-retinal lesions and suggested a modulation of neuronal sensitivity coupled to a circadian oscillator being the cause for this variability. ARV are particularly prone to sensitivity changes during the course of the day and they may increase with higher levels of alertness and decrease with fatigue. Here, the thresholds of light detections may be markedly elevated in the corresponding intact zones of the visual field. In our laboratory there was a particularly interesting case of a patient who had suffered partial visual field loss with improvements of the visual field after carrying out vision training for many months. When he showed up for a follow-up investigation one day, his performance had dropped unexpectedly to pre-training levels and he did poorly in his visual fields. An experienced neurologist suggested that he go to the cafeteria and have a coffee and a bite to eat. Afterwards the patient was back at the level of performance he had on the other days. This is an example how the state of vigilance (perhaps influenced by a temporary hypoglycaemic state) had a direct impact on his vision performance in ARV.
Another factor contributing to visual field variability is local attention which was documented by the following study: Poggel et al.5 asked patients to particularly focus their attention on ARV by placing a cue (a target window) onto ARV in the border region of the visual field defect. This led to an immediate shift of the visual field border which reverted back to the pre-attention performance level once the cue-task was no longer carried out.
While spontaneous fluctuations are an expression of residual capacities of the damaged brain, residual functions are visible only under optimal physiological conditions. Therapeutic progress is only achieved if such residual functions can be improved permanently. In recent years there has been some progress in this field. It was found that even after spontaneous recovery has reached a plateau, plasticity could be initiated again by repetitive stimulation of residual visual field areas using visual field training.6, 7 Also, as shown recently, transorbital electric stimulation can be used to improve visual functions.8 Systematic, repeated stimulation of areas of residual vision with light stimuli, e.g. by visual field training, led to functional gains observed with both standard and high resolution perimetric measurements and improvements were achieved primarily in ARV.7Based on these observations we proposed that a basic mechanism of vision restoration is an increase of neuronal activity in ARV where injured neurons mediate residual vision at a sub-threshold capacity and where training increases the function of these partially damaged areas.6 This was confirmed by recent studies where computer-simulations using data-mining methods confirmed that, indeed, areas of visual field improvements (“hot spots”) are primarily found in regions with considerable residual vision.9
Figure 1 shows the results of a computer-based perimetric evaluation of the visual field before and after visual field training. Here, supra-threshold stimuli were presented and patients were required to hit a computer space bar whenever stimuli were presented. The patients either detected or missed the stimuli and this information was used to generate visual field charts on the computer. By superimposing three to five of such visual field charts, ARV could be documented where the stimuli were missed sometimes but not always. Such areas where stimulus detection was inconsistent (“relative defects”) were then displayed in the chart by grey squares. One may argue that eye movements away from fixation explain such inconsistent performances and any changes in visual field borders (such as presumed improvements) are not the results of true functional improvements but rather an eye movement artefact. To study this possibility we have therefore measured the eye positions with an eye-tracker and adjusted the visual field charts to where the eye was located for each of the stimulus positions.
|When the fixation accuracy of the patient deviated from the fixation point the presented stimulus position was recalculated and accordingly adjusted. Adjusted stimulus positions show high topographic congruency with non-adjusted stimulus positions. Increased stimulus detection and thus improved visual field was concentrated on the border zone of the visual field where residual functions are located.|
As Figure 1 shows, after but not before visual field border training supra-threshold light stimuli were detected in the “absolutely” blind visual field regions and also in areas of residual vision. This suggests that eye movements cannot explain the new responses in the blind field which is in agreement with our prior studies showing that visual field training effects cannot be explained by eye movements.10 Also Bergsma & Van der Wildt11 recently observed not only increased light stimulus detection but also improved visual acuity, critical flicker fusion frequency and color vision in the regained visual field areas after visual field training with simple light stimuli.
The ultimate test for vision restoration is its documentation by functional imaging of the brain before vs. after therapy. One would expect that functional changes go along with cortical activation pattern changes (re-activation), particularly in areas of residual vision. Marshall et al.12 measured the blood oxygenation level dependent (BOLD) in six patients with chronic homonymous hemianopia in areas of residual vision (border zone) and at an intact seeing position before and after one month of visual field training. They observed an alteration of brain activity which was associated with an attentional shift from the intact visual field towards the area of residual visual function at the borderzone of the blind field. This suggests that attention plays a central role in vision restoration and this is also what was observed in our laboratory, where repeated training with localized attention (induced by cues) was found to permanently improve visual functions in those attended locations.13
The concept of post-lesion plasticity in the adult visual system is now becoming increasingly accepted. The following statement is worth quoting which was made by Torsten Wiesel, recipient of the Nobel Prize for his discovery of how the visual system works. He stated:
”Restoration of vision after damage is an issue I am very interested in and I think that there is progress; to find different means of restoring visual functions are very interesting and encouraging ... My experiments (on receptive field plasticity) are hard evidence that it is possible to restore (visual) function through time. In this case we did not make any special effort by stimulating the eyes, in a similar way that Bernhard Sabel has done with clinical cases … trying to restore visual function. This kind of experiment and also from the clinical work gives you hope that there is more to learn … that it should be possible to have patients restore vision in spite of an initial apparent lack of vision. As we know from earlier work, plasticity in a monkey is pretty much over after one year (of age), and (our) monkeys were older. So we can look upon this experiment that receptive field reorganization happens in the adult (monkey brain) similar to the experiments in clinical cases” (quoted from a lecture on “Specificity and Plasticity of the Visual System”, April 6, 2005 at the “Symposium on Restoration of vision after brain damage”, presented at the meeting “VISION 2005” in London, Royal National Institute of the Blind (RNIB), chaired by Bernhard Sabel and Torsten Wiesel).
While short-term improvements similar to those observed during spontaneous performance fluctuations might be scientifically interesting, it is of far greater interest to look for evidence of vision restoration and to find clinically useful means to achieve long-lasting functional changes with as little effort as possible. So far, many months of training (150 days) are required to achieve good restoration with clinical benefits.6, 7, 13 Efforts are now underway to develop new methods to speed up restoration even more. We are currently evaluating non-invasive electrical brain stimulation to achieve visual field improvements in as little as 10-20 days of therapy. Initial findings are encouraging and suggest that such a goal may actually be achievable.14 Before properly controlled clinical trials with positive outcome are published, however, it remains open if such methods can be recommended for routine clinical use.
In summary, it is clear that the visual system, thought to be irreparably damaged after brain injury, has a plasticity potential of its residual functions that basically does not differ much from plasticity that are well known in other functional systems. The brain is not only able to adapt to the damage by spontaneous recovery and repair, but it is rather receptive to rehabilitation by vision training or by non-invasive electrical brain stimulation.
- Huxlin KR. Perceptual plasticity in damaged adult visual systems. Vision Res 2008;48:2154-66.
- Pascual-Leone A, Amedi A, Fregni F, et al. The plastic human brain cortex. Annu Rev Neurosci 2005;28:377-401.
- Zhang X, Kedar S, Lynn MJ, et al. Natural history of homonymous hemianopia. Neurol 2006;66:901-5.
- Zihl J, Pöppel E, Cramon von D. Diurnal variation of visual field size in patients with postretinal lesions. Exp Brain Res 1977;27:245-9.
- Poggel DA, Kasten E, Müller-Oehring EM, Bunzenthal U, Sabel BA. Improving residual vision by attentional cueing in patients with brain lesions. Brain Res 2006;1097:142-148
- Kasten E, Wuest S, Sabel BA. Residual vision in transition zones in patients with cerebral blindness. J Clin Exp Neuropsy 1998;20:581-98.
- Sabel B A, Kenkel S, Kasten, E. Vision restoration therapy (VRT) efficacy as assessed by comparative perimetric analysis and subjective questionnaires. Restor Neurol Neurosci 2004;22:399-420.
- Fujikado T, Morimoto T, Matsushita K, et al. Effect of transcorneal electrical stimulation in patients with nonarteritic ischemic optic neuropathy or traumatic optic neuropathy. Jpn J Ophthalmol 2006;50:266–73.
- Guenther T, Mueller I, Preuss M, Kruse R, Sabel B. Treatment Outcome Prediction Model of Visual Field Recovery Using Self-Organizing-Maps. IEEE Trans Biomed Eng. 2008. [Epub ahead of print]
- Kasten E, Bunzenthal U, Sabel BA. Visual field recovery after vision restoration therapy (VRT) is independent of eye movements: an eye-tracker study, Behav Brain Res 2006;75:18-26.
- Bergsma DP, Van der Wildt GJ. Properties of the regained visual field after visual detection training of hemianopsia patients. Restor Neurol Neurosci 2008;26:365-75.
- Marshall RS, Ferrera JJ, Barnes A, et al. Brain activity associated with stimulation therapy of the visual borderzone in hemianopic stroke patients. Neurorehabil Neural Repair 2008;22:136-44.
- Poggel DA, Kasten E, Sabel BA. Attentional cueing improves vision restoration therapy in patients with visual field defects. Neurol 2004;63:2069-76.
- Sabel BA, Gall C, Jobke S, Chibisova AN, Chibisova JM, Fedorov AB. Restoration of vision with non-invasive electrical brain synchronization after optic nerve damage, Abstract for International Conference of Transcranial Stimulation, 01. - 04.10.2008, Göttingen