74 nature neuroscience • volume 1 no 1 • may 1998 article The perception of movement in auditory space by humans depends on a number of cues. The movement of sounds in space produces changes in the sound considered at each ear alone 1 and changes in the relative phase and amplitude of the sound’s spectral components between the two ears 2 . For broad spectrum sounds, the spatial fil- tering properties of the pinnae 3 provide another cue for the per- ception of spatial sound properties and are responsible for the external quality of the perception. The neural basis for the perception of sound movement is not firmly established by animal work. The detection of differences in sound phase and amplitude at the two ears depends on the con- vergence of the inputs to each ear. This convergence first occurs in nuclei of the superior olive in the brainstem 4,5 . Neurons sensitive to changing sound phase or amplitude cues at the two ears have been described in the inferior colliculus 6 and primary auditory cortex 7 . The detection of cues based on the spectral shape of sound may depend on a different brainstem pathway. Neurons in the auditory cortex have also been shown to be selectively activated by the actual movement of sounds in space 8 . Animal lesions that involve the primary auditory cortex produce a deficit in the per- formance of tasks based on the detection of dynamic phase differ- ences between the ears 9 . The auditory pathway described in animals from the cochlea to the auditory cortex is therefore necessary for the perception of sound movement. Human lesion work also suggests that the convergence of the inputs to the two ears in the brainstem is important for the detection of interaural cues that produce a perception of sound movement 10 . However, another human lesion study suggests that sound movement perception depends on processing beyond the auditory cortex in the right hemi- sphere 11 , and that the pathway up to the auditory cortex is therefore not sufficient to underlie the perceptual process. This study specifically addresses the question of which brain areas are responsible for the conscious perception of sound movement in humans, as opposed to the simple detection of the underlying cues. The study’s hypothesis was that sound movement perception depends on processing in cortical areas beyond the primary audi- tory cortex. We have used novel stimuli that contain identical changes in the phase and amplitude at the two ears but produce different perceptions of sound movement. These stimuli allow us to identify brain areas where activity increases with the perception of sound movement. That activity has been measured in two separate exper- iments by BOLD (blood oxygen level-dependent) fMRI response and by regional cerebral blood flow as measured by PET. Result s PSYCHOPHYSICAL DATA All subjects underwent psychophysical testing using a novel 500-Hz stimulus, presented over one second, containing a linear change in both the phase and the amplitude difference between the ears. Con- sidered alone, both the phase and the amplitude change may be per- ceived as sound movement toward one side. When the directions of the perceived movements generated by these two cues are opposite to one another, cancellation or trading can occur; the stimulus is a dynamic extension for those used in static time-intensity trading experiments 12 . For all subjects, a point could be defined at which there was exact cancellation of the movement percept due to the two cues and a single static sound image was perceived in the midline (Fig. 1). The point of cancellation for the phase and amplitude ramps varied between subjects (Table 1). The mean amplitude/phase ratio at the point of trading was 0.86 (standard deviation 0.34). During the PET and fMRI experiments, subjects were present- ed with two sound conditions (Fig. 2). The cancel condition repre- sents the dynamic time-intensity-traded stimulus discussed above. Right parietal cortex is involved in the perception of sound movement in humans Timothy D. Griffiths 1,2,3 , Geraint Rees 1 , Adrian Rees 2 , Gary G.R. Green 2 , Caroline Witton 2 , Dominic Rowe 4 , Christian Büchel 1 , Robert Turner 1 and Richard S.J. Frackowiak 1 1 Wellcome Department of Cognitive Neurology, Institute of Neurology, 12 Queen Square, London, WC1N 3BG, UK 2 Department of Physiological Sciences and 3 Department of Clinical Neuroscience, Newcastle University Medical School, Newcastle upon Tyne, NE2 4HH,UK 4 National Hospital for Neurology, Queen Square, London, WC1N 3BG, UK Correspondence should be addressed to T.D.G. (t.griffiths@fil.ion.ucl.ac.uk) Changes in the delay (phase) and amplitude of sound at the ears are cues for the analysis of sound movement. The detection of these cues depends on the convergence of the inputs to each ear, a process that first occurs in the brainstem. The conscious perception of these cues is likely to involve higher centers. Using novel stimuli that produce different perceptions of movement in the presence of identical phase and amplitude modulation components, we have demonstrated human brain areas that are active specifically during the perception of sound movement. Both functional magnetic resonance imaging (fM RI) and positron emission tomography (PET) demonstrated the involvement of the right parietal cortex in sound movement perception with these stimuli.