A N400 ERP Study in letter recognition after passive tactile stimulation training in blind children and sighted controls

Abstract

Background:

We previously demonstrated that using a sensory substitution device (SSD) for one week, tactile stimulation results in faster activation of lateral occipital complex in blind children than in seeing controls.

Objective:

We used long-term haptic tactile stimulation training with an SSD to test if it results in stable cross-modal reassignment of visual pathways after six months, to provide high level processing of tactile semantic content.

Methods:

We enrolled 12 blind and 12 sighted children. The SSD transforms images to a stimulation matrix in contact with the dominant hand. Subjects underwent twice-daily training sessions, 5 days/week for six months. Children were asked to describe line orientation, name letters, and read words. ERP sessions were performed at baseline and 6 months to analyze the N400 ERP component and reaction times (RT). N400 sources were estimated with Low Resolution Electromagnetic Tomography (LORETA). SPM8 was used to make population-level inferences.

Results:

We found no group differences in RTs, accuracy of identifications, N400 latencies or distributions with the line task at 1 week or at 6 months. RTs on the letter recognition task were also similar. After 6 months, behavioral training increased accurate letter identification in both seeing and blind children (Chi 2 = 11906.934, p = 0.000), but the increase was larger in blind children (Chi 2 = 8.272, p = 0.004). Behavioral training shifted peak N400 amplitude to left occipital and bilateral parietal cortices in blind children, but to left precentral and postcentral and bilateral occipital cortices in sighted controls.

Conclusions:

Blind children learn to recognize SSD-delivered letters better than seeing controls and had greater N400 amplitude in the occipital region. To the best of our knowledge, our results provide the first published example of standard letter recognition (not Braille) by children with blindness using a tactile delivery system.

Abbreviations

ERP

Event Related Potentials

EEG

Electroencephalogram

SSD

Sensory Substitution Device

Reaction time

LORETA

Low Resolution Electromagnetic Tomography

BMA

Bayesian Model Averaging

1. Introduction

Neuroplasticity is a process in which neurons modify their interconnections, modulating their function as a consequence of experience and learning (Feldman & Brecht, 2005). Following sensory deprivation in one modality, individuals develop significant neural plastic changes in distributed cortical and subcortical areas (Merabet et al., 2008; Merabet & Pascual-Leone, 2010; Kim, Kanjlia, Merabet, & Bedny, 2017). Cross-modal plasticity is a well-known phenomenon in visually deprived individuals (Sathian & Stilla, 2010; Merabet & Pascual-Leone, 2010; Fiehler & Rösler, 2010; Bedny, 2017), and can be defined as the re-purposing of sensory cortex in the absence of its specific sensory input. Thus, primary sensory regions deprived their natural inputs are remapped to inputs coming from spared sensory sources (Bedny, 2017). The best established example is the adaptive recruitment of visually related areas in blind subjects by auditory or tactile stimuli (Amedi, Raz, Azulay, Malach, & Zohary, 2010; Ortiz et al., 2011; Sathian & Stilla, 2010; Bedny, 2017).

Evidence regarding the potential enhanced processing ability of blind individuals for tactile stimuli is controversial (Sathian & Stilla, 2010; Norman & Bartholomew, 2011), but is supported by enhanced passive tactile acuity (Goldreich & Kanics, 2003; Wong, Gnanakumaran, & Goldreich, 2011; Gurtubay-Antolin & Rodríguez-Fornells, 2017), superior handling non-spatial tactile distractors (Frings, Amendt, & Spence, 2011) and haptic discrimination (Norman & Bartholomew, 2011). Others have suggested that practice effects may explain the differences in tactile performance in blind subjects compared to sighted controls (Bhattacharjee, Ye, Lisak, Vargas, & Goldreich, 2010; Stronks, Nau, Ibbotson, & Barnes, 2015).

The brain of individuals with blindness may process tact of letters differently from other tactile non-linguistic spatial information (Sathian & Stilla, 2010; Ortiz Alonso et al., 2015), but processing may differ between early and late blind subjects (Voss, 2013). Rauschecker proposed that word perception begins in occipital areas (Rauschecker, 2011), while others consider bilateral dorsal intra-parietal regions as the origin of the letter-by-letter word reading (Cohen, Dehaene, Vinckier, Jobert, & Montavont, 2008). In addition, it has been demonstrated that the Visual Word Form Area, located in the left fusiform gyrus, responds selectively to letters in blind and sighted individuals (Schlaggar & McCandliss, 2007; Striem-Amit, Cohen, Dehaene, & Amedi, 2012).

Most research about blindness and PTS has been conducted with short or relatively short stimulation periods (Sathian & Stilla, 2010; Kupers & Ptito, 2014). Likewise, most of the research regarding letter perception has focused on Braille (Veispak, Boets, & Ghesquière, 2013) or on critical periods for plasticity (Burton, Sinclair, & Dixit, 2010), even though one report addressed processing of non-Braille tactile reading without extensive previous training (Burton, McLaren, & Sinclair, 2006).

ERPs are useful for investigating the timing in the brain processing under tactile stimulation (Ortiz Alonso et al., 2015) and qualitative differences of language processing (Friederici, 2004). The negative component N400 is associated with semantic processes (Hata, Homae, & Hagiwara, 2011; Deacon, Shelley-Tremblay, Ritter, & Dynowska, 2013) lexical post integration (Deacon, Hewitt, Yang, & Nagata, 2000), and other cognitive processes associated with language (Ortells, Kiefer, Castillo, Megías, & Morillas, 2016). N400 is part of the normal brain automatic (or unconscious) response to words and other semantically meaningful or potentially meaningful stimuli (Rohaut & Naccache, 2017). Unlike P300, which typically represents novelty, the later negative component is specifically engaged by attentional processes to meaning. Indeed, the N400 seems to index an early verbal semantic network and unconscious stage of processing whereas the P300 indexes processing of auditory novelty, and sustained P600 indexes conscious access to semantic knowledge (Rohaut & Naccache, 2017). Therefore, we chose to analyze N400 to address neural processing of letters in an attempt to understand the specific semantic (language) processing component of the crossmodal task. N400 ERP is associated with different levels of perceptive and phonological processing (Kutas & Federmeier, 2011); in superior, middle and inferior frontal gyrus it correlates with complex tasks such as decision making, integration of spatial information, inhibition and visual recognition (Sabsevitz, Medler, Seidenberg, & Binder, 2005), whereas regional sources in left and right lateral prefrontal cortex, right temporal cortex, and both anterior and posterior cingulate were responsive to semantic manipulation (Frishkoff, Tucker, Davey, & Scherg, 2004). Left hemisphere activity preceded right hemisphere activity, and semantic effects in frontal regions began earlier and were more sustained than the transient effects within posterior cortical regions (Frishkoff et al., 2004).

In a previous study in the same subjects we demonstrated that short term SSD stimulation, either without or with semantic content, results in faster activation of lateral occipital complex (LOC) in children with blindness when compared to seeing controls (Ortiz et al., 2015). We also found that children with blindness had a more intense activation of the left inferior temporal gyrus during the same tasks, suggesting engagement of the ventral visual processing stream (Ortiz et al., 2015). We now present data with long-term behavioral training with a SSD to provide visual information, to test if stable cross-modal reassignment of visual pathways follows proficiency in recognizing tactile semantic content (series of letters). To the best of our knowledge, our results provide the first published example of standard letter recognition (not Braille) by children with blindness using a tactile SSD.

2. Material and methods

2.1 Demographics

We enrolled 24 children, 12 with blindness (5 girls and 7 boys), (mean age = 9.9 years old, SD = 1.1) and 12 sighted controls (5 girls and 7 boys), (mean = 9.6 years old, SD = 0.7). Age ranges from 7 to 11 years old. All children had similar school performance, and cultural background. Causes of blindness are described in Table 1. Participants were chosen among students from 12 randomly selected schools serving children with blindness from Madrid, Spain. Inclusion criteria for the participants were (a) age between 7 and 11 years old, (b) active schooling and, (c) normal school performance. School performance was verified from the psychological school reports and grades. Exclusion criteria were: (a) having another sensorial deficit different than blindness, (b) present or past neuropsychiatric disease or, (c) history of obstetric trauma with cerebral hypoxia. The research protocol was approved by the Ethical Committee of the Hospital Clínico Universitario San Carlos (Madrid) and was in full compliance with the Declaration of Helsinki. Written information about the experiment was provided to all participant schools. After principals and teachers approved the research protocol, a formal presentation was organized in each school, and parents and teachers were provided detailed information (verbally and in writing) about the nature and purpose of the experiment. Interested parents provided written consent following individual informative sessions. Children were provided the same information verbally, and encouraged to ask questions in the context of group and individual sessions prior to providing their assent to participate in the study.

Table 1
Characteristics of the children with blindness

Age Age at blindness Light perception Handedness Cause of blindness

8 6 years yes right Pigmentary Retinal Dystrophy

11 birth yes right Hereditary Retinal Dystrophy

8 birth yes right Pigmentary Retinal Dystrophy

9 birth yes right Optic disc coloboma

10 birth yes right Pigmentary Retinal Dystrophy

7 birth no right Hereditary Retinal Dystrophy

10 birth yes right Partial optic atrophy

9 birth yes right Hereditary Retinal Dystrophy

9 birth yes right Pigmentary Retinal Dystrophy

10 birth no right Anophtalmia

10 1 year yes right Optic atrophy

12 3 years yes right Optic atrophy

Age	Age at blindness	Light perception	Handedness	Cause of blindness
8	6 years	yes	right	Pigmentary Retinal Dystrophy
11	birth	yes	right	Hereditary Retinal Dystrophy
8	birth	yes	right	Pigmentary Retinal Dystrophy
9	birth	yes	right	Optic disc coloboma
10	birth	yes	right	Pigmentary Retinal Dystrophy
7	birth	no	right	Hereditary Retinal Dystrophy
10	birth	yes	right	Partial optic atrophy
9	birth	yes	right	Hereditary Retinal Dystrophy
9	birth	yes	right	Pigmentary Retinal Dystrophy
10	birth	no	right	Anophtalmia
10	1 year	yes	right	Optic atrophy
12	3 years	yes	right	Optic atrophy

2.2 Tactile stimulation SSD

The tactile stimulation SSD was described previously (Ortiz et al., 2011). Briefly, a micro-camera (visual receptor) mounted on an eyeglass frame transmits images to a tactile stimulator (stimulation matrix) either wirelessly or through a cable. The stimulator includes a microprocessor that transforms images captured by the micro-camera into vibro-tactile impulses using ad hoc algorithms with features such as borders processing, zoom and image stabilization, automatic contrast adjustment and separation of planes. The stimulation matrix has 28 x 28 stimulation points, corresponding to binned pixels of the image captured by the micro-camera. The child passively touches the stimulation matrix with his/her dominant hand. Images projected on a flat screen are captured by the camera occupying the whole field of view. The sweep signal of the entire image is performed every 75 milliseconds.

2.3 Behavioral training and training sessions

Subjects underwent twice-daily (morning and afternoon) training sessions, 5 days a week with using a SSD. Each session lasted 30 minutes, and consisted of systematic, orderly and organized repetition of the visual stimuli transformed into tactile stimulation by the SSD. The sighted control group underwent exactly the same protocol but was blindfolded during each training session. The content of each session changed during training. The first week sessions consisted of displaying lines with varying spatial orientation (vertical, horizontal and diagonal). During week 2, all capital letters were added. For the remaining of the behavioral training, all sessions included both lines and letters. During the sessions children were asked to describe the orientation of the lines or to name the letters, and given feedback about the correctness of their identification.

2.4 Event related potentials task

Data were collected in a dimly lit room isolated from external noise. Children sat in an armchair, 75 cm in front of a 19 inch LCD screen (refreshed at a rate of 100 Hz) used to display the stimuli, and were provided with a keyboard to enter responses to recognized shapes. They were asked to be as relaxed as possible. The task consisted of 300 randomly ordered stimuli, each presented in the center of the screen for 300 msec and separated from the subsequent stimulus by a black screen interval of 700 msec. The total duration of the task was 5 min. To obtain ERPs an oddball paradigm with just two stimuli was used. A high frequency line or letter was presented 80 % of the time, whereas a low frequency stimulus (20 %) was designed as target and a motor response (press the space bar) was required whenever it appeared on the screen (Ortiz Alonso et al., 2015). ERPs were obtained one week after the onset of behavioral training but before letters were introduced as stimuli (pre-training). Six months after onset of behavioral training, when children were fully proficient performing the task, two sets ERPs were obtained, one with lines of variable orientation, and one with letters (Fig. 1).

Fig. 1

Event Related Potentials (ERP) wave following tactile presentation of infrequent SSD delivered stimuli. Black line represents a typical ERP following presentation of lines with variable orientation (target: vertical). Red line represents a typical ERP following presentation of variable letters (target: N). Time frame to analyze N400 components were 330–450 ms and it was determined by searching for the maximal negative amplitude in the respective time window at the Pz electrode. The BMA analysis was made opening a time window of 40 ms (–20 to +20 ms) starting from the high amplitude pick measured in Pz electrode (grey bar).

2.5 Electrophysiology recording and analysis

High-density (128 channel) EEG recordings were obtained during tactile stimulation using a custom-designed electrode Neuroscan cap and an ATI EEG system (Advantek SRL). Impedances were kept under 5kOhms. Additional channels were included to monitor eye movement (right and left lateral-canthi and superior and inferior orbits of the left eye) the reference electrodes were placed on the mastoids, the ground electrode was placed on the forehead. Data were processed to an average reference following acquisition with a band-pass filter of 0.05–30 Hz and a sample rate of 512 Hz. An artifact rejection criterion of 100 microV was used to exclude eye blinks. Individual subject averages were visually inspected to insure that clean recordings were obtained. Eye and muscle movement artifacts were identified off-line on a trial-by-trial basis through visual inspection, and they were removed prior to data averaging and ERP analysis. Noisy channels were sparingly replaced with linear interpolations from clean channels (around 6 +/– 3.5 channels per record and subject). From the remaining artifact-free trials, averages were computed for each participant and each condition using 1,000 msec epocs (including a 100 ms pre-stimulus baseline). EEG analysis was carried out on frequent (non-target) trials to avoid contamination by motor-related neural activity associated with making a response. ERPs were averaged separately for each condition and each subject. We analyzed the N400 ERP component generated between 300–400 msec time-window after the trigger (Fig. 1, red line). Reaction times (RT) slow responses exceeding 900 msec as well as incorrect responses were eliminated. The behavioral learning, intended to allow users with blindness to read using the device, was independent of the electrophysiological testing. The latter was intended to assess automatic semantic processing and to provide a physiological mapping of it before and after the behavioral training, were independent from each other.

2.6 Statistics

2.6.1 Group effects

We designed the analysis to test specific effects of training with a sensory substitution device in blind children. The primary outcome measure was accuracy in a letter recognition task. We included reaction times as a control for non-perceptual group differences in the task execution. To understand the mechanism of any differential effects, we analyzed the latency and distribution of N400 ERPs testing for differences in the cortical representation of the tactile stimulation before and after training in blind children and controls. Shapiro-Wilk test was used to test the normality of continuous variables. Main effects and interactions of SDS training and condition (blindness vs normal vision) were tested using generalized linear models. For accuracy counts, we estimated the maximum likelihood using a Poisson loglinear distribution with N400 latency and reaction times as covariants, and training and condition as predictors. Additional generalized linear model analysis were performed with N400 latency and reaction times as responses. All statistical analyses were carried out using SPSS 22 statistics software.

2.6.2 Source localization reconstruction

The sources of the N400 component were estimated from 123 electrode recordings in 12 blind children and 12 sighted controls. Localization was derived through the solution of the EEG inverse problem using the Bayesian Model Averaging (BMA) approach (Trujillo-Barreto, Aubert-Vázquez, & Valdés-Sosa, 2004) with an opening time window of – 20 to +20 msec, starting from the highest negative amplitude peak measured in Pz electrode. Individual models were solved with Low Resolution Electromagnetic Tomography (LORETA) (Pascual-Marqui, Michel, & Lehmann, 1994). Each model was defined by constraining the solution to a particular anatomical structure or combination of them using the Statistical Parametric Mapping software package (SPM8, The MathWorks, Natick, MA). Source localization analyses were based on the highest amplitude for the Pz wave.

2.6.3 Source localization group analysis

Primary current density (PCD) was estimated for each subject’s N400 component pre and post sensory substitution training. SPM8 was used to make population-level inferences over the calculated sources. Subsequently, SPMs were computed based on a voxel-by-voxel Hotelling T2 test against zero (Carbonell et al., 2004) to estimate statistically significant sources for N400 ERP components for each session (Pre and Post-training). Statistically significant effects of training and condition were computed using the voxel-by-voxel Hotelling T2 test for independent groups. The resulting probability maps were thresholded at a false discovery rate of 0.05 (Lage-Castellanos, Martínez-Montes, Hernández-Cabrera, & Galán, 2010) and depicted as 3D activation images overlaid on the Montreal Neurological Institute (MNI) average brain. Anatomical structures according to the Automated Anatomical Labeling reconciliation tool (Tzourio-Mazoyer et al., 2002) were identified, and local maxima were measured and located according to MNI coordinates system.

3. Results

3.1 Overall effects of the SSD behavioral training at week 1 and after six months

Reaction times for line orientations were unchanged from week 1 in blind (621±53) and seeing children (638±35), to month six (621±52 and 638±24, respectively) of behavioral training. Likewise, accuracy of recognition of line orientation did not change between week 1 and month 6 in either group (not shown). Table 2 describes the main effects of SSD behavioral training on letter recognition. As expected, blind and seeing children had similar reaction times on the behavioral task. On the other hand, training resulted in increased number of accurate letter identifications through the SSD in both seeing and blind children (omnibus Wald Chi Square = 11906.934, p = 0.000). Notably, the improvement in performance was significantly larger in children with blindness (diagnosis^*condition interaction Wald Chi Square = 8.272, p = 0.004).

Table 2
Main outcomes of the SSD behavioral training

	n	Assessment	Reaction time	N400 Latency	Accuracy
Seeing	12	Week 1	668.4 + /– 20.5	334.4 + /– 12.3	21.2 + /– 3.8
		After 6 months	696.3 + /– 20.1	344.4 + /– 12.1	28.1 + /– 3.8 *
Blind	12	Week 1	699.9 + /– 20.9	336.1 + /– 12.6	17.7 + /– 3.8
		After 6 months	666.8 + /– 21.7	353.6 + /– 12.9	36.7 + /– 3.8 **,#

3.2 Event related potentials for lines and letters

As shown in Fig. 1, after six months of twice daily behavioral training on the use of the SSD, ERPs for non-semantic content (line orientation) did not generate a N400 component (black line), which was, on the other hand, present for processing of infrequent letters (red line).

3.3 N400 source localization maps before and after training

Table 3 summarizes the distribution of N400 peak amplitudes at the onset (1 week) and after (six months) SSD behavioral training. At the beginning of behavioral training, N400 peak amplitude localized to the right inferior frontal gyrus in both groups, but to occipital cortex in blind children, and posterior temporal cortex in sighted controls. After behavioral training, overall speed of processing, as reflected by the N400 latency, was unchanged by training and similar in both groups. However, peak N400 amplitude shifted to left occipital and bilateral parietal cortices in blind children, but to left precentral and postcentral and bilateral occipital cortices in sighted controls (Fig. 2).

Table 3
Distribution of N400 peak amplitudes before and after SSD bevavioral training

		AAL	Brodmann Area	Voxels	T2 Max	p
Seeing	1 week	Right Frontal Inf	46	4	1120500	0.00036
		Right Temporal Mid	21	82	1160790	0.00034
		Right Temporal Sup	37	82	1156490	0.00030
	6 months	Left Calcarine	17	37	4177160	0.00002
		Left Lingual	17	22	3677110	0.00036
		Left Precentral	6	104	3988830	0.00007
		Left Postcentral	2	74	4414910	0.00027
		Right Lingual	17	30	1139490	0.00094
Blind	1 week	Right Frontal Inf	46	4	1820500	0.00018
		Right Occipital Sup	18	31	4134180	0.00045
	6 months	Left Calcarine	17	16	1220160	0.00038
		Left Precuneus	5	129	5548600	0.00026
		Right Precuneus	5	76	2463760	0.00391

Fig. 2

N400 mean electrical maps during week 1 and after six months of SSD behavioral training in blind children and matched seeing controls. SPMs were computed based on a voxel-by-voxel Hotelling T2 test against zero. Maximal intensity projection areas are displayed in yellow/red color.

4. Discussion

In developed countries, most of the causes of blindness are irreversible and few restorative treatments, if any, have been identified for this population. In the case of blind children, one particular challenge is the limited ability to optimally learn and succeed in a classroom environment that relies heavily on visual information along sighted peers. Here, we present one of the first studies done in blind children applying a SSD based on PTS as the first step for a possible prosthetic treatment. We analyzed amplitude and distribution of the N400 ERP, a known marker of linguistically demanding cognitive processes (Ortells et al., 2016). Blind children learned to recognize letters passively translated from a video-camera to tactile stimulation, in less time and with fewer errors than their seeing counterparts (Table 2).

Is it correct to consider our SSD “passive” touch? The distinction between passive and active touch has been established from the work of Loomis and Lederman on the basis of involvement of kinesthesis in the latter, but not the former sensory processes. Under a broad definition, passive tactile perception is derived exclusively from cutaneous information, whereas a narrow definition of passive tactile perception includes afferent kinesthesis. In both definitions, active haptic perception requires efferent kinesthesis and active control. In other words, unless manipulation of objects is included in the perception, touch perception remains passive. Therefore, since in our experiment head motion is only likely to affect centering of the image in the camera, it is unlikely to mimic the manipulation required for active touch perception to occur.

In blind individuals, but not in seeing controls, tactile acuity (the ability to discriminate dots) using active touch has been shown to be preserved throughout lifetime (Legge, Madison, Vaughn, Cheong, & Miller, 2008). Somewhat unexpectedly, tactile acuity does not correlate with Braille reading speed, the amount of daily reading, or the age at which braille was learned (Legge et al., 2008); these findings are consistent with studies reporting lack of practice-induced improvement in tactile acuity (Reuter, Voelcker-Rehage, Vieluf, & Godde, 2012). Nevertheless, repetitive tactile stimulation is a well-stablished inducer of sensory cortical plasticity (Reuter, Voelcker-Rehage, Vieluf, Winneke, & Godde, 2014). Indeed, differences in brain electrical activity between blind and sighted children suggest a reorganization of brain areas in the blind subjects that may contribute to the improved spatial resolution for tactile sources (Röder et al., 1999).

In the baseline condition of our study, blind and seeing children activated frontal cortical structures during the task, possibly as a result of attentional processes (Frey, Zlatkina, & Petrides, 2009; Davenport & Coulson, 2013; Qu et al., 2017), as well as the recognition of novel or unpredictable stimuli since subjects had no prior experience with the SSD. Prior to behavioral training, peak activations during the ERP task were influenced by visual capacity. In blind children maximum amplitude was found in right superior occipital cortex (BA18), whereas in seeing children maximum aplitude was found in right middle temporal (BA21) and right occipitotemporal cortex (BA 37) (Table 3). BA18 represents the bulk of the visual association area, whose neurons are tuned to simple visual characteristics (such as orientation, color or shape), as well as to various complex shape characteristics (such as the orientation of illusory). This structure plays a very important role in object recognition memory as well as in conversion of short-term object memories into long-term memories. Therefore, its peak activation in blind children during the letter recognition ERP task is strongly suggestive of reassignment of its function to tactile-driven semantic information.

The dominant BA21 is associated with selective processing of text and speech semantic processing as well as word and sentence generation, but its role on the non-dominant hemisphere is less defined. It may participate in prosodic integration, processing of complex sounds and, together with BA37, attribution of intention to others (mirror neurons) (Brunet, Sarfati, Hardy-Baylé, & Decety, 2000). Left BA37 includes the visual word form area, but in the right hemisphere this area is engaged by complex visual-spacial tasks, face recognition, and theory of mind as discussed (Weiner & Zilles, 2016; Brunyé, Moran, Holmes, Mahoney, & Taylor, 2017). It seems likely, therefore, that both areas (21 and 37) are engaged by the SSD ERP task in seeing children because of their experience with multimodal language processing (Ricciardi, Bonino, Pellegrini, & Pietrini, 2014; Murray, Lewkowicz, Amedi, & Wallace, 2016). Conversely, blind children engage visual association cortex as a consequence of cross-modal compensatory plasticity (Fiehler & Rösler, 2010; Murray et al., 2016).

Our results suggest that blind children learn to recognize SSD-delivered letters more effectively than their seeing peers. Notably, after behavioral training sighted children had greater N400 amplitude during the SSD ERP task in language related temporal areas whereas activation was greatest in the occipital region of blind children. Long term behavioral training altered the localization of the peak N400 amplitude. In both seeing and blind children, behavioral training resulted in recruitment of the left occipital cortex (Table 3). Somatosensory circuits also became engaged in both groups, with some differences. In blind children, bilateral somatosensory association cortex (BA5) became activated after long term behavioral training, whereas in seeing children left primary somatosensory (BA2) and bilateral primary visual cortex (BA17) were newly engaged. These results are expected because novel stimuli are typically processed by the right hemisphere (Garrett et al., 2000; Flowers et al., 2004), whereas the left is associated with episodic learning and language (Ortiz Alonso et al., 2015). Our data also indicate that left occipital cortex is engaged by language processing in blind children (Burton, Diamond, & McDermott, 2003; Bedny, 2017). In blind adults, left occipital cortex may be involved in long-term verbal memory consolidation (Amedi, Raz, Pianka, Malach, & Zohary, 2003; Raz, Amedi, & Zohary, 2005), possibly explaining its predominant activation after the SSD ERP task.

Thus, plasticity was influenced by early experience with vision. The switch from predominant right hemisphere involvement to engagement of left hemispheric cortical structures by the end of behavioral training suggests that subjects learn to process SSD information as language. In late bilinguals, asymmetry for words increases with years of experience (Grossi, Savill, Thomas, & Thierry, 2012), suggesting that engagement of right hemispheric structures may be due primarily to the novelty of the SSD use. In any case, hemispheric switching after SSD behavioral training suggests that subjects effectively transform initial spatial somatosensory stimuli to a representation in the linguistic space of the brain.

Footnotes

Acknowledgments

The work was supported by a generous gift of Fundación Eugenio Rodriquez Pascual to TO, and Valley Baptist Legacy Foundation, Grant/Award number 450000942 to GAdE.

References

Amedi, A., Raz, N., Azulay, H., Malach, R., & Zohary, E. (2010). Cortical activity during tactile exploration of objects in blind and sighted humans. Restorative Neurology and Neuroscience, 28(2), 143–156. https://doi.org/10.3233/RNN-2010-0503

Amedi, A., Raz, N., Pianka, P., Malach, R., & Zohary, E. (2003). Early ``visual'' cortex activation correlates with superior verbal memory performance in the blind. Nature Neuroscience, 6(7), 758–766. https://doi.org/10.1038/nn1072

Bedny, M. (2017). Evidence from Blindness for a Cognitively Pluripotent Cortex. Trends in Cognitive Sciences, 21(9), 637–648. https://doi.org/10.1016/j.tics.2017.06.003

Bhattacharjee, A., Ye, A.J., Lisak, J.A., Vargas, M.G., & Goldreich, D. (2010). Vibrotactile masking experiments reveal accelerated somatosensory processing in congenitally blind braille readers. The Journal of Neuroscience: The Official Journal of the Society for Neuroscience, 30(43), 14288–14298. https://doi.org/10.1523/JNEUROSCI.1447-10.2010

Brunet, E., Sarfati, Y., Hardy-Baylé, M.C., & Decety, J. (2000). A PET investigation of the attribution of intentions with a nonverbal task. NeuroImage, 11(2), 157–166. https://doi.org/10.1006/nimg.1999.0525

Brunyé, T.T., Moran, J.M., Holmes, A., Mahoney, C.R., & Taylor, H.A. (2017). Non-invasive brain stimulation targeting the right fusiform gyrus selectively increases working memory for faces. Brain and Cognition, 113, 32–39. https://doi.org/10.1016/j.bandc.2017.01.006

Burton, H., Diamond, J.B., & McDermott, K.B. (2003). Dissociating cortical regions activated by semantic and phonological tasks: a FMRI study in blind and sighted people. Journal of Neurophysiology, 90(3), 1965–1982. https://doi.org/10.1152/jn.00279.2003

Burton, H., McLaren, D.G., & Sinclair, R.J. (2006). Reading embossed capital letters: an fMRI study in blind and sighted individuals. Human Brain Mapping, 27(4), 325–339. https://doi.org/10.1002/hbm.20188

Burton, H., Sinclair, R.J., & Dixit, S. (2010). Working memory for vibrotactile frequencies: comparison of cortical activity in blind and sighted individuals. Human Brain Mapping, 31(11), 1686–1701. https://doi.org/10.1002/hbm.20966

10.

Carbonell, F., Galán, L., Valdés, P., Worsley, K., Biscay, R.J., Díaz-Comas, L., … & Parra, M. (2004). Random field-union intersection tests for EEG/MEG imaging. NeuroImage, 22(1), 268–276. https://doi.org/10.1016/j.neuroimage.2004.01.020

11.

Cohen, L., Dehaene, S., Vinckier, F., Jobert, A., & Montavont, A. (2008). Reading normal and degraded words: Contribution of the dorsal and ventral visual pathways. NeuroImage, 40(1), 353–366. https://doi.org/10.1016/j.neuroimage.2007.11.036

12.

Davenport, T., & Coulson, S. (2013). Hemispheric asymmetry in interpreting novel literal language: An event-related potential study. Neuropsychologia, 51(5), 907–921. https://doi.org/10.1016/j.neuropsychologia.2013.01.018

13.

Deacon, D., Hewitt, S., Yang, C., & Nagata, M. (2000). Event-related potential indices of semantic priming using masked and unmasked words: evidence that the N400 does not reflect a post-lexical process. Brain Research. Cognitive Brain Research, 9(2), 137–146.

14.

Deacon, D., Shelley-Tremblay, J.F., Ritter, W., & Dynowska, A. (2013). Electrophysiological evidence for the action of a center-surround mechanism on semantic processing in the left hemisphere. Frontiers in Psychology, 4, 936. https://doi.org/10.3389/fpsyg.2013.00936

15.

Feldman, D.E., & Brecht, M. (2005). Map plasticity in somatosensory cortex. Science (New York, N.Y.), 310(5749), 810–815. https://doi.org/10.1126/science.1115807

16.

Fiehler, K., & Rösler, F. (2010). Plasticity of multisensory dorsal stream functions: evidence from congenitally blind and sighted adults. Restorative Neurology and Neuroscience, 28(2), 193–205. https://doi.org/10.3233/RNN-2010-0500

17.

Flowers, D.L., Jones, K., Noble, K., VanMeter, J., Zeffiro, T.A., Wood, F.B., & Eden, G.F. (2004). Attention to single letters activates left extrastriate cortex. NeuroImage, 21(3), 829–839. https://doi.org/10.1016/j.neuroimage.2003.10.002

18.

Frey, S., Zlatkina, V., & Petrides, M. (2009). Encoding touch and the orbitofrontal cortex. Human Brain Mapping, 30(2), 650–659. https://doi.org/10.1002/hbm.20532

19.

Friederici, A.D. (2004). Event-related brain potential studies in language. Current Neurology and Neuroscience Reports, 4(6), 466–470.

20.

Frings, C., Amendt, A., & Spence, C. (2011). When seeing doesn't matter: assessing the after-effects of tactile distractor processing in the blind and the sighted. Journal of Experimental Psychology. Human Perception and Performance, 37(4), 1174–1181. https://doi.org/10.1037/a0022336

21.

Frishkoff, G.A., Tucker, D.M., Davey, C., & Scherg, M. (2004). Frontal and posterior sources of event-related potentials in semantic comprehension. Brain Research. Cognitive Brain Research, 20(3), 329–354. https://doi.org/10.1016/j.cogbrainres.2004.02.009

22.

Garrett, A.S., Flowers, D.L., Absher, J.R., Fahey, F.H., Gage, H.D., Keyes, J.W., & Wood, F.B. (2000). Cortical activity related to accuracy of letter recognition. NeuroImage, 11(2), 111–123. https://doi.org/10.1006/nimg.1999.0528

23.

Goldreich, D., & Kanics, I.M. (2003). Tactile acuity is enhanced in blindness. The Journal of Neuroscience: The Official Journal of the Society for Neuroscience, 23(8), 3439–3445.

24.

Grossi, G., Savill, N., Thomas, E., & Thierry, G. (2012). Electrophysiological cross-language neighborhood density effects in late and early english-welsh bilinguals. Frontiers in Psychology, 3, 408. https://doi.org/10.3389/fpsyg.2012.00408

25.

Gurtubay-Antolin, A., & Rodríguez-Fornells, A. (2017). Neurophysiological evidence for enhanced tactile acuity in early blindness in some but not all haptic tasks. NeuroImage, 162, 23–31. https://doi.org/10.1016/j.neuroimage.2017.08.054

26.

Hata, M., Homae, F., & Hagiwara, H. (2011). Semantic relatedness between words in each individual brain: an event-related potential study. Neuroscience Letters, 501(2), 72–77. https://doi.org/10.1016/j.neulet.2011.06.044

27.

Kim, J.S., Kanjlia, S., Merabet, L.B., & Bedny, M. (2017). Development of the visual word form area requires visual experience: Evidence from blind Braille readers. The Journal of Neuroscience, 0997–17. https://doi.org/10.1523/JNEUROSCI.0997-17.201710.1523/JNEUROSCI.0997-17.2017

28.

Kupers, R., & Ptito, M. (2014). Compensatory plasticity and cross-modal reorganization following early visual deprivation. Neuroscience and Biobehavioral Reviews, 41, 36–52. https://doi.org/10.1016/j.neubiorev.2013.08.001

29.

Kutas, M., & Federmeier, K.D. (2011). Thirty years and counting: finding meaning in the N400 component of the event-related brain potential (ERP). Annual Review of Psychology, 62, 621–647. https://doi.org/10.1146/annurev.psych.093008.131123psych.093008.131123

30.

Lage-Castellanos, A., Martínez-Montes, E., Hernández-Cabrera, J.A., & Galán, L. (2010). False discovery rate and permutation test: an evaluation in ERP data analysis. Statistics in Medicine, 29(1), 63–74. https://doi.org/10.1002/sim.3784

31.

Legge, G.E., Madison, C., Vaughn, B.N., Cheong, A.M.Y., & Miller, J.C. (2008). Retention of high tactile acuity throughout the life span in blindness. Perception & Psychophysics, 70(8), 1471–1488. https://doi.org/10.3758/PP.70.8.1471

32.

Merabet, L.B., Hamilton, R., Schlaug, G., Swisher, J.D., Kiriakopoulos, E.T., Pitskel, N.B., … & Pascual-Leone, A. (2008). Rapid and reversible recruitment of early visual cortex for touch. PloS One, 3(8), e3046. https://doi.org/10.1371/journal.pone.0003046

33.

Merabet, L.B., & Pascual-Leone, A. (2010). Neural reorganization following sensory loss: the opportunity of change. Nature Reviews. Neuroscience, 11(1), 44–52. https://doi.org/10.1038/nrn2758

34.

Murray, M.M., Lewkowicz, D.J., Amedi, A., & Wallace, M.T. (2016). Multisensory Processes: A Balancing Act across the Lifespan. Trends in Neurosciences, 39(8), 567–579. https://doi.org/10.1016/j.tins.2016.05.003

35.

Norman, J.F., & Bartholomew, A.N. (2011). Blindness enhances tactile acuity and haptic 3-D shape discrimination. Attention, Perception, & Psychophysics, 73(7), 2323–2331. https://doi.org/10.3758/s13414-011-0160-4

36.

Ortells, J.J., Kiefer, M., Castillo, A., Megías, M., & Morillas, A. (2016). The semantic origin of unconscious priming: Behavioral and event-related potential evidence during category congruency priming from strongly and weakly related masked words. Cognition, 146, 143–157. https://doi.org/10.1016/j.cognition.2015.09.012

37.

Ortiz Alonso, T., Santos, J.M., Ortiz Terán, L., Borrego Hernández, M., Poch Broto, J., & de Erausquin, G.A. (2015). Differences in Early Stages of Tactile ERP Temporal Sequence (P100) in Cortical Organization during Passive Tactile Stimulation in Children with Blindness and Controls. PloS One, 10(7), e0124527. https://doi.org/10.1371/journal.pone.0124527

38.

Ortiz, T., Poch, J., Santos, J.M., Requena, C., Martínez, A.M., Ortiz-Terán, L., … & Pascual-Leone, A. (2011). Recruitment of Occipital Cortex during Sensory Substitution Training Linked to Subjective Experience of Seeing in People with Blindness. PLoS ONE, 6(8), e23264. https://doi.org/10.1371/journal.pone.0023264

39.

Pascual-Marqui, R.D., Michel, C.M., & Lehmann, D. (1994). Low resolution electromagnetic tomography: a new method for localizing electrical activity in the brain. International Journal of Psychophysiology: Official Journal of the International Organization of Psychophysiology, 18(1), 49–65.

40.

Qu, J., Qian, L., Chen, C., Xue, G., Li, H., Xie, P., & Mei, L. (2017). Neural Pattern Similarity in the Left IFG and Fusiform Is Associated with Novel Word Learning. Frontiers in Human Neuroscience, 11, 424. https://doi.org/10.3389/fnhum.2017.0042410.3389/fnhum.2017.00424.

41.

Rauschecker, J.P. (2011). An expanded role for the dorsal auditory pathway in sensorimotor control and integration. Hearing Research, 271(1-2), 16–25. https://doi.org/10.1016/j.heares.2010.09.001j.heares.2010.09.001.

42.

Raz, N., Amedi, A., & Zohary, E. (2005). V1 activation in congenitally blind humans is associated with episodic retrieval. Cerebral Cortex (New York, N.Y.: 1991), 15(9), 1459–1468. https://doi.org/10.1093/cercor/bhi026

43.

Reuter, E.-M., Voelcker-Rehage, C., Vieluf, S., & Godde, B. (2012). Touch perception throughout working life: effects of age and expertise. Experimental Brain Research, 216(2), 287–297. https://doi.org/10.1007/s00221-011-2931-5

44.

Reuter, E.-M., Voelcker-Rehage, C., Vieluf, S., Winneke, A.H., & Godde, B. (2014). Extensive occupational finger use delays age effects in tactile perception-an ERP study. Attention, Perception & Psychophysics, 76(4), 1160–1175. https://doi.org/10.3758/s13414-014-0634-2

45.

Ricciardi, E., Bonino, D., Pellegrini, S., & Pietrini, P. (2014). Mind the blind brain to understand the sighted one! Is there a supramodal cortical functional architecture?? Neuroscience and Biobehavioral Reviews, 41, 64–77. https://doi.org/10.1016/j.neubiorev.2013.10.006

46.

Röder, B., Teder-Sälejärvi, W., Sterr, A., Rösler, F., Hillyard, S.A., & Neville, H.J. (1999). Improved auditory spatial tuning in blind humans. Nature, 400(6740), 162–166. https://doi.org/10.1038/22106

47.

Rohaut

, & Naccache

Disentangling conscious from unconscious cognitive processing with event-related EEG potentials.

Rev Neurol (Paris). 2017 Jul-Aug; 173(7-8):521-528. doi:10.1016/j.neurol.2017.08.001. Epub 2017 Aug 24.

48.

Sabsevitz, D.S., Medler, D.A., Seidenberg, M., & Binder, J.R. (2005). Modulation of the semantic system by word imageability. NeuroImage, 27(1), 188–200. https://doi.org/10.1016/j.neuroimage.2005.04.01210.1016/j.neuroimage.2005.04.012.

49.

Sathian, K., & Stilla, R. (2010). Cross-modal plasticity of tactile perception in blindness. Restorative Neurology and Neuroscience, 28(2), 271–281. https://doi.org/10.3233/RNN-2010-0534

50.

Schlaggar, B.L., & McCandliss, B.D. (2007). Development of neural systems for reading. Annual Review of Neuroscience, 30, 475–503. https://doi.org/10.1146/annurev.neuro.28.061604.135645neuro.28.061604.135645

51.

Striem-Amit, E., Cohen, L., Dehaene, S., & Amedi, A. (2012). Reading with sounds: sensory substitution selectively activates the visual word form area in the blind. Neuron, 76(3), 640–652. https://doi.org/10.1016/j.neuron.2012.08.026

52.

Stronks, H.C., Nau, A.C., Ibbotson, M.R., & Barnes, N. (2015). The role of visual deprivation and experience on the performance of sensory substitution devices. Brain Research, 1624, 140–152. https://doi.org/10.1016/j.brainres.2015.06.033

53.

Trujillo-Barreto, N.J., Aubert-Vázquez, E., & Valdés-Sosa, P.A. (2004). Bayesian model averaging in EEG/MEG imaging. NeuroImage, 21(4), 1300–1319. https://doi.org/10.1016/j.neuroimage.2003.11.00810.1016/j.neuroimage.2003.11.008.

54.

Tzourio-Mazoyer, N., Landeau, B., Papathanassiou, D., Crivello, F., Etard, O., Delcroix, N., & Joliot, M. (2002). Automated anatomical labeling of activations in SPM using a macroscopic anatomical parcellation of the MNI MRI single-subject brain. NeuroImage, 15(1), 273–289. https://doi.org/10.1006/nimg.2001.0978

55.

Veispak, A., Boets, B., & Ghesquière, P. (2013). Differential cognitive and perceptual correlates of print reading versus braille reading. Research in Developmental Disabilities, 34(1), 372–385. https://doi.org/10.1016/j.ridd.2012.08.012

56.

Voss, P. (2013). Sensitive and critical periods in visual sensory deprivation. Frontiers in Psychology, 4, 664. https://doi.org/10.3389/fpsyg.2013.0066410.3389/fpsyg.2013.00664

57.

Weiner, K.S., & Zilles, K. (2016). The anatomical and functional specialization of the fusiform gyrus. Neuropsychologia, 83, 48–62. https://doi.org/10.1016/j.neuropsychologia.2015.06.033chologia.2015.06.033

58.

Wong, M., Gnanakumaran, V., & Goldreich, D. (2011). Tactile spatial acuity enhancement in blindness: evidence for experience-dependent mechanisms. The Journal of Neuroscience: The Official Journal of the Society for Neuroscience, 31(19), 7028–7037. https://doi.org/10.1523/JNEUROSCI.6461-10.2011JNEUROSCI.6461-10.2011