In this paper, we show that in the absence of vision since birth, the brain adapts and allows extraneous information coming from other senses, to compensate for the lost sense. We provide scientific data showing that the tongue is a viable organ to transfer electrotactile information to the visual cortex in the visually-deprived brain. This calls for a concept termed sensory substitution and refers to the replacement of a lost sense (in this case vision) by another one (audition or touch) while preserving the key features of the lost sense. Many animal and human studies have so far proven the existence of sensory substitution in the blind and have unravelled some of its neural bases through a mechanism of cross-modal plasticity .
It is well known that if brain damage occurs during development, abnormal neuronal connectivity patterns can develop. If, for example, in hamsters central retinal targets are destroyed at a very early age, new and permanent retinofugal projections into non-visual sites such as the thalamic auditory nucleus are formed  (Fig. 1A). These rewired hamsters possess visually responsive neurons within their auditory cortex, whose physiological properties (e.g., motion and orientation selectivity) are similar to those encountered in a normal visual cortex .
These new connections to the auditory cortex are also functional at the behavioural level. Indeed, rewired hamsters can learn visual discrimination tasks as well as normal animals and a lesion of the auditory cortex abolishes this function  (Fig.1B). In fact, rewired hamsters with auditory cortex lesions exhibit cortical blindness similar to non-rewired hamsters with visual cortex lesions. It therefore seems that the auditory cortex of early blind animals does acquire properties concerned with the lost sense through cross-modal plasticity.
Fig. 1: Animal studies:
A. Formation of retinal projection in MGB illustrated with intraocular injection of CTB.
B. Behavioral training of visual discriminations.
Visual deprivation from birth also leads to profound modifications in the architecture of the human brain that have an impact on various behaviours. In recent years, many sensory substitution devices have been developed to help blind people cope with their visual loss. These devices usually convert visual images into sound (such as the vOICe system®) or touch (such as the Tongue display Unit-TDU). Tactile display methodologies are the most widely used and are thoroughly described in Grunwald,M (2008) . In our studies on cross-modal plasticity in human blindness, we have widely used the TDU as a mean to translate visual information into electro-tactile pulses applied to the tongue (Fig. 2A). This translated visual input into touch is then funnelled to the brain [1, 12].
Substituting vision with touch
In a first experiment, we trained a group of congenitally blind and blindfolded sighted control subjects to use the TDU in an orientation discrimination task . We measured brain activity using positron emission tomography (PET) before and after a 1-week training period with the TDU. Both groups learned the task equally well, although the blind tended to be faster than the sighted participants. As expected, the PET data before training did not show activations in visual cortical areas in either group. In sharp contrast, after training, blind but not blindfolded sighted control subjects activated large parts of their visual cortex. Interestingly, the activated clusters in the occipital and occipitoparietal areas showed a strong resemblance with the areas reported to be activated when sighted subjects perform a visual orientation task.
The visual system is classically subdivided into a dorsal “where” and a ventral “what” pathway. We therefore addressed the question whether the basic architecture of a dorsal and ventral pathway is preserved in subjects lacking vision from birth since. To that end, we used functional magnetic resonance imaging (fMRI) while participants used the TDU in tasks designed to activate either the dorsal or the ventral visual pathway. In a first study, congenitally blind and sighted participants were trained to detect the motion direction of random dot patterns [6, 14]. Stimuli were either moving in a coherent manner (left, right), randomly or remained static. The fMRI data showed that following training, blind subjects activated large parts of the dorsal extrastriate visual pathway including the motion sensitive area hMT + (Fig. 2B, left). These data suggest that the recruitment of the hMT + cpmplex is not mediated by visual-based mental imagery and that visual experience is not necessary for the development of this cortical system. In a subsequent fMRI study, we trained blind and blindfolded sighted subjects to use the TDU in a shape recognition task. Participants were presented four different shapes (a triangle, rectangle, square and the letter E) and they had to indicate which of the four shapes had been presented. In line with our hypothesis, the fMRI data showed that during non-haptic shape recognition, blind subjects activated their primary visual cortex and key areas of the ventral visual stream, including the inferotemporal cortex, lateral occipital area and fusiform gyrus  (Fig. 2B, right).
The above results lead to the following conclusions. First, the segregation of the efferent projections of the primary visual cortex into a dorsal and ventral visual stream is preserved in people blind from birth. Second, cortical “visual” association areas are capable of processing and interpreting information carried by non-visual sensory modalities.
Fig. 2: Integrity of the two visual streams in the congenitally blind:
A. The motion (dorsal) stream.
B. The shape (ventral) stream.
What about navigation?
Vision is undoubtedly an important facilitator of navigation. The access to visual information explains why sighted individuals can easily select a navigational path through a hallway scattered with obstacles. Avoiding obstacles and creating a cognitive map of the environment is obviously more difficult in the absence of vision and remains one of the greatest navigational challenges faced by blind individuals. Notwithstanding, congenitally blind subjects are able to generate spatial representations, probably through tactile, auditory, and olfactory cues, as well as motion-related cues arising from the vestibular and proprioceptive systems, and they preserve the ability to recognize a travelled route and to represent spatial information mentally .
We tested the potential of the TDU in a series of behavioural and brain imaging studies. In a first study, we trained a group of blind and blindfolded sighted control subjects to use the TDU in a life-size obstacle course . The obstacle course was composed of two hallways in which obstacles of different sizes and shapes were placed. Although both groups learned to detect and avoid the obstacles, blind subjects performed significantly better than the sighted. These data underscore the potential of the TDU as a navigational aid in people lacking vision from birth.
The neural correlates of navigation in the blind have remained elusive. This is largely due to the technical difficulty of designing a navigation task for blind subjects wihin the context of a brain imaging paradigm. We circumvented this difficulty by presenting a navigational task using the TDU . We trained congenitally blind and blindfolded sighted participants in a route navigation and a route recognition task. Once fully trained, participants repeated the route recognition task inside the MRI scanner. The fMRI data revealed that during route recognition, blind subjects activated large parts of the visual cortex, the right parahippo- campus, posterior parietal cortex and pre-cuneus (Fig. 3A). In sharp contrast, blindfolded sighted controls did not show taskdependent BOLD signal increases in the parahippocampus or in the visual cortex; instead they activated more frontal cortical areas not seen in blind subjects In a second fMRI experiment, we demonstrated that the areas activated by the blind participants are the same as those activated by sighted subjects when they did the same navigational task under full vision (Fig. 3B). These data suggest not only cross-modal plasticity in spatial coding but also that visual experience is not necessary for the development of spatial navigation networks in the brain.
Fig. 3: Neural correlates of navigation. A. Flat map of the cortex showing the visual areas and the parahippocampus in the blind. B. Flat map of the cortex in subjects doing the same tasks in full vision. Note the similarities between the various brain activations.
Anatomy of the blind human brain
How does absence of vision from birth affect the organisation of the human brain and through which pathways can non-visual information be funnelled to the occipital cortex in the visually deprived brain? Using whole brain voxel-based-morphometry (VBM) analysis, we demonstrated a significant grey matter (GM) atrophy of all structures of the visual pathways, including the lateral geniculate and posterior pulvinar nuclei, the striate and extra-striate visual areas and parts of the ventral and dorsal streams  (Fig. 2A). These volume reductions ranged from 25% in the primary visual cortex up to 20% in extrastriate visual areas . Volume reductions also occur in non-visual areas such as the hippocampus . Besides the volume reductions in GM, congenitally blind subjects show increases in cortical thickness in the cuneus (Fig. 2B) that are likely due to a reduction in cortical pruning in the early maturation process of the cortex as a consequence of the loss of visual input.
Changes in white matter include atrophy of the optic tracts and optic chiasm, the optic radiations, the splenium of the corpus callosum and the inferior longitudinal fasciculus , a pathway connecting the occipital cortex with the temporal lobe. No studies found direct evidence for the establishment of new pathways, although volume increases in the occipito-frontal fasciculus, the superior longitudinal fasciculus and the genu of the corpus callosum have been reported .
There is also indirect evidence for an increased functional connectivity between parietal and visual areas in the blind . Taken together, since no de novo tracts have been demonstrated in congenitally blind subjects, the data suggest that cross-modal functionality of the visual cortex in early blindness is primarily mediated by preserved or strengthened cortico-cortical connections.
In this paper, we summarized recent findings from our laboratory in animals and humans on the anatomical and functional changes that take place in the brain lacking vision from birth. Our studies with the TDU emphasize that the functional organization of extra-striate visual cortical areas remains largely preserved in congenitally blind individuals. These studies seem to concur that the blind brain should not be considered as a “disabled” brain but rather as a truly “differentially able” brain.