One set of sounds, two tonotopic maps: exploring auditory cortex with amplitude-modulated tones

https://doi.org/10.1016/j.clinph.2003.12.034Get rights and content

Abstract

Objective: The possibility of simultaneously observing activation of primary and secondary auditory cortices has been demonstrated by Engelien et al. [Hear Res 2000;148:153–60].

Methods: Such a dual monitoring by means of neuromagnetic recordings can be achieved when a subject is stimulated by brief pulses of 40 Hz-modulated tones. Depending on the frequency filter applied, either the steady-state field (SSF) or the N1m can be extracted from the evoked magnetic field complex.

Results: Using this ‘combined’ (two-maps) paradigm with 4 carrier frequencies, we show that it is possible to synchronously screen two tonotopic maps—one map each reflected either by the SSF or the N1m. Indicators are the systematic variation in the location (higher frequencies are more posterior) and orientation (higher frequencies oriented differently in the saggital plane) of the equivalent current dipole (ECD). These parameters were compared with those obtained from ‘classic’ (one map) paradigms in which either a pure tone elicits an N1m or a 40 Hz continuous (3 s) stimulation produces an SSF. Overall the results were similar, however, systematic differences between the paradigms were found for ECD localization, dipole strength, amplitude, and phase.

Conclusions and Significance: One possible interpretation of these results is that different tonotopically arranged cortical fields were involved in the generation of the components.

Introduction

From both animal (Kaas et al., 1999, Merzenich et al., 1975) and human studies (Pantev et al., 1993, Pantev et al., 1996, Romani, 1986, Romani et al., 1982a, Romani et al., 1982b) it is well-known that neurons in the auditory cortex are arranged tonotopically (i.e. frequency-specific), thus resembling the spatial order of the inner hair cells on the basilar membrane. However, speaking of the tonotopic organization of the auditory cortex can be misleading. The auditory cortex can be divided into several areas, each comprising multiple subareas. Each subarea exhibits tonotopy to a varying degree and manner (Rouiller, 1997). According to Kaas et al. (1999) the main hierarchically organized areas in primates are the core (including the primary auditory field, AI), the belt and the parabelt, out of which the core can be considered the actual primary auditory cortex. The different areas of the core show the highest tonotopic order with an either anterior/anterolateral to posterior/posteromedial or directly opposite gradient depending on the field (see also Ehret, 1997).

So far, magnetencephalography (MEG) has proven to be a successful non-invasive technique to study the tonotopic organization of the auditory cortex in humans. The standard MEG approach to investigate tonotopy consists of source localization of the N1m (operationalized as a single equivalent current dipole (ECD) in a homogeneous sphere) elicited by pure tones of a few hundred milliseconds duration with varying frequency. Although, the first study (Romani et al., 1982a, Romani et al., 1982b) to non-invasively demonstrate tonotopy in man used steady-state stimuli, such attempts have been scarce (Pantev et al., 1996). Typical auditory steady-state stimuli are amplitude- or frequency-modulated tones. Initial perturbations of the signal can be discarded if the stimuli are sufficiently long (seconds to minutes). Typically a modulation rate of 35–40 Hz is chosen in order to obtain maximum response energy (Galambos et al., 1981, Hari et al., 1989, Pastor et al., 2002, Roß et al., 2000).

In the following, we will refer to these two conventional approaches (i.e. elicitation of the N1m by pure tones and the steady-state field (SSF) by continuous 40 Hz modulated tones) as classic paradigms. They are differentiated from a combined paradigm, a term introduced by Engelien et al. (2000). It refers to the possibility of extracting the transient N1m- and the steady-state response to the same stimulus, which has been first shown by Mäkelä and Hari (1987). A simultaneous—and thus time-economic—measurement of both components is desirable, since sources for these components are associated with activations from primary or secondary auditory cortex, respectively (see below). Engelien et al. raised the question whether the N1m and SSF gained at the same time via the combined approach would lead to identical source localizations as if assessed by the classic approach. These authors presented a pure tone with a carrier frequency of 250 Hz and a duration of 500 ms for elicitation of the N1m, and a continuous 200 s 39 Hz amplitude-modulated tone was used for elicitation of the SSF, thus representing classic paradigms. A third—combined—paradigm presented the same carrier frequency as 500 ms 39 Hz amplitude-modulated tones, thus combining features of both aforementioned paradigms, in order to gain the N1m and SSF simultaneously. The authors demonstrated that the components obtained simultaneously and their classcial counterparts (a) showed the same pattern of differences and (b) seemed not to differ in location. However, only one single carrier frequency was used, i.e. no information was obtained with respect to tonotopy. The main goal of the present study was to see if and which tonotopic maps can be obtained by using the combined paradigm described above on 4 different carrier frequencies.

In the past, 3 neuromagnetic measures served as indicators for the presence of tonotopically arranged neuronal populations (fields) in the auditory cortex in humans: (a) change of the location of a component as a function of the frequency of the eliciting stimulus, (b) spatial separation of tonotopic gradient in dependence of the component investigated, and (c) orientation of the dipole.

For the N1m some authors have found a low-to-high frequency-specific arrangement of ECD sources on the anterior–posterior axis (Elberling et al., 1982). However, the great majority of MEG studies localize the main tonotopic gradient on a medial–lateral axis (Elbert et al., 2002, Lütkenhöner and Steinsträter, 1998, Pantev et al., 1993) with the depth of the source linearly increasing with the logarithm of the carrier frequency. Those studies investigating SSF have also reported tonotopy for high-to-low frequencies in the medial–lateral direction (Pantev et al., 1996, Romani et al., 1982a, Romani et al., 1982b, Roß et al., 2000). While the SSF map seems consistent with primate studies (Kaas et al., 1999), BOLD imaging in humans confirms the significance of the spatial gradient for the N1m showing that the center of activation shifts in a posterior and medial directions for higher frequencies (Lauter et al., 1985, Wessinger et al., 2001).

Although the direction of the tonotopic gradient seems to be the same for the N1m and SSF, ECD sources have significantly different locations, thus suggesting that different areas of the auditory cortex may be activated. In comparison to the N1m, which has been associated with activation of the secondary auditory cortex and lateral parts of Heschl's gyrus (Godey et al., 2001, Lütkenhöner and Steinsträter, 1998), sources for SSF seem to be located more anterior and more medial and have been related to activation of the primary auditory cortex (Engelien et al., 2000, Pantev et al., 1996, Tiihonen et al., 1989). Overall, data from neuromagnetic studies confirm that the human auditory cortex—like the one in primates—can be subdivided into different tonotopically arranged fields. This view is supported by recent fMRI (Talavage et al., 2000) and post-mortem histochemical studies (Wallace et al., 2002). However, the exact number of different fields in human auditory cortex remains a matter of controversy (Talavage et al., 2000).

For the N1m, some authors reported a systematic relationship between frequency and dipole orientation (Tiitinen et al., 1993, Verkindt et al., 1995). Generally, when viewed from the lateral surface in the saggital plane, the angle in the right hemisphere rotates clockwise with increasing frequency of the stimulus. According to Verkindt et al. (1995) this might be due to gyral folding characteristics varying with depth. Tiitinen et al. (1993) were able to show the same pattern for the Mismatch Negativity. To the best of our knowledge, no work has yet investigated whether such an association between dipole orientation and frequency also exists for the SSF.

According to the criteria mentioned above dipole localizations on two of the 3 coordinate axes (posterior–anterior, medial–lateral) and dipole orientation in the saggital plane served as indicators of tonotopy. Sources of the N1m and SSF were regarded as activations from different auditory fields if they were separable in space. The results from the combined paradigm were compared to results obtained with the classic paradigms to elicit an N1m and an SSF. We compared the following dependent variables: localization on the 3 coordinate axes, dipole orientation, dipole strength (Q), peak amplitude and peak latency/phase (in case of the N1m or SSF, respectively).

Section snippets

Subjects

Eleven right-handed participants (6 females; age range 20–28 years) without neurological or otological disorders gave informed consent after obtaining written and oral information on the nature of the study. They received € 15 for participation.

Neuromagnetic recording

Neuromagnetic data were recorded (A/D conversion rate: 678.17 Hz; 0.1–200 Hz bandpass) with a 148 channel whole-head magnetometer (4D Neuroimaging Inc., San Diego, CA). Vertical and horizontal electro-oculogram (EOG) were measured from above and below

N1m peak latency

From N1m waveforms depicted in Fig. 1a it can be seen that the N1m deflections peaked at approximately 100 ms. However, for the combined paradigm they showed a consistent delay of about 7 ms for all the frequencies (M=104.54 ms, SE 1.62) compared to their counterparts derived from the classic paradigm (M=97.47 ms, SE 1.64; F1,10=46.66, P<0.001). A significant effect on latency was found for the factor frequency (F3,30=3.89, P<0.02). This effect was due to a later N1m response for the 6000 Hz

Discussion

Compared to other MEG-studies the direction of tonotopy we found is rather unusual, yet not surprising. It is in agreement with the study by Elberling et al. (1982), animal studies (see Kaas et al., 1999) and functional imaging studies (Lauter et al., 1985, Wessinger et al., 2001). From neuroanatomy it is also known that Heschl's gyrus does not only extend medially, but also in a posterior direction (Talavage et al., 2000). The systematic relationship between the carrier frequency and the

Acknowledgements

This work was supported by the Deutsche Forschungsgemeinschaft (DFG; EL 101/20).

References (36)

  • T.M. Talavage et al.

    Frequency-dependent responses exhibited by multiple regions in human auditory cortex

    Hear Res

    (2000)
  • I.M. Tarkka et al.

    Electric source localization of the auditory P300 agrees with magnetic source localization

    Electroencephalogr Clin Neurophysiol

    (1995)
  • C. Verkindt et al.

    Tonotopic organization of the human auditory cortex: N100 topography and multiple dipole model analysis

    Electroencephalogr Clin Neurophysiol

    (1995)
  • G. Ehret

    The auditory cortex

    J Comp Physiol

    (1997)
  • C. Elberling et al.

    Auditory magnetic fields: source location and ‘tonotopical organization’ in the right hemisphere of the human brain

    Scand Audiol

    (1982)
  • T. Elbert et al.

    Expansion of the tonotopic area in the cortex of the blind

    J Neurosci

    (2002)
  • R. Galambos et al.

    A 40-Hz auditory potential recorded from the human scalp

    Proc Natl Acad Sci

    (1981)
  • C. Gallen et al.

    Reliability and validity of auditory neuromagnetic source localization using a large array biomagnetometer

  • Cited by (15)

    • Frequency characteristics of neuromagnetic auditory steady-state responses to sinusoidally amplitude-modulated sweep tones

      2016, Clinical Neurophysiology
      Citation Excerpt :

      The tonotopy of the ASSR component is mapped in an area of approximately 10–20 mm3 (Romani et al., 1982), with a localization shift of approximately 10 mm in each of the xyz-coordinates, along with the periodic amplitude modulation (AM) at fm (Pantev et al., 1996). Multiple sources for the ASSR generators (Gutschalk et al., 1999; Weisz et al., 2004a) and multiple tonotopic organizations within the auditory cortex (Talavage et al., 2004) have been reported. Together with the large individual deviation among the small number of subjects (n = 8, SD = 4.3–36.8 mm per fc per hemisphere), the non-static aspect of the ASSR generator and the oversimplification of the actual source configuration by the single ECD model may have obscured the results.

    • Phase coherence of auditory steady-state response reflects the amount of cognitive workload in a modified N-back task

      2015, Neuroscience Research
      Citation Excerpt :

      The results of this study clearly showed that the modulation of workload is reflected in the PLI of the ASSR gradually, suggesting the possibility that the PLI of the ASSR could be an index of the amount of workload in the brain. Previous MEG and MRI studies reported that ASSR is evoked by auditory stimuli, and their source has been mainly localized in the primary auditory cortex (Engelien et al., 2000; Gutschalk et al., 1999; Herdman et al., 2002; Kuriki et al., 1995; Pantev et al., 1996; Weisz et al., 2004). In this study, we used MEG for investigating the relationship between the cognitive workload and ASSR in detail.

    View all citing articles on Scopus
    View full text