Neural plasticity in processing of sound location by the early blind: an event-related potential study

Electroencephalography and clinical Neurophysiology, 84 (1992) 469-472

469

© 1992 Elsevier Scientific Publishers Ireland, Ltd. 0013-4649/92/$05.00

EVOPOT 91698

Short c o m m u n i c a t i o n

Neural plasticity in processing of sound location by the early blind: an event-related potential study * Teija Kujala, Kimmo Alho, Petri Paavilainen, Heikki Summala and Risto N?i?it~inen Department of Psychology, University of Helsinki, SF-O0170 Helsinki (Finland) (Accepted for publication: 26 June 1992)

Summary Event-related potentials (ERPs) to a change in the locus of origin of a repetitive sound were studied in early blind human subjects. It was found that the N2b component of the ERP was posteriorly distributed on the scalp to that in the sighted control subjects. This suggests that the blind might use, to a larger extent than the sighted, parietal, or perhaps even occipital, brain areas in sound localization. The present results thus appear to demonstrate plastic changes in neural populations involved in processing of auditory space following early loss of vision. Key words: Blind; Neural plasticity; Audition; Sound localization; Event-related potentials; Visual deprivation Early deprivation of one sensory modality causes plastic changes in neural populations involved in sensory processing. Hyviirinen et al. (1981a,b) showed that in the visual and parietal cortices of the monkey, early visual deprivation decreases the proportion of neurons responsive to visual stimuli and increases that of neurons responsive to active somatic exploration of the surroundings. In addition, Uhl et al. (1991) found a posteriorly larger slow negative DC-potential shift during Braille reading and during a non-reading tactile task in early blind than sighted subjects. Moreover, Neville et al.'s (1983) ERP data suggest enhanced visual processing in auditory areas of deaf subjects. According to Wanet-Defalque et al. (1988), the macroanatomy of the visual cortex of early blind humans is normal. Interestingly, higher glucose metabolism of visual cortex has been reported in blind than (blindfolded) control subjects. The higher metabolism was observed both during rest with no stimulation (Wanet-Defalque et al. 1988; Phelps et al. 1981) and during tactile and auditory discrimination tasks (Wanet-Defalque et al. 1988). The present study aims at determining whether the brain areas, normally processing visual information, participate in processing of auditory spatial information in early blind persons. This will be accomplished by comparing the scalp distributions of ERPs involved in discrimination of a change in the locus of origin of a sound in early blind and sighted control subjects. A physically deviant stimulus in a sequence of repetitive auditory, "standard," stimuli elicits a mismatch negativity (MMN) peaking at 100-200 msec from stimulus onset and reaching its maximum amplitude at frontal or frontoeentral scalp areas (for a review, see N~iiit~inen 1992). The MMN is probably elicited by any discriminable change in an auditory stimulus, also by a change in its locus of origin (Paavilainen et al. 1989), and it is elicited even when the stimuli are unattended. Thus the MMN appears to reflect an automatic, pre-

Correspondence to: Teija Kujala, Department of Psychology, University of Helsinki, Ritarikatu 5, SF-00170 Helsinki (Finland). Tel.: +358-0-90-1913455; Fax: +358-0-90-1913443. * Supported by The University of Helsinki and The Academy of Finland.

perceptual discrimination process (N~i~it~inen 1990). A substantial part of the MMN is generated in supratemporal auditory cortex (see, e.g., Sams et al. 1985). The MMN is followed, and partly overlapped, by the N2b component when the stimulus sequence is attended (N~i~it~inen and Gaillard 1983; N~i~it~inen 1992). The N2b is elicited not only by a stimulus mismatch but also by a stimulus match, depending on which event is the infrequent one (Ritter et al. 1992). The N2b is usually followed by a relatively sharp, frontocentral or central P3a positivity and a slower, parietal positivity (N~iiit~inen 1992), and might be associated with the orienting reflex (N~i~it~inen and Gaillard 1983). In previous studies, auditory ERPs of the blind have shown shorter N1 latencies (Niemeyer and Starlinger 1981) and larger N1, P2, and P3 amplitudes (Woods et al. 1985) than those of the controls. In the present study, the scalp distributions of the MMN and N2b to an infrequent change in the apparent sound location were compared between early blind and sighted subjects. It was hypothesized that the blind might be using, to a larger extent than the sighted, posterior brain areas in the processing of auditory space. If so, this might be reflected by more posterior scalp distributions of the ERP components elicited by a change in the sound location in early blind than control subjects.

Methods There were 9 blind (median age 23 years) and 9 normally sighted (median age 22 years) subjects, 3 males and 6 females in each group. All subjects were students and none of them had previously participated in any ERP experiment. All the blind subjects had a diagnosed peripheral deficit which caused early blindness beginning within the first 2 years of life. Four of them had a retinopathy of prematurity, one had an inherited retinal degeneration, one was blinded due to mother's rubella during pregnancy causing degeneration of eyes, one's eyes were ablated at the age of 1 year due to retinal cancer, and one had a congenital atrophy of the optic tracts. The etiology of one subject's blindness (diagnosed as peripheral) was unknown. Five blind subjects reported still being able to differentiate between dark and light. In addition, one had been capable of doing so until the age

470

T. KUJALA ET AL.

of 5 years and another until the age of 2 years. Every blind subject reported being unable to discriminate the direction of the light. Binaural tone pips (sine-wave bursts of 600 Hz, duration 60 msec including 10 msec rise and fall times, intensity at each ear 75 dB SPL) were presented through headphones at a constant rate of 1 tone/610 msec in blocks of 500 stimuli. The stimulus blocks consisted of "standard" ( P = 0.9) and "deviant" ( P = 0.1) tones occurring in a random order. For the standard tones, the inputs were simultaneous for the two ears, the tones appearing "in the middle of the head." For the deviant tones, the onset of the right-ear input led

that to the left ear by 0.7 msec, the tones therefore appearing "at the right." In the "attend" condition (3 blocks), subjects were instructed to count all deviant tones. In the "ignore" condition (3 blocks), they were instructed to ignore auditory stimuli and to concentrate on reading self-chosen text (Braille text in the blind). The order of the attend and ignore blocks was randomized for each subject. The E E G (0.1-100 Hz, - 3 dB points) was recorded with Ag/AgC1 electrodes at 19 scalp sites. The electrodes were placed on the midline, over the left and right posterior scalp areas, and on the tilted coronary line connecting the mastoids via Fz (see Fig. la). Two

SIGHTED

BLIND ATTEND I

I

/

\

a.

IGNORE

. ~ .

l ....

-' ] "%" ~-'-,

l'v

!

1.,.,~,.. _

Fig. 1. ERPs to standard tones appearing in the middle of the head (dashed line) and to deviant tones appearing "at right" (solid line) for the blind (a and c) and sighted (b and d) subjects in the attend condition (deviants were counted; a and b) and in the ignore condition (subjects were reading; c and d). LM = left mastoid; R M = right mastoid. L3, L1, L2 and L4 are electrodes placed equidistantly on the tilted coronary line connecting the mastoids via Fz.

S O U N D L O C A L I Z A T I O N IN T H E BLIND

471

ATTEND

additional E O G electrodes were placed at the outer canthi of the eyes for recording voltage changes caused by horizontal eye movements. Vertical eye movements wi~re monitored with Fpz electrode. All E E G and E O G electrodes were referred to an electrode attached at the nose. The E E G analysis epochs of 420 msec (sampling rate 250 Hz) began 50 msec before stimulus onset. The first 10 epochs of each block as well as epochs contaminated by blinks, eye movements, muscle activity, or other extracerebral artifacts (voltage variation during an epoch exceeding 150/zV at any electrode) were automatically omitted from averaging. Frequencies higher than 30 Hz were digitally filtered out (FFT-based filter) from the ERPs. Difference waves were obtained by subtracting the standard-tone ERPs from the deviant-tone ERPs separately for the attend and ignore conditions. The MMN and N2b amplitudes were measured from the difference waves in reference to a 50 msec prestimulus baseline. The MMN was identified as the largest negative peak within 70-165 msec post-stimulus period in the ignore condition (see Paavilainen et al. 1989), the N2b as the second negative peak within 135-250 msec post-stimulus period in the attend condition (see N~iii#inen and Gaillard 1983; Paavilainen et al. 1989). Both E R P components were measured at all electrodes as the mean amplitude over a 50 msec period beginning 25 msec before and ending 25 msec after the peak latency (determined at Fz separately for each subject). In Results, the reported significance levels for the F values from analyses of variance (BMDP, see Dixon et al. 1988) are Greenhouse-Geisser corrected when appropriate.

a.

"L

IGNORE

Results

There were no marked differences in the standard-tone ERPs between the groups (Fig. 1 and Table I). The deviant-tone ERPs were negatively displaced in relation to the standard-tone ERPs in both groups (Fig. 1). In the attend condition, two successive negative phases can be discerned. The earlier phase corresponds to the MMN, the later to the N2b (e.g., N~iit~inen 1990). In the ignore condition, only the earlier phase (the MMN) was elicited. Both the MMN and N2b were largest at fronto-central scalp areas (Figs. 1 and 2). A 2-factor analysis of variance (factors: Group, Condition; repeated measures on the latter factor) for the MMN amplitude and latency at Fz showed neither significant differences between the groups or conditions nor significant Group x Condition interactions (see Fig. 2 and Table I). Furthermore, in a single-factor analysis of variance for the N2b amplitude and latency at Fz no significant differences between the groups were found.

MMN

r~

b.

TABLE I The N1, MMN, N2b amplitudes, peak latencies and standard deviations (in parentheses) for the blind and sighted subjects. The N1 was measured at Cz from the standard-tone ERPs, the MMN and the N2b at Fz from the difference waves (obtained by subtracting ERPs to the standards from those to the deviants). Amplitude (/zV) (S.D.)

Latency (msec) (S.D.)

Blind

Sighted

Blind

N1 Attend Ignore

- 1.8 (1.38) - 1.0 (1.08)

- 2 . 0 (1.74) - 1.1 (1.00)

87 (15) 84 (21)

92 (14) 87 (14)

MMN

-

3.5 (3.04)

- 3.1 (1.93)

122 (33)

123 (33)

N2b

- 4.5 (4.17)

- 4.6 (4.17)

208 (32)

209 (34)

Sighted

Fig. 2. The difference waves, corresponding to Fig. 1, obtained by subtracting the ERPs to the standard tones from those to the deviant tones. Continuous line = blind subjects; dashed line = sighted subjects.

In the blind, the N2b amplitudes tended to be larger at the posterior electrode sites compared with those of the sighted (Fig. 2; Pz: - 3 . 9 / z V vs. - 1 . 3 / z V , F (1, 16)=3.50, P < 0.08; 0 3 : - 4 . 0 / z V vs. - 0 . 1 /xV, F (1, 16) = 5.15, P < 0.04; 0 1 : - 3 . 2 /zV vs. 0.4/zV, F (1, 16)= 3.99, P < 0.065; Oz: - 2 . 7 ~V vs. 0.4 /zV, F (1, 16)= 3.38, P < 0.09; 0 2 : - 3 . 0 3 / z V vs. 0.20 p.V, F (1, 16)= 2.84, P < 0.12; 04: - 4 . 0 ~V vs. - 0 . 5 /zV, F (1, 16)= 3.18, P < 0.10). To compare the N2b amplitude distributions between the groups, these amplitudes at the midline electrodes (Fpz, Fz, Cz, Pz, Oz) were separately normalized for each subject by their corresponding vector length (cf.,

472

T. KUJALA El" AL.

McCarthy and Wood 1985). Two-factor analyses of variance (factors: Group, Electrode; repeated measures on both factors) were performed for the normalized amplitudes within each group. The same procedures were also performed for the MMN. For the normalized midline N2b amplitude, there was a significant Group× Electrode interaction, the N2b amplitude in the blind being posteriorly distributed to that i ~ d a e sighted ( F (4, 64)= 4.21, P < 0.025). No significant differenc~~in the MMN midline distribution between the groups could be found in either condition, although there was a trend toward this distribution in the blind being posterior to that in the sighted (see Pz in Fig. 2). In contrast to the N2b, the P3 amplitude distribution of the blind appeared to be anterior to that in the sighted (see Fig. la). In many subjects, however, the P3 peak latency, unfortunately, exceeded the end of the present post-stimulus analysis period of 370 msec. Therefore, this latency, as well as the P3 amplitude, could not be measured. Discussion

In the blind, the N2b scalp distribution for changes in apparent sound location was posterior to that in the sighted. The MMN showed a similar, but insignificant, tendency. The large posterior N2bs of the blind might originate from parietal, or even occipital, brain areas. The disuse of the visual system for peripheral reasons deprives large brain areas of their normal sensory input. These areas might then develop new, non-visual, functions, at least if the blindness has started at a very early age. In a parallel study (Alho et al. submitted), the present blind subjects showed a selective-attention effect on ERP (the processing negativity) for attended sound locations which was distributed on the scalp posteriorly to that of the sighted. These converging data hence suggest that attention-related brain mechanisms might be especially plastic. Such changes in the neural basis of attentive functions have also been reported elsewhere: blindness with an early onset in the monkey causes an increase in the proportion of cells responsive to active tactile task but not to passive somatosensory stimulation in the occipital (Hyv~irinen et al. 1981a) and parietal (Hyviirinen et al. 1981b) areas. In addition, Neville et al. (1983) reported plastic changes in response to attended visual stimuli in the neural structures of the deaf. In conclusion, the present results, in parallel with those of Alho et al. (submitted), suggest a high degree of neural plasticity underlying perception of auditory space. This could be due to the high survival value of spatial information for the blind, for example in locomotion (see, e.g., Strelow and Brabyn 1982). Our preliminary pilot data suggest, however, similar plastic changes in the N2bs to pitch changes as in those to location changes in the blind. In addition, Simpson and his colleagues have recently found N2bs which were posterior in the blind to those in the sighted in response to a duration change of a sound (G.V. Simpson, personal communication, March 1992). ~ "~

References

Alho, K., Kujala, T., Paavilainen, P., Summala, H. and N~iiit~inen, R. Selective-attention effect on the auditory event-related potential in early blind human subjects. Submitted.

'\

Dixon, W.J., Brown, M.B., Engelman, L., Hill, M.A. and Jennrich, R.I. (Eds.). BMDP Statistical Software Manual. University of California Press, Berkeley, CA, 1988. Hyviirinen, J., Carlson, S. and Hyv~irinen, L. Early visual deprivation alters modality of neuronal responses in area 19 of monkey cortex. Neurosci. Lett., 1981a, 26: 239-243. Hyv~irinen, J., Hyv~irinen, L. and Linnankoski, I. Modification of parietal association cortex and functional blindness after binocular deprivation in young monkeys. Exp. Brain Res., 1981b, 42: 1-8. McCarthy, G. and Wood, C.C. Scalp distributions of event-related potentials: an ambiguity associated with analysis of variance models. Electroenceph. clin. Neurophysiol., 1985, 62: 203-208. N~iiitiinen, R. The role of attention in auditory information processing as revealed by event-related brain potentials. Behav. Brain Sci., 1990, 13: 201-288. N~iiitiinen, R. Attention and Brain Function. Erlbaum, Hillsdale, N J, 1992. N~iiitiinen, R. and Galliard, A.W.K. The orienting reflex and the N2 deflection of the event-related potential (ERP). In: A.W.K. Galllard and W. Ritter (Eds.), Tutorials in ERP Research: Endogenous Components. North-Holland Publishing Company, Amsterdam, 1983: 119-141. Neville, H.J., Schmidt, A, and Kutas, M. Altered visual-evoked potentials in congenitally deaf adults. Brain Res., 1983, 266: 127-132. Niemeyer, W. and Starlinger, I. Do the blind hear better? Investigations on auditory processing in congenital or early acquired blindness. II. Central functions. Audiology, 1981, 20: 510-515. Paavilainen, P., Karlsson, M.L., Reinikainen, K. and Niiiitiinen, R. Mismatch negativity to change in spatial location of an auditory stimulus. Electroenceph. clin. Neurophysiol., 1989, 73: 129-141. Phelps, M.E., Mazziotta, J.C., Kuhl, D.E., Nuwer, M., Packwood, J., Metter, J. and Engel, J. Tomographic mapping of human cerebral metabolism: visual stimulation and deprivation. Neurology, 1981, 31: 517-529. Ritter, W., Paavilainen, P., Lavikainen, J., Reinikainen, K., Alho, K., Sams, M. and N~iiit§nen, R. Event-related potentials to repetition and change of auditory stimuli. Electroenceph. din. Neurophysiol., 1992, in press. Sams, M., H~imiil~iinen, M., Antervo, A., Kaukoranta, E., Reinikainen, K. and Hari, R. Cerebral neuromagnetic responses evoked by short auditory stimuli. Electroenceph. clin. Neurophysiol., 1985, 61: 254-266. Strelow, E.R. and Brabyn, J.A. Locomotion of the blind controlled by natural sound cues. Perception, 1982, 11: 635-640. Uhl, F., Franzen, P., Lindinger, G., Lang, W. and Deecke, L. On the functionality of the visually deprived occipital cortex in early blind persons. Neurosci. Lett, 1991, 124: 256-259. Wanet-Defalque, M.-C., Veraart, C., De Voider, A., Metz, R., Michel, C., Dooms, G. and Goffinet, A. High metabolic activity in the visual cortex of early blind human subjects. Brain Res., 1988, 446: 369-373. Woods, D.L., Clayworth, C.C. and Bach-y-Rita, P. Early blindness reorganizes auditory processing in humans. In: Abstracts, Society for Neuroscience, 15th Annual Meeting, Dallas, TX, Oct. 20-25, 1985: 449.