Seeing through the tongue: cross-modal plasticity in the congenitally blind

International Congress Series 1270 (2004) 79 84 Seeing through the tongue: cross-modal plasticity in the congenitally blind Ron Kupers a, *, Maurice Ptito b www.ics-elsevier.com a Center for Functionally Integrative Neuroscience (CFIN), Aarhus University, and Aarhus University Hospital, Noerrebrogade 44, 8000 Aarhus Denmark b School of Optometry, Université de Montréal, CP 6128 Montreal H3C 3J7, QC, Canada Abstract. Sensory substitution refers to the capacity of the brain to replace the functions of a lost sense by another sensory modality. This cross-modal plasticity has been documented both in animals and humans deprived of a particular sensory modality, such as vision or audition. Discovering new ways to exploit this cross-modal plasticity may help to optimize the recovery of sensory loss. The most commonly used form of sensory substitution is Braille reading, which enables the blind to read by using the somatosensory system. Recently, a human machine interface, the tongue display unit (TDU), which uses the tongue as a medium for visual substitution in blind subjects, has been developed. We trained six congenitally blind and five blindfolded, sighted controls to use the TDU to perform a visual orientation discrimination task. Subjects were positron emission tomography (PET) scanned before and after an intensive 1-week training program with the TDU. Before training, no increased activity was measured in the visual cortex of either group during the orientation detection task. However, after training, patterned stimulation of the tongue activated the visual cortex in congenitally blind subjects. Sighted controls did not show occipital activation post-training despite equivalent performance on the same task. These data reveal the development of cross-modal plasticity in the brains of congenitally blind subjects. They further show that the time course of neuroplasticity in humans can be remarkably rapid. D 2004 Elsevier B.V. All rights reserved. Keywords: Cross-modal plasticity; Congenitally blind; Transcranial magnetic stimulation; Occipital cortex; Sensory substitution 1. Introduction Sight is probably the most important of our senses. Since we live in a very visual world, the loss of vision is one of the most incapacitating events that may overcome a person. It is therefore not surprising that many attempts have been undertaken to develop artificial forms of vision. The best-known example of artificial vision is Braille reading. Although Braille is an important tool for the blind, a major limitation is that it does not allow one to convey information from objects placed outside the egocentric space. * Corresponding author. Tel.: +45-8949-3081; fax: +45-8949-4400. E-mail address: ron@pet.auh.dk (R. Kupers). 0531-5131/ D 2004 Elsevier B.V. All rights reserved. doi:10.1016/j.ics.2004.04.053

80 R. Kupers, M. Ptito / International Congress Series 1270 (2004) 79 84 Over the past few decades, many efforts have been undertaken to develop alternative forms of artificial vision. Some of these approaches are based on highly invasive procedures, such as electrical stimulation of the retina [1] or visual cortex [2] and stem cell transplants in the eyes [3]. However, new sensory substitution systems replacing vision with touch or audition have been developed recently, offering valuable non-invasive alternatives [4 7]. 2. Tactile vision sensory substitution (TVSS) systems The system that we have been using is based upon the principle of tactile image projection and was developed by Bach-Y-Rita [5]. The system uses the tongue as a substrate for electrotactile stimulation. There are several good reasons to use the tongue as a substrate for electrotactile stimulation. First, the tongue is permanently covered with saliva and the sensory receptors are located close to the surface. As a result, stimulation can be applied with much lower voltage and current than is required for the skin. Second, the tongue is more densely populated with touch-sensitive nerves than most other parts of the body. Therefore, the tongue can convey higher-resolution information than the skin can [8]. Finally, the tongue is in the protected environment of the mouth and is normally out of sight and out of the way, which makes it cosmetically acceptable. 3. Behavioral training and positron emission tomography (PET) study 3.1. Methods Six congenitally blind and five blindfolded, sighted control subjects participated in a behavioral training program and a PET study that was conducted before and after the behavioral training program. During behavioral training, subjects learned to use a TVSS system to discriminate the orientation of a series of Snellen patterns. Our system consists of a laptop, a tongue display unit (TDU), an electrode array and special image processing software. The electrode array measures 3 3 cm and consists of 144 gold-plated electrodes arranged in a 12 12 square matrix. Subjects had to detect the orientation of Snellen T s, which were randomly presented in one of four possible orientations. The images were presented on a laptop and the subjects could use a computer mouse to explore the image (Fig. 1). Electrical pulses were generated when the cursor was superimposed on any of the pixels on the screen forming the letter T. The first PET session (see further) took place before the onset of training. Thereafter, the behavioral training program started and subjects underwent a 1-h daily training session with the TDU on the orientation task. The criterion that was set for successful learning was an 85% correct response on two successive training sessions. Once subjects had reached the criterion, they were PET scanned for the second time. Subjects were scanned under the following three conditions in a semi-random order: (1) Rest: the electrode array was placed on the tongue with no stimulation; (2) Noise: subjects were presented with a random noise pattern without meaningful shape and they were asked to detect potential changes in the stimulus; and (3) Orientation task: tumbling T s were randomly presented in four different orientations and subjects had to indicate the correct orientation using the left thumb. The presentation of the tumbling T s started simulta-

R. Kupers, M. Ptito / International Congress Series 1270 (2004) 79 84 81 Fig. 1. Experimental setup. Subject uses a computer mouse to explore the image presented on the laptop. Insert shows the TDU control box and the electrode array with 144 contacts. neously with tracer injection and was continued throughout the rest of the scanning time. Each scanning condition was repeated four times. The Aarhus Ethics Committee approved the study and subjects gave written, informed consent. 3.2. Results and discussion During the first PET scanning session and at the beginning of training, subjects performed around chance level on the orientation detection task. The subjects performance rapidly improved over time. At the end of the 1-week behavioral training period, there was no statistically significant difference in performance between both groups. Prior to training, significant increases in regional cerebral blood flow (rcbf) during the orientation detection task in both groups were restricted to parietal and prefrontal areas. In sharp contrast, the blind showed significant activations in large portions of occipital, occipito-parietal and occipito-temporal cortices post-training (Fig. 2). Blindfolded controls did not activate the visual cortex post-training despite equal performance on the orientation detection task. It is also noteworthy that the posterior parietal activation was significantly stronger in the congenitally blind following training. These results indicate that the visual cortex is recruited in congenitally blind subjects who have learned to use the tongue in a visual orientation task. In contrast, trained, sighted, blindfolded controls showed a strong activation of the tongue area of the somatosensory cortex and of the prefrontal cortex. These results are reminiscent of those reported by others in a Braille reading task in blind subjects [9 11]. A difference with

82 R. Kupers, M. Ptito / International Congress Series 1270 (2004) 79 84 Fig. 2. PET results after training. Blind subjects activated large parts of the occipital cortex during the orientation detection task. RCBF increases in trained, blindfolded controls were restricted to parietal and prefrontal areas. our study, however, is that blind subjects in these studies were already fluent in Braille reading at the time of scanning, thus making it impossible to delineate the time course of the plastic processes. Our results demonstrate that such plasticity can occur in a matter of hours. It could be argued that mental imagery is at the basis of the visual cortex activation in blind subjects [12]. This is unlikely in the present experiment because our congenitally blind subjects never had any visual experience and never reported engaging in visual imagery. An alternative hypothesis is that the visual cortex activation in early blind subjects is mediated by the superior parietal lobe [10]. Tactile information is processed in the anterior part of the superior parietal cortex (area 7a), whereas visual information is processed more posteriorly, in area 7b. Therefore, tactile information may reach the visual system through increased connectivity between areas 7a and 7b after loss of vision, as suggested by the stronger activation of superior parietal areas in the blind. A pathway involving the lateral occipital tactile-visual region may also play a role in the activation of the visual cortex. This area is activated by visual and haptic exploration of objects [13]. Our results are, therefore, consistent with the hypothesis that somatosensory stimulation in the blind activates the occipito-temporal region and further expands into earlier retinotopic areas. 4. Trancranical magnetic stimulation (TMS) study In order to study the functional relevance of the PET activations described above, we performed a TMS study in three early blind and five normal control subjects trained with the TDU device. Since preliminary TMS studies on some normal volunteers revealed possible contaminating attentional and motor effects during application of TMS pulses, we chose for a design using low frequency repetitive TMS (rtms) with behavioral assessment after the TMS sequence. Previous studies showed long aftereffects of at least 20 min after a low-frequency rtms train of 15 min over the occipital cortex [14]. We used a Magstim RapidR magnetic stimulator connected to a double 9.0-cm figure- 8-shaped coil with a maximal stimulator output of 1.2 T. We first identified the phosphene and motor thresholds using single TMS pulses. For occipital stimulation, the coil was

R. Kupers, M. Ptito / International Congress Series 1270 (2004) 79 84 83 positioned in a vertical position with its inferior limit 1 cm above the inion. For parietal stimulation, the coil was placed 2 cm anterior to the interaural line and 9 cm from the vertex on the left hemisphere, corresponding with the known representation of the tongue area [15]. Stimulus intensity was set to 110% of phosphene or motor threshold. One Hz rtms was applied without interruption for 15 min (900 pulses). Before and after the TMS sequence, subjects performed the orientation detection task using the TDU. We measured reaction time and percentage of correct responses. Behavioral testing post-tms started 2 min after the end of the TMS sequence. No significant effects of rtms on task performance were found in either group for both parietal and occipital rtms. Reaction times and percentage of correct responses were not significantly different pre- and post-rtms. In order to test whether a more dorsal stimulation site in the occipital cortex might produce different results, we retested the blind subjects with the coil positioned 2.5 cm above the inion. Also for this coil position, there was no interference with task performance. These data are different from those reported by Cohen et al. [16] showing interference with Braille reading when the occipital cortex was stimulated in early blind subjects. There are several important methodological differences between our study and the study by Cohen et al., which may explain the discrepancy in results. Despite these negative data, we obtained some highly interesting results during the phosphene threshold assessment. In two of the congenitally blind subjects, single pulse TMS over the occipital cortex induced clear tactile sensations in the tongue. We mapped the area over which tactile phosphenes could be induced, and found a somatotopic representation of the tongue area with the left part of the tongue represented in the left occipital cortex, right part of the tongue in the right occipital cortex and middle part of the tongue on the medial occipital cortex. No such tactile phosphenes were elicited in any of the five trained, sighted subjects. These preliminary TMS data confirm our PET findings of the development of cross-modal plasticity and cortical reorganization in the early blind [17,18]. 5. Conclusion Our results reveal that the tongue can be used as a substitute for the eye and can act as a portal to the visual cortex of congenitally blind people. The TDU has the important advantage of being hands-free and providing sensory input corresponding to distal stimuli positioned beyond egocentric space. It would represent a major breakthrough in the quality of life of the blind. Acknowledgements This study was supported by a grant from the Danish Medical Research Council, the Danish Grundforskningsfonden and the FRSQ-Vision Network (Qc). We are indebted to Dr. E. Sampaio (Université de Strasbourg, France) for lending us the TDU device, S. Moesgaard (Aarhus University) for practical help during training of the subjects, Dr. P.-E. Buchholtz Hansen (Psychiatric Hospital Aarhus) for giving us access to his TMS laboratory and Dr. A. Fumal (University of Liège, Belgium) for helping us with the TMS experiments.

84 R. Kupers, M. Ptito / International Congress Series 1270 (2004) 79 84 References [1] M.S. Humayun, et al., Visual perception in a blind subject with a chronic microelectronic retinal prosthesis, Vis. Res. 43 (2003) 2573 2581. [2] E.M. Schmidt, et al., Feasibility of a visual prosthesis for the blind based on intracortical microstimulation of the visual cortex, Brain 119 (1996) 507 522. [3] I. Fine, et al., Long-term deprivation affects visual perception and cortex, Nat. Neurosci. 6 (2003) 1 2. [4] P. Bach-y-Rita, et al., Vision substitution by tactile image projection, Nature 221 (1969) 963 964. [5] P. Bach-Y-Rita, S.W. Kercel, Sensory substitution and the human machine interface, Trends Cogn. Sci. 7 (2003) 541 546. [6] C. Capelle, et al., A real time experimental prototype for enhancement of vision rehabilitation using auditory substitution, IEEE Trans. Biomed. Eng. 45 (1998) 1279 1293. [7] P. Arno, et al., Occipital activation by pattern recognition in the early blind using auditory substitution for vision, NeuroImage 13 (2001) 632 645. [8] R.L. Ringel, S.J. Ewanowski, Oral perception: 1. Two-point discrimination, J. Speech Hear. Res. 8 (1965) 389 398. [9] N. Sadato, et al., Activation of the primary visual cortex by Braille reading in blind subjects, Nature 380 (1996) 526 528. [10] C. Buchel, et al., Different activation patterns in the visual cortex of late and congenitally blind subjects, Brain 121 (1998) 409 419. [11] H. Burton, et al., Adaptive changes in early and late blind: a fmri study of Braille reading, J. Neurophysiol. 87 (2002) 589 607. [12] A. Aleman, et al., Visual imagery without visual experience: evidence from congenitally totally blind people, NeuroReport 12 (2001) 2601 2604. [13] A. Amedi, et al., Visuo-haptic object-related activation in the ventral visual pathway, Nat. Neurosci. 4 (2001) 324 330. [14] A. Fumal, et al., Effects of repetitive transcranial magnetic stimulation on visual evoked potentials: new insights in healthy subjects, Exp. Brain Res. 15 (2003) 332 340. [15] R.M. Rodel, R. Laskawi, H. Markus, Tongue representation in the lateral cortical motor region of the human brain as assessed by transcranial magnetic stimulation, Ann. Otol. Rhinol. Laryngol. 12 (2003) 71 76. [16] L.G. Cohen, et al., Functional relevance of cross-modal plasticity in blind humans, Nature 389 (1997) 180 183. [17] T. Kujala, K. Alho, R. Naatanen, Cross-modal reorganization of human cortical functions, Trends Neurosci. 23 (2000) 115 120. [18] J.P. Rauschecker, Cortical map plasticity in animals and humans, Prog. Brain Res. 138 (2002) 73 88.