**1. Introduction: the development of space representation**

The development of a multisensory space representation of the environment is crucial for humans to interact with objects and each other. Different sensory modalities represent space using varying reference systems: vision relies on retinotopic coordinates, audition on head-centered coordinates, and touch on bodycentered ones. To perceive a multisensory world, human brains must combine the spatial information arriving from all the sensory modalities into a coherent representation. The visual modality seems to have a crucial role in this important step, and specifically in the process of developing an integrated multisensory representation of the environment. If vision is so important, then an obvious question arises: what happens to space representation when the visual input is missing? Studies of animals suggest that the lack of vision in the first period of life alters the development of space representation. For example, auditory spatial maps of juvenile barn owls change after visual adaptation with prismatic spectacles [1]. Likewise, total

visual deprivation in young ferrets is associated with the development of disordered auditory spatial maps [2]. Similar transitory effects occur in humans. In a number of studies, auditory space representation altered after short periods of adaptation to non-aligned auditory and visual stimuli [3, 4]. In agreement with this idea, research shows the representation of the auditory space is dominated by visual experience among young children [5]. Taken together, these results support not only the idea that vision is important for developing auditory space representation, but also that its absence may interfere with such development.

#### **2. Space representation and blindness**

Since blindness represents a unique condition to investigate the role of the visual modality in the development of space representation, many researchers have investigated this topic. However, contradictory results have been found. Lack of visual experience is associated with an enhancement of auditory (e.g., [6–11]) and tactile modalities [12] in blind compared with sighted individuals according to some studies. Results show that early blind subjects have enhanced skills in auditory pitch discrimination [13], localization of peripheral sounds in the horizontal plane [7, 10, 11], and ability to form spatial topographical maps underlying simple auditory localization [14, 15]. In particular, Lessard et al. [10] investigated the three-dimensional spatial mapping in early blind individuals by considering monaural and binaural listening conditions. Authors observed that early blind subjects show equal or better accuracy compared to sighted subjects when localizing single sounds. Moreover, they observed that early blind people could correctly localize sounds monaurally compared to sighted participants. Neurophysiological results indicate a clear response of the occipital cortex of blind individuals to auditory stimuli (e.g., [8, 16–19]), revealing even topographic organization [20–24]. The absence of visual input also drives anatomical changes in the auditory cortex (e.g., [25, 26]). On the other side, other studies in humans and animals show that lack of vision is associated with spatial deficits. For example, studies show that blindness affects one's ability to estimate the absolute distance of auditory cues [27–29], audio metric tasks [30, 31], auditory distance discrimination, and proprioceptive reproduction [32]. Research has also demonstrated poorer skills of blind compared to sighted people for localization of sounds along the mid-sagittal plane [11].

These results suggest that the mechanisms that subtend the development of space representation remain unclear. They also support that the role of the visual modality in space representation varies based on spatial properties, producing in some cases enhanced or impaired skills in blind individuals. The mechanisms behind this require clarification.

## **3. Audio metric impairment in blind individuals**

Another exception of the enhanced skills of blind individuals in space representation is the ability to perform an audio spatial bisection task [30]. Contrary to previous works studying pitch and timbre discrimination [13, 33], or localization of single sounds in space [7, 10], the bisection task requires estimation and comparison of different locations in space. While sighted children of 6 years of age can perform it [5], our study found that blind individuals were strongly impaired in this task. The results were in agreement with previous findings of our group showing

**287**

**Figure 1.**

difficulty.

*Audio Cortical Processing in Blind Individuals DOI: http://dx.doi.org/10.5772/intechopen.88826*

left) or to the last (i.e., right) sound.

that, during development, the visual modality dominates the multisensory spatial percept in audio-visual conditions of the bisection task, suggesting that the visual input might be crucial for the development of audio spatial bisection skills [5]. During the task, participants sat 180 cm from the center of a bank of 23 speakers and perceived three sounds: the first speaker (on the left) and the last speaker (on the right) delivered the first and the third sounds, respectively. The second sound came from an intermediate speaker between the first and the last one (see **Figure 1**). Participants verbally reported whether the second sound was closer to the first (i.e.,

While sighted individuals succeeded at the task, with responses varying systematically as a function of speaker position (a standard deviation of 4.3°), blind individuals provided almost random responses. However, the same deficit was absent for other tasks, such as the minimal audible angle task in which participants were asked to evaluate which sound was from the left in a sequence of two sounds (this result is in agreement with previous studies [10]). The deficit we reported for the bisection task was far larger than the perceptual enhancements that have been reported before, and it was highly consistent among blind individuals. **Figure 2** reports the individual thresholds for the minimum audible angle against individual thresholds for the bisection task: the thresholds of blind individuals are over the equality line. While the study observed no difference between groups for the minimum audible angle (*t* test, *p* = 0.21), groups significantly differed for the bisection thresholds (Wilcoxon signed ranks test, *p* < 0.01; bootstrap sign-test: *p* < 10–5). We performed other tests we do not report here so as to show the specificity of the deficit and its independency from the kind of sound used (e.g., the pointing task; for more details, see [30]). We also performed a temporal version of the bisection task, in which the participants performed the same task in the temporal domain. Participants had to report if the second sound was closer to the first or to the last in time. In this task, no deficit emerged, suggesting that the deficit was not due to a general/aspecific impairment, to task incomprehension, or to attention and memory problems associated with task

*Description of the spatial bisection task. Participants were aligned with the central speaker (i.e., 0°) and listened to a sequence of three sounds. The first and the third sound were delivered from the first speaker on the left (i.e., −25°) and the last speaker on the right (i.e., +25°) respectively, whereas the second sound derived from an intermediate speaker between the first and the last one (i.e., between −25° and +25°). Participants were asked whether the second sound was closer to the first (i.e., left, −25°) or the last (i.e., right, +25°) sound. Upper panel reports an exemplar trial in which the second sound is closer to the last (i.e., right) sound.*

#### *Audio Cortical Processing in Blind Individuals DOI: http://dx.doi.org/10.5772/intechopen.88826*

*Visual Impairment and Blindness - What We Know and What We Have to Know*

its absence may interfere with such development.

**2. Space representation and blindness**

along the mid-sagittal plane [11].

behind this require clarification.

**3. Audio metric impairment in blind individuals**

visual deprivation in young ferrets is associated with the development of disordered auditory spatial maps [2]. Similar transitory effects occur in humans. In a number of studies, auditory space representation altered after short periods of adaptation to non-aligned auditory and visual stimuli [3, 4]. In agreement with this idea, research shows the representation of the auditory space is dominated by visual experience among young children [5]. Taken together, these results support not only the idea that vision is important for developing auditory space representation, but also that

Since blindness represents a unique condition to investigate the role of the visual modality in the development of space representation, many researchers have investigated this topic. However, contradictory results have been found. Lack of visual experience is associated with an enhancement of auditory (e.g., [6–11]) and tactile modalities [12] in blind compared with sighted individuals according to some studies. Results show that early blind subjects have enhanced skills in auditory pitch discrimination [13], localization of peripheral sounds in the horizontal plane [7, 10, 11], and ability to form spatial topographical maps underlying simple auditory localization [14, 15]. In particular, Lessard et al. [10] investigated the three-dimensional spatial mapping in early blind individuals by considering monaural and binaural listening conditions. Authors observed that early blind subjects show equal or better accuracy compared to sighted subjects when localizing single sounds. Moreover, they observed that early blind people could correctly localize sounds monaurally compared to sighted participants. Neurophysiological results indicate a clear response of the occipital cortex of blind individuals to auditory stimuli (e.g., [8, 16–19]), revealing even topographic organization [20–24]. The absence of visual input also drives anatomical changes in the auditory cortex (e.g., [25, 26]). On the other side, other studies in humans and animals show that lack of vision is associated with spatial deficits. For example, studies show that blindness affects one's ability to estimate the absolute distance of auditory cues [27–29], audio metric tasks [30, 31], auditory distance discrimination, and proprioceptive reproduction [32]. Research has also demonstrated poorer skills of blind compared to sighted people for localization of sounds

These results suggest that the mechanisms that subtend the development of space representation remain unclear. They also support that the role of the visual modality in space representation varies based on spatial properties, producing in some cases enhanced or impaired skills in blind individuals. The mechanisms

Another exception of the enhanced skills of blind individuals in space representation is the ability to perform an audio spatial bisection task [30]. Contrary to previous works studying pitch and timbre discrimination [13, 33], or localization of single sounds in space [7, 10], the bisection task requires estimation and comparison of different locations in space. While sighted children of 6 years of age can perform it [5], our study found that blind individuals were strongly impaired in this task. The results were in agreement with previous findings of our group showing

**286**

that, during development, the visual modality dominates the multisensory spatial percept in audio-visual conditions of the bisection task, suggesting that the visual input might be crucial for the development of audio spatial bisection skills [5]. During the task, participants sat 180 cm from the center of a bank of 23 speakers and perceived three sounds: the first speaker (on the left) and the last speaker (on the right) delivered the first and the third sounds, respectively. The second sound came from an intermediate speaker between the first and the last one (see **Figure 1**). Participants verbally reported whether the second sound was closer to the first (i.e., left) or to the last (i.e., right) sound.

While sighted individuals succeeded at the task, with responses varying systematically as a function of speaker position (a standard deviation of 4.3°), blind individuals provided almost random responses. However, the same deficit was absent for other tasks, such as the minimal audible angle task in which participants were asked to evaluate which sound was from the left in a sequence of two sounds (this result is in agreement with previous studies [10]). The deficit we reported for the bisection task was far larger than the perceptual enhancements that have been reported before, and it was highly consistent among blind individuals. **Figure 2** reports the individual thresholds for the minimum audible angle against individual thresholds for the bisection task: the thresholds of blind individuals are over the equality line. While the study observed no difference between groups for the minimum audible angle (*t* test, *p* = 0.21), groups significantly differed for the bisection thresholds (Wilcoxon signed ranks test, *p* < 0.01; bootstrap sign-test: *p* < 10–5). We performed other tests we do not report here so as to show the specificity of the deficit and its independency from the kind of sound used (e.g., the pointing task; for more details, see [30]). We also performed a temporal version of the bisection task, in which the participants performed the same task in the temporal domain. Participants had to report if the second sound was closer to the first or to the last in time. In this task, no deficit emerged, suggesting that the deficit was not due to a general/aspecific impairment, to task incomprehension, or to attention and memory problems associated with task difficulty.

#### **Figure 1.**

*Description of the spatial bisection task. Participants were aligned with the central speaker (i.e., 0°) and listened to a sequence of three sounds. The first and the third sound were delivered from the first speaker on the left (i.e., −25°) and the last speaker on the right (i.e., +25°) respectively, whereas the second sound derived from an intermediate speaker between the first and the last one (i.e., between −25° and +25°). Participants were asked whether the second sound was closer to the first (i.e., left, −25°) or the last (i.e., right, +25°) sound. Upper panel reports an exemplar trial in which the second sound is closer to the last (i.e., right) sound.*

**Figure 2.**

*Individual data, plotting bisection thresholds against minimal audible angle. Arrows at the margin show the geometric means of each group as well as the shaded areas of 95% confidence intervals. The blue and green arrows show the average thresholds for 7- and 10-year-old children, respectively (taken from a previous study) [5]. The dashed diagonal line is the equality line: while the thresholds of sighted subjects are scattered around this line, all except one non-sighted subject are above it. Indeed, the only non-sighted subject with bisection threshold that falls within the control range had a threshold for minimal audible angle that was six times lower than the mean of the controls, meaning the subject's data point sits well above the bisection line. With permission from Gori et al. [30].*

#### **4. Cortical processing of space and blindness**

Scientific evidence suggests that the auditory and somatosensory systems colonize the visual cortex of congenitally blind individuals to a certain extent (e.g., [16, 34]). For example, studies that were performed with fMRI [35–38] and event-related potentials (ERPs [39, 40]) show that the visual cortex shows a strong and reliable response to sound presented alone. Tomasello et al. [41] have recently proposed a neurocomputational model to explain the visual cortex recruitment during language processing in congenitally blind individuals. For what concerns space representation, Collignon and colleagues [42] compared the brain activity of early blind and sighted individuals during a spatial and pitch task using the same stimuli for both. Authors observed that the processing of sounds recruited the occipital cortex, and the spatial processing of audio spatial stimuli also activated the dorsal occipital stream involved in visuospatial/motion processing in sighted individuals. They concluded that some regions of the right dorsal occipital stream specialize toward processing spatial information without the necessity of visual experience. Not only are visual areas activated during auditory tasks, but also localization abilities of blind subjects are strongly associated with the magnitude of visual cortex activity [8, 43, 44]. For example, early blind people localize sounds more accurately than those who are sighted under monaural conditions [10]. Their activation in right-hemisphere striate and ventral extrastriate areas correlates with the performance in a pointing task to monaurally presented sounds [8]. These results suggest that the enhancement of some auditory skills of blind individuals

**289**

**Figure 3.**

*Audio Cortical Processing in Blind Individuals DOI: http://dx.doi.org/10.5772/intechopen.88826*

and not contralateral to the sound spatial position.

may reflect in the recruitment of the visual cortex. From this arises the question: what about the impaired skills, such as in the case of the bisection task? If the visual information is important for the development of audio space bisection [45], as we reported in the previous section, then we may expect the visual cortex of sighted and not of blind individuals [30] should be recruited for this audio processing. We recently used EEG to measure activation of the occipital cortex of sighted and blind individuals during the audio bisection task [45, 46]. **Figure 3** illustrates the scalp maps elicited by the second sound of the spatial bisection task when it was delivered from the left (i.e., −4.5°, see left panel) and the right side (i.e., +4.5°, see right panel) in the 50- to 90-ms time window after sound onset, for sighted (**Figure 3A**) and blind participants (**Figure 3B**). In the case of both groups, two strong positivities emerged: one involving central areas and one involving parieto-occipital areas. However, the latter positivity showed a specific contralateral pattern during the spatial bisection task that was only in sighted subjects (**Figure 3A**). In early blind participants (**Figure 3B**), the parieto-occipital response was strongly attenuated

To provide evidence that the early contralateral component that we observed over the occipital scalp actually involved generators in occipital areas, we performed comparisons between groups at the source level (**Figure 4**). Results suggest that sighted subjects showed a stronger occipital and temporal activation contralateral to the physical sound position, while early blind subjects exhibited reduced activation in contralateral cortical areas and an increased activation in ipsilateral cortical areas. In early blind individuals, the laterality was absent, which means that early visual experience mediates development of this contralateral early occipital response. The data suggest that visual modality plays a key role in the development of an early occipital response that is specific for space perception and auditory stimuli. In sighted subjects, the acoustic recruitment of the visual brain may be necessary to build a spatial metric of the environment using high resolution and flexibility that only the visual brain is capable of implementing. Lack of vision

*Scalp maps of the mean ERP amplitude in the selected time window (50–90 ms) after the second sound of the spatial bisection task, for sighted (A) and blind (B) groups. Left and right panels of the figure report the conditions in which S2 was presented from either −4.5° (i.e., narrow first distance) or + 4.5° (i.e., wide first distance), respectively, and independently of timing (±250 ms). With permission from Campus et al. [46].*

#### *Audio Cortical Processing in Blind Individuals DOI: http://dx.doi.org/10.5772/intechopen.88826*

*Visual Impairment and Blindness - What We Know and What We Have to Know*

**4. Cortical processing of space and blindness**

Scientific evidence suggests that the auditory and somatosensory systems colonize the visual cortex of congenitally blind individuals to a certain extent (e.g., [16, 34]). For example, studies that were performed with fMRI [35–38] and event-related potentials (ERPs [39, 40]) show that the visual cortex shows a strong and reliable response to sound presented alone. Tomasello et al. [41] have recently proposed a neurocomputational model to explain the visual cortex recruitment during language processing in congenitally blind individuals. For what concerns space representation, Collignon and colleagues [42] compared the brain activity of early blind and sighted individuals during a spatial and pitch task using the same stimuli for both. Authors observed that the processing of sounds recruited the occipital cortex, and the spatial processing of audio spatial stimuli also activated the dorsal occipital stream involved in visuospatial/motion processing in sighted individuals. They concluded that some regions of the right dorsal occipital stream specialize toward processing spatial information without the necessity of visual experience. Not only are visual areas activated during auditory tasks, but also localization abilities of blind subjects are strongly associated with the magnitude of visual cortex activity [8, 43, 44]. For example, early blind people localize sounds more accurately than those who are sighted under monaural conditions [10]. Their activation in right-hemisphere striate and ventral extrastriate areas correlates with the performance in a pointing task to monaurally presented sounds [8]. These results suggest that the enhancement of some auditory skills of blind individuals

*Individual data, plotting bisection thresholds against minimal audible angle. Arrows at the margin show the geometric means of each group as well as the shaded areas of 95% confidence intervals. The blue and green arrows show the average thresholds for 7- and 10-year-old children, respectively (taken from a previous study) [5]. The dashed diagonal line is the equality line: while the thresholds of sighted subjects are scattered around this line, all except one non-sighted subject are above it. Indeed, the only non-sighted subject with bisection threshold that falls within the control range had a threshold for minimal audible angle that was six times lower than the mean of the controls, meaning the subject's data point sits well above the bisection line. With* 

**288**

**Figure 2.**

*permission from Gori et al. [30].*

may reflect in the recruitment of the visual cortex. From this arises the question: what about the impaired skills, such as in the case of the bisection task? If the visual information is important for the development of audio space bisection [45], as we reported in the previous section, then we may expect the visual cortex of sighted and not of blind individuals [30] should be recruited for this audio processing. We recently used EEG to measure activation of the occipital cortex of sighted and blind individuals during the audio bisection task [45, 46]. **Figure 3** illustrates the scalp maps elicited by the second sound of the spatial bisection task when it was delivered from the left (i.e., −4.5°, see left panel) and the right side (i.e., +4.5°, see right panel) in the 50- to 90-ms time window after sound onset, for sighted (**Figure 3A**) and blind participants (**Figure 3B**). In the case of both groups, two strong positivities emerged: one involving central areas and one involving parieto-occipital areas. However, the latter positivity showed a specific contralateral pattern during the spatial bisection task that was only in sighted subjects (**Figure 3A**). In early blind participants (**Figure 3B**), the parieto-occipital response was strongly attenuated and not contralateral to the sound spatial position.

To provide evidence that the early contralateral component that we observed over the occipital scalp actually involved generators in occipital areas, we performed comparisons between groups at the source level (**Figure 4**). Results suggest that sighted subjects showed a stronger occipital and temporal activation contralateral to the physical sound position, while early blind subjects exhibited reduced activation in contralateral cortical areas and an increased activation in ipsilateral cortical areas.

In early blind individuals, the laterality was absent, which means that early visual experience mediates development of this contralateral early occipital response. The data suggest that visual modality plays a key role in the development of an early occipital response that is specific for space perception and auditory stimuli. In sighted subjects, the acoustic recruitment of the visual brain may be necessary to build a spatial metric of the environment using high resolution and flexibility that only the visual brain is capable of implementing. Lack of vision

#### **Figure 3.**

*Scalp maps of the mean ERP amplitude in the selected time window (50–90 ms) after the second sound of the spatial bisection task, for sighted (A) and blind (B) groups. Left and right panels of the figure report the conditions in which S2 was presented from either −4.5° (i.e., narrow first distance) or + 4.5° (i.e., wide first distance), respectively, and independently of timing (±250 ms). With permission from Campus et al. [46].*

#### **Figure 4.**

*Average source activity within the selected time window (50–90 ms) compared between sighted and blind subjects. Left and right panels of the figure report the conditions in which S2 was presented from either the left (i.e., −4.5°, narrow first distance) or the right side (i.e., +4.5°, wide first distance), respectively. We report results of paired two tailed* t *tests with the scale in terms of t-statistic. We also display significant values of t statistic: reddish and bluish colors indicate stronger activations in sighted and early blind subjects, respectively, while intensity indicates magnitude of t (i.e., strength of difference). Only t values corresponding to* p *< 0.0001 after FDR correction appear. Adapted with permission from Campus et al. [46].*

seems to impact the development of this processing and underlying neural circuits, thereby impairing understanding of Euclidean relationships, such as those involved in solving a spatial bisection task. These findings agree with our previous behavioral results [30], at the same time revealing that the neural correlates of the audio space bisection deficit reported in blind individuals might correspond to reduction of early occipital contralateral activation. We speculate that cortical activation underlying the C1 ERP component (usually elicited by visual stimuli) plays a fundamental role in the construction of metrics in the spatial domain independently of the involved sensory modality. Moreover, the construction of spatial metrics may depend on visual experience.
