Functional brain imaging of tinnitus-like perception induced by aversive auditory stimuli

Size: px

Start display at page:

Download "Functional brain imaging of tinnitus-like perception induced by aversive auditory stimuli"

Augustus Wright
5 years ago
Views:

1 BRAIN IMAGING Functional brain imaging of tinnitus-like perception induced by aversive auditory stimuli Frank Mirz, 1,2,3,CA Albert Gjedde, 2 Hans Sùdkilde-Jrgensen 3 and Christian Brahe Pedersen 1 1 Department of Otorhinolaryngology, 2 PET-Centre, and 3 MR-Research Centre Aarhus University Hospitals, Nùrrebrogade 44, 8200 Aarhus C, Denmark CA,1 Corresponding Author and Address Received 10 November 1999; accepted 10 December 1999 Acknowledgements: We thank the experimental subjects for their patience and efforts. This study was supported by grants from Landsforening for Bedre Hùrelse. MeÂnieÁre-Tinnitus-Forening and MRC Denmark. Tinnitus is an aversive auditory percept of unknown origin. We tested the speculation that tinnitus may share neuronal processing mechanisms with aversive auditory percepts of known origin. This study revealed the functional neuroanatomy of the perception of aversive auditory stimuli. The stimuli were presented to 12 healthy volunteers so as to mimic the psychoacoustical features of tinnitus and its affective response in tinnitus sufferers. The regional cerebral blood ow distribution was measured by PET during four auditory processing conditions and one control condition. The aversive auditory stimuli activated primary and secondary auditory areas bilaterally, dorsolateral prefrontal attention areas, and structures in the limbic system which subserve emotional processing. Based on these results and ndings from other functional neuroimages of tinnitus, we hypothesize that the perception of tinnitus may involve the functional linkage of these brain areas: secondary auditory cortex, dorsolateral prefrontal cortex, and limbic system. NeuroReport 11:633±637 & 2000 Lippincott Williams & Wilkins. Key words: Attention; Aversion; Emotion; Positron emission tomography; Tinnitus INTRODUCTION Tinnitus-related neuronal activity may originate nearly anywhere in the auditory system [1].Only a few hypotheses concerning the pathomechanisms and the involved generator sites have been supported by neurophysiological evidence from human or animal studies [2,3]. In animal studies tinnitus induced pharmacologically (salicylate) or by noise trauma, has been attributed to aberrant activity in the inferior colliculus and other auditory brain stem nuclei [4,5]. Other studies implicated the reorganization of auditory cortical elds due to cortical plasticity as a result of deafferentation of the periphery of the auditory system in the perception of tinnitus [6]. Most current models of the tinnitus phantom percept comply with the notion that the conscious perception of tinnitus must involve the cerebral cortex, as do all conscious percepts [1]. PET and fmri revealed a variety of auditory processing areas, prefrontal attention sites or limbic system structures involved in the perception of tinnitus [7±11]. Throughout these studies the various patterns of tinnitus-related activity seem to be caused by the heterogeneity of the psychoacoustical features of the examined tinnitus and the varying affective impacts of tinnitus in the lives of the recruited patients. To avoid the unpredictable in uences of these factors, we created a model of tinnitus in healthy volunteers. To reveal the cortical centres involved in an auditory process comparable to the perception of tinnitus, we mimicked tinnitus by presenting different aversive auditory sounds. The sounds produced auditory sensations psycho-acoustically similar to tinnitus and evoked adequate negative emotional responses. MATERIALS AND METHODS Subjects: Twelve right-handed, healthy volunteers ( ve women, seven men, aged 20±49 years, mean age 25 years) without tinnitus participated in the study. Complete otoneurological and audiological evaluations revealed no abnormal ndings. The protocol followed the Helsinki Declaration II and was approved by the Aarhus County Research Ethics Committee. Auditory stimuli: Twenty different aversive sounds were synthesized to imitate tinnitus. These sounds elicited adequate auditory sensations and emotional responses similar to the psycho-acoustic effects of tinnitus. The sound features were based on psychoacoustical descriptions of & Lippincott Williams & Wilkins Vol 11 No 3 28 February

2 F. MIRZ ET AL. tinnitus by a large number of tinnitus patients treated at our department and other descriptions [12]. Ten subjects not involved in the PET study determined and selected the four most aversive stimuli (T1±T4) by rating annoyance, dislike, nuisance and disturbance caused by those sounds on a visual analog scale (VAS) consisting of 100 mm lines with endpoints denoted by the words total absence and maximum of annoyance, dislike, nuisance and disturbance. The four chosen sounds achieved signi cantly higher scores on the VAS than the rest of the aversive sounds. Two of the highest rated sounds were pure tones with a pitch of 8 khz and a loudness of 85 db sound pressure level. The rst of these sounds (T1) contained no bandwidth, whereas the second sound (T2) had a 30 Hz bandwidth around the centre frequency. A daily life sound, produced by scraping a knife against a plate, was recorded and resampled to sound like an engine room. This sound was presented either without (T3) or with (T4) 40 randomly placed gaps of 20±50 ms duration. Both sounds were presented at a sound pressure level of 85 db. The auditory stimuli were contrasted to a baseline scan in silence (B1). Tasks T1±T4 and B1 were counterbalanced across subjects. Auditory stimuli were recorded on a digital tape recorder Sony TCD-D7 and delivered binaurally using an Orbiter 922 audiometer and EARtone 3A earphones. Data acquisition, image and statistical analyses: PET imaging with H 15 2 O as tracer was performed with the ECAT Exact HR47 tomograph (Siemens/CTI) in 3D mode. For attenuation correction GA-68 transmission tomograms were acquired in 2D mode. Five emission scans were initiated at true cps after bolus injection of 500 Mbq H 15 2 O in single 40 s frames. All scans were performed in a quiet darkened room. Subjects were supine with the head held xed in a vacuum pillow. The presentation of auditory stimuli was initiated 10 s prior to injection. PET images were reconstructed after correction for attenuation and scatter in a resolution of 12 mm FWHM (Hann- lter, cutoff frequency 0.15 cycles/s). T1-weighted MR brain images were acquired on a Philips 1.5-T gyroscan (fast- eld-echo sequence, 64 sagittal 2 mm slices, TE ˆ 21.6 ms, TR ˆ 41.7 ms) and co-registered to the PET-images and the Talairach coordinate system [13]. Besides Talairach coordinates tentative transformations of these coordinates into Brodmann areas (BA) were used for anatomical localization of the obtained PET results. t-statistic maps were created after a pixel-by-pixel subtraction of PET volumes using DOT (two-tailed t-statistic, approximated to a standard Gaussian distribution; pooled s.d. of all intracranial voxels) [14]. Searching the cerebral cortex and limbic structures in the temporal lobes (600 ml), t. 4.4 equals signi cance at p, Subtraction analyses were performed on PET volumes of individual aversive stimulations vs baseline (T1 B1, T2 B1, T3 B1, T4 B1), as well as on a combined PET volume of all stimulations vs baseline (Tavg B1). RESULTS Analysis of the combined subtraction (Tavg B1) of all tasks showed signi cantly increased activity in the primary auditory cortex (Brodmann area, [BA] 41) in both hemispheres, and in associative auditory regions in the right hemisphere (BA 21, 22; Table 1, Fig. 1). Several sites in the dorsolateral prefrontal cortex (middle and inferior frontal gyri) were activated bilaterally although with a preponderance of right-sided activations (BA ). Anterior midline structures (superior frontal and medial frontal gyri, BA 6 8 9), sites in the inferior parietal lobule (BA ), as well as structures in the limbic system (amygdala/parahippocampal gyrus, hippocampus) showed increased activity (Table 1, Fig. 1). Analyses of subtractions of the individual tasks (T1 B1, T2 B1, T3 B1, T4 B1) showed the same tendency of activation, although without complete consistency. DISCUSSION The processing and perception of aversive auditory stimuli was associated with activation of primary and associative auditory areas. The activation of Heschl's gyrus (BA 41) is attributed to the early sensation of the externally presented stimuli, whereas activation of associative auditory areas located more anteriorly (BA 21), and more posteriorly at the temporo-parietal junction (BA 22), may represent higher order processing [15,16]. Recent brain imaging studies of tinnitus patients revealed sites of activation in similar parts of the auditory cortex. Arnold et al. discovered an increased metabolic activity in the left primary auditory cortex (Heschl's gyrus, BA 21) when comparing the central activity measured with FDG-PET of tinnitus patients and healthy volunteers [7]. Cacace et al. examined individuals who after complete unilateral deafferentation of the auditory-vestibular system were able to evoke tinnitus by cutaneous stimulation of the ngertip [8]. With functional magnetic resonance scanning they found tinnitus-associated cortical activity in the temporo-parietal region contralateral to the cutaneous stimulation. In a study by Lockwood et al., tinnitus patients with changeable tinnitus loudness based on oral-facial movements were PETscanned in different conditions revealing tinnitus-relevant activity in the auditory cortex (BA 21/41) and in a part of the limbic system [9]. Giraud and co-workers scanned patients who could evoke tinnitus by speci c eye movements [10]. Their study showed that associative auditory areas in the temporo-parietal region (BA 22/42) were activated during the perception of tinnitus. In a PET study by Mirz et al. tinnitus was modi ed pharmacologically with lidocaine or acoustically with narrow-band noise [11]. A within-subject paradigm was adopted to compare conditions with and without tinnitus. The study showed that auditory association areas in the right temporal cortex (BA 21) and multimodal attention sites in the right prefrontal part of the brain (BA 8) were associated with tinnitus. The tinnitus suppressing effect of lidocaine was also used in a SPECT study by Staffen et al. [17]. Tinnitus-related neuronal activity was located predominantly in the right primary auditory cortex (BA 41). The modulation of the spatial distribution and extent of activated auditory areas appear also to depend on the degree of attention-related processing. The magnitude and extent of activation in auditory areas is enhanced when attention is involved [18]. However, auditory attention also activates neuronal systems other than those involved in auditory processing [19]. In the present study the perception and further proces- 634 Vol 11 No 3 28 February 2000

3 FUNCTIONAL BRAIN IMAGING OF TINNITUS-LIKE PERCEPTION INDUCED BY AVERSIVE AUDITORY STIMULI Table 1. Sites of activation. Anatomical localization Subtraction T1 B1 T2 B1 T3 B1 T4 B1 Tavg B1 BA x,y,z (t) BA x,y,z (t) BA x,y,z (t) BA x,y,z (t) BA x,y,z (t) R transverse temporal gyrus 41 44, 23,14 (4.7) 41 39, 30,11 (5.1) 41 43, 23,11 (5.8) 41 40, 23,6 (7.8) 41 42, 23,9 (9.8) L transverse temporal gyrus 41 39, 32,12 (3.9) 41 40, 16,2 (4.6) 41 42, 31,12 (5.3) 41 42, 26,0 (7.7) 41 40, 33,12 (9.2) R superior temporal gyrus 22 67, 40,11 (4.3) 42 58, 32,11 (4.5) 42 66, 37,12 (4.7) 22 60, 33,11 (5.7) 22 64, 38,11 (8.1) L superior temporal gyrus 42 42, 35,12 (3.2) R middle temporal gyrus 21 64, 2, 9 (2.7) 21 50, 4, 21 (4.0) 21 59, 14,0 (4.4) R inferior temporal gyrus 21 51, 6, 24 (3.8) L inferior temporal gyrus 44 47,8,21 (2.8) 20 47, 26, 21 (2.6) 20 46, 33, 15 (3.6) R superior frontal gyrus 9 17,49,41 (3.2) 8 11,29,60 (3.8) 8 7,46,51 (4.8) L superior frontal gyrus 8 15,36,48 (2.6) 9 21,48,38 (3.3) 9 1,49,35 (3.8) R middle frontal gyrus 46 47,48,26 (2.7) 6 50,8,47 (3.9) 8 56,12,39 (4.6) R inferior frontal gyrus 45 60,29,9 (3.2) 44 47,17,20 (4.0) 45 58,36,11 (3.3) 47 46,30, 6 (4.4) L inferior frontal gyrus 45 46,25,5 (3.4) 45 44,24,6 (4.3) Right medial frontal gyrus 8 7,44,50 (3.0) 6 1,6,56 (2.8) 9 9,51,17 (3.4) Left medial frontal gyrus 6 3,10,65 (3.3) 6 1,8,62 (4.4) R inferior parietal lobule 7 34, 56,60 (3.4) L inferior parietal lobule 39 46, 64,38 (2.9) R hippocampus 20, 49,11 (3.0) 28, 11, 14 (3.1) L hippocampus 21, 37,5 (3.0) 23, 35,6 (3.6) 21, 30, 2 (3.4) 21, 33,5 (5.6) R parahippocampal gyrus 16, 4, 15 (5.1) 16, 9, 27 (3.2) 16, 7, 26 (5.7) L parahippocampal gyrus 20, 19, 12 (2.6) R thalamus 15, 16,17 (3.7) L thalamus R cingulate gyrus 24 11, 16,45 (3.3) L insula 38, 4,11 (3.0) The anatomical localization of the activated sites is based on the Talairach coordinate system and combined with a tentative transformation of the coordinates into Brodman areas (BA) [13]. Talairach coordinates in mm: x (medial-lateral position relative to midline, right ( )/left ( )), y (anterior-posterior position relative to anterior commissure, anterior ( )/posterior ( )), and z (superior-inferior position relative to the intercommissural plane, superior ( )/inferior ( )). Vol 11 No 3 28 February

4 F. MIRZ ET AL. Fig. 1. The t-statistic map of the analysis of all test conditions (Tavg B1) superimposed on transaxial (upper rows) and coronal (lower rows) slices of the average brain MR image of all 12 subjects. Dotted lines indicate position of the intersectional plane between transaxial and coronal images. Abbreviations and coordinates (y and z) according to Talairach [12]. GC, cingulate gyrus; GFd, medial frontal gyrus; GFi, inferior frontal gyrus; GFm, middle frontal gyrus; GFs, superior frontal gyrus; GTi, inferior temporal gyrus; GTm, middle temporal gyrus; GTs, superior temporal gyrus; GTT, transverse temporal gyrus; Hi, hippocampus; LPi, inferior parietal lobule; NA/GR, amygdale/parahippocampal-gyrus complex; R, right; L, left. sing of aversive auditory stimuli activated dorsolateral prefrontal structures and areas in the inferior parietal lobe. These ndings are consistent with results from several neuroimaging studies concerned with attentional processing of multimodal stimuli [20]. The recruitment of attentional resources has also been implicated in the processing of the auditory phantom perception of tinnitus [3,11,21]. Perception of tinnitus is hypothesized to cause continuous attention to the sound or sounds which consequently may result in a vicious circle of reinforcement of the strength of the perceived initial tinnitus-related neuronal activity [22]. The limbic structures activated in the present study (amygdala/parahippocampal gyrus) are involved in processing the affective content of externally presented or internally generated information, as shown in different modalities [23]. Moreover, these structures also subserve emotional memory [24]. Emotionally arousing stimuli, such as the aversive sounds in this study, seem to be stored in memory systems which, when presented with these sounds on later occasions, will evoke much faster and probably more distinct emotional responses. When applied to tinnitus this explanation means that the initial neuronal activity related to tinnitus may trigger an auditory sensation, which then evokes a negative emotional response. This 636 Vol 11 No 3 28 February 2000

5 FUNCTIONAL BRAIN IMAGING OF TINNITUS-LIKE PERCEPTION INDUCED BY AVERSIVE AUDITORY STIMULI pattern of activation is stored in the emotional memory system, which then persistently delivers the perceptually negative experience of the tinnitus sound. Activation of the insula, a polymodal convergence area, which has been suspected of relaying sensory information into the limbic system [25], may re ect the mediation between sensory and affective functions. The combined activation of limbic system structures and the insula support the notion that the sounds used in the present study indeed were aversive and evoked a signi cant emotional response in the subjects. Descriptions given by tinnitus patients of the loudness and other psychoacoustical features of their perceived tinnitus sound or sounds are similar to the descriptions given by subjects who evaluated the sounds used in the present study [12]. The affective responses to the aversive stimuli in this study (measured on VAS) were also comparable to results from studies evaluating the impact of tinnitus on the psychological state of tinnitus sufferers [26]. It appears that the perceptual experience of speci c aversive auditory stimuli led to other perceptual experiences: a phenomenon known as synaesthesia. Thus, our results indicate that the acoustic perception as well as the induced affective response in our subjects successfully imitated the state of tinnitus sufferers. CONCLUSION Our results suggest that the perception of tinnitus-like auditory stimuli engages auditory sensory and processing areas, recruits attentional resources in the right prefrontal areas of the brain and generates emotional responses based on activation of the limbic system. As the psychoacoustical features of tinnitus are similar to the stimuli used in the present study, we hypothesize that the perception of tinnitus may involve the same functionally linked brain areas. Results from recent brain imaging studies on tinnitus patients support this hypothesis. Most results obtained in these previous studies were replicated in the present study. Our tinnitus model may help to specify the parts and processes of the brain to be studied in the future to reveal more information on the central mechanisms of tinnitus perception. REFERENCES 1. Jastreboff PJ. Neurosci Res 8, 221±254 (1990). 2. Melcher JR, Sigalovsky IS, and Levine RA. Tinnitus-related fmri activation patterns in human auditory nuclei. In: Hazell J, ed. Proceedings of the Sixth International Tinnitus Seminar. London: The Tinnitus and Hyperacusis Centre, 1999: 166± Norena A, Cransac H and Chery-Croze S. Clin Neurophysiol 110, 666±675 (1999). 4. Chen G-D and Jastreboff P-J. Hear Res 82, 158±178 (1995). 5. WallhaÈusser-Franke E and Langner G. Central activation patterns after experimental tinnitus induction in an animal model. In: Hazell J, ed. Proceedings of the Sixth International Tinnitus Seminar. London: The Tinnitus and Hyperacusis Centre, 1999: 155± Eggermont JJ and Komiya H. Auditory cortex reorganization after noise trauma: Relation to tinnitus? In: Hazell J, ed. Proceedings of the Sixth International Tinnitus Seminar. London: The Tinnitus and Hyperacusis Centre, 1999: 146± Arnold W, Bartenstein P, Oestreicher E et al. ORL (Basel) 58, 195±199 (1996). 8. Cacace, AT, Cousins JP, Parnes SM et al. Audiol Neurootol 4, (1999). 9. Lockwood AH, Salvi RJ, Coad ML et al. Neurology 50, 114±120 (1998). 10. Giraud AL, Chery-Croze S, Fischer G et al. Neuroreport 10, 1±5 (1999). 11. Mirz F, Pedersen CB, Ishizu K et al. Hear Res 134, 133±144 (1999) 12. Meikle M and Taylor-Walsh E. J Laryngol Otol Suppl. 9, 17±21 (1984). 13. Talairach J and Tournoux P. Co-planar Stereotaxic Atlas of the Human Brain: 3-Dimensional Proportional System: An Approach to Cerebral Imaging. Stuttgart: Georg Thieme Verlag, Worsley KJ, Evans AC, Marrett S et al. J Cerebr Blood Flow Metab 12, 900±918 (1992). 15. Herzog H. Cortical activation by auditory stimulation studied with positron emission tomography. In: Thatcher RW, Hallett M, Zef ro T et al., eds. Functional Neuroimaging. Technical Foundations. San Diego: Academic Press, 1994: 59± Alho K, Tervaniemi M, Huotilainen M et al. Psychophysiology 33, 369± 375 (1996). 17. Staffen W, Biesinger E, Trinka E et al. Audiology 38, 53±57 (1999). 18. Grady CL, Van Meter JW, Maisog JM et al. Neuroreport 8, 2511±2516 (1997). 19. Benedict RH, Lockwood AH, Shucard JL et al. Neuroreport 9, 121±126 (1998). 20. Pardo JV, Fox PT and Raichle ME. Nature 349, 61±64 (1991). 21. Jacobson GP, Calder JA, Newman CW et al. Hear Res 97, 66±74 (1996). 22. Jastreboff PJ. The neurophysiological model of tinnitus and hyperacusis. In: Hazell J, ed. Proceedings of the Sixth International Tinnitus Seminar. London: The Tinnitus and Hyperacusis Centre, 1999: 32± Lane RD, Reiman EM, Bradley MM et al. Neuropsychologia 35, 1437±1444 (1997). 24. LeDoux JE. Behav Brain Res 58, 69±79 (1993). 25. Roland PE. Brain Activation. New York: Wiley-Liss, Van Veen ED, Jacobs JB and Bensing JM. J Laryngol Otol 112, 258±263 (1998). Vol 11 No 3 28 February

Positron emission tomography of cortical centers of tinnitus

Hearing Research 134 (1999) 133^144 Positron emission tomography of cortical centers of tinnitus Frank Mirz a;b; *, Christian Brahe Pedersen a, Koichi Ishizu b, Peter Johannsen b, Therese Ovesen a, Hans