Evidence for the role of the right auditory cortex in fine pitch resolution

Abstract

The neural basis of human pitch perception is not fully understood. It has been argued that the auditory cortices in the two hemispheres are specialized, such that certain right auditory cortical regions have a relatively finer resolution in the frequency domain than homologous regions in the left auditory cortex, but this concept has not been tested directly. Here, we used functional magnetic resonance imaging (fMRI) to test this specific prediction. Healthy volunteers were scanned while passively listening to pure-tone melodic-like sequences in which the pitch distance between consecutive tones was varied in a parametric fashion. As predicted, brain activation in a region of right lateral auditory cortex, corresponding to the planum temporale, was linearly responsive to increasing pitch distance, even across the fine changes in pitch. In contrast, the BOLD signal at the homologous left cortical region was relatively constant as a function of pitch distance, except at the largest pitch change. The results support the model of relative hemispheric specialization and indicate that the right secondary auditory cortex has a finer pitch resolution than the left.

Introduction

Pitch is the perceptual correlate of acoustic frequency and can be considered along at least two perceptual dimensions, pitch height and pitch chroma (Shepard, 1982). Pitch height, is related to spectral energy distribution and is illustrated by the octave on the keyboard. In contrast, pitch chroma, or the cycle of notes within the octave, provides a basis for acoustic patterns (melodies). Several studies have investigated the processing of sequential or melodic pitch, which is critical for musical perception, however, its neural correlates are not fully understood.

Many findings have shown that musical pitch processing preferentially involves right auditory cortical structures. For example, studies of brain-lesioned patients have shown that the right auditory cortex is critical for melody discrimination (Milner, 1962), perception of missing fundamental pitch (Zatorre, 1988), perception of melody in terms of its global contour (Peretz, 1990), direction of pitch change (Johnsrude, Penhune, & Zatorre, 2000), and in using melodic contextual cues in pitch judgments (Warrier & Zatorre, 2004). Consistent evidence comes from neuroimaging studies of normal subjects, showing that right secondary auditory regions are central in various aspects of musical pitch processing, such as in melodic processing (Patterson, Uppenkamp, Johnsrude, & Griffiths, 2002; Zatorre, Evans, & Meyer, 1994), in the maintenance of pitch while singing (Perry et al., 1999), and in imagery for tunes (Halpern & Zatorre, 1999). In contrast, left auditory regions seem to be specialized for rapid temporal processing as required in speech (e.g. Belin et al., 1998; Jancke, Wustenberg, Scheich, & Heinze, 2002; Phillips & Farmer, 1990; Zaehle, Wustenberg, Meyer, & Jancke, 2004).

Zatorre, Belin, and Penhune (2002) have recently proposed that the auditory system has developed two parallel and complementary systems, one in each hemisphere, specialized for differential resolution in the spectral and temporal domains, as a need to optimally process incoming simultaneous spectral and temporal acoustic information from the environment. A similar proposition has been made by Poeppel (2003), who suggested that different time integration windows characterize the left and right auditory cortices. Support for this model of hemispheric asymmetry comes from neuropsychological (Robin, Tranel, & Damasio, 1990), electrophysiological (Liegeois-Chauvel, Giraud, Badier, Marquis, & Chauvel, 2001) and neuroimaging studies (Boemio, Fromm, Braun, & Poeppel, 2005; Brechmann & Scheich, 2005; Jamison, Watkins, Bishop, & Matthews, 2006; Schönwiesner, Rubsamen, & von Cramon, 2005; Zatorre & Belin, 2001). The present investigation was based on a study by Zatorre and Belin (2001), in which normal subjects were scanned while passively listening to pure-tone sequences that varied parametrically either in terms of spectral complexity or temporal rate. Pure-tones (or sine wave tones) are tones with a single frequency, in contrast to complex tones that are made up of multiple frequencies. Responses to the spectral features were weighted towards right auditory areas, whereas responses to the temporal features were weighted towards the left. These findings were interpreted as reflecting that the right auditory cortex has a finer spectral resolution, whereas the left has a higher temporal resolution, but this explanation was not tested directly, and it has not been universally accepted (Scott & Wise, 2004).

Here, we used functional magnetic resonance imaging (fMRI) to test the prediction arising from the hemispheric specialization hypothesis that the right auditory cortex is more sensitive to small spectral changes relative to the left. Subjects were scanned while passively listening to pure-tone melodic-like sequences in which the pitch distance between consecutive tones was varied in a parametric fashion. We reasoned that if the right auditory cortex (particularly secondary auditory regions) has a higher resolution for spectral change, then it should show correlated changes in the BOLD response as a function of even small changes in the pitch of a pure-tone sequence; the left auditory cortex, should show increased BOLD response only to large changes in pitch. In other words, a high resolution for spectral change (as in the right auditory cortex) should result in a high sensitivity to small spectral changes, whereas a coarser spectral resolution (as in the left auditory cortex) should result in a lower sensitivity when stimulus frequency changes are small. This hypothesis is based on the assumption that there exist neural populations with frequency selectivity. We expand on this theory in Section 4.