Behavioral/Systems/CognitiveNeuromagnetic Correlates of Streaming in HumanAuditory CortexAlexander Gutschalk,1,2,3Christophe Micheyl,2Jennifer R. Melcher,3,4,5Andre´ Rupp,1Michael Scherg,1andAndrew J. Oxenham2,41Department of Neurology, University of Heidelberg, 69120 Heidelberg, Germany,2Research Laboratory of Electronics, Massachusetts Institute ofTechnology, Cambridge, Massachusetts 02139,3Eaton–Peabody Laboratory, Massachusetts Eye and Ear Infirmary, Boston, Massachusetts 02114,4Programin Speech and Hearing Bioscience and Technology, Harvard–Massachusetts Institute of Technology Division of Health Sciences and Technology,Cambridge, Massachusetts 02139, and5Department of Otology and Laryngology, Harvard Medical School, Boston, Massachusetts 02115The brain is constantly faced with the challenge of organizing acoustic input from multiple sound sources into meaningful auditoryobjects or perceptual streams. The present study examines the neural bases of auditory stream formation using neuromagnetic andbehavioral measures. The stimuli were sequences of alternating pure tones, which can be perceived as either one or two streams. In thefirst experiment, physical stimulus parameters were varied between values that promoted the perceptual grouping of the tone sequenceinto one coherent stream and values that promoted its segregation into two streams. In the second experiment, an ambiguous tonesequence produced a bistable percept that switched spontaneously between one- and two-stream percepts. The first experiment demon-strated a strong correlation between listeners’ perception and long-latency (⬎60 ms) activity that likely arises in nonprimary auditorycortex. Thesecond demonstrated a covariation between this activity and listeners’ perception in the absence of physical stimulus changes.Overall, the results indicate a tight coupling between auditory cortical activity and streaming perception, suggesting that an explicitrepresentation of auditory streams may be maintained within nonprimary auditory areas.Key words: auditory cortex; magnetoencephalography; scene analysis; stream segregation; bistable percepts; adaptationIntroductionListening to one speaker or following an instrumental melodicline in the presence of competing sounds relies on the perceptualseparation of acoustic sources. As acoustic events unfold overtime, the brain generally succeeds in correctly assigning soundfrom the same source to one “auditory stream” while keepingcompeting sources separate in a process known as “auditorystream segregation” or streaming (Bregman, 1990).The auditory system can segregate sounds into streams basedon various acoustic features (Carlyon, 2004). A compelling dem-onstration (van Noorden, 1975) is provided by a sequence ofrepeating tone triplets separated by a pause (ABA_). When thefrequency difference (⌬F ) between the A and B tones is small andthe interstimulus interval (ISI) is long, the sequence is generallyheard as a single stream with a characteristic galloping rhythm. Incontrast, when the ⌬F is large and the ISI is brief, the sequencesplits into two streams, the galloping rhythm is no longer heard,and instead two regular, or isochronous, rhythms are heard, one(A-tone stream) at twice the rate of the other (B-tone stream). Atintermediate ⌬F and ISI values, the percept depends on the lis-tener’s attentional set as well as the duration of listening to thesequence (Anstis and Saida, 1985; Carlyon et al., 2001), and spon-taneous switches in percept can occur just as they do in vision forambiguous figures or in situations involving binocular rivalry(Blake and Logothetis, 2002).The present study examined the neural basis of streamingusing magnetoencephalography (MEG). Our experiments weremotivated by a hypothesized relationship between streaming andneural adaptation (Fishman et al., 2001) that can be understoodby considering the auditory evoked N1m, a surface negative wavethat occurs in auditory cortex 80 –150 ms after stimulus onset.For a sequence of tones at the same frequency, the N1m evoked byeach successive tone adapts to a steady-state value that dependson the ISI between tones (Ritter et al., 1968; Hari et al., 1982;Imada et al., 1997). For a sequence of tones alternating in fre-quency, the N1m amplitude and its dependence on ISI remainsthe same, as long as the ⌬F between tones is small. However, forlarge ⌬F, the amplitude (and hence the degree of adaptation)becomes consistent with the longer ISI between successive tonesof the same frequency rather than the shorter ISI between tem-porally adjacent tones of different frequency, indicating that ad-aptation occurs selectively on the basis of tone frequency (Butler,1968; Picton et al., 1978; Na¨a¨ta¨nen et al., 1988). This is consistentwith perceptual organization in a streaming task: when all tonesare perceived as one stream, the perceived rate is uniquely deter-mined by the ISI between successive tones (plus the tone dura-Received Jan. 25, 2005; revised April 26, 2005; accepted April 26, 2005.This research was supported primarily by Deutsche Forschungsgemeinschaft Grant GU 593/2-1 (A.G.) as well asby National Institutes of Health Grant R01 DC 05216 (A.J.O.). We thank Josh McDermott and Chris Plack for helpfulcomments.Correspondence should be addressed to Alexander Gutschalk, Eaton–Peabody Laboratory, Massachusetts Eyeand Ear Infirmary, 243 Charles Street, Boston, MA 02114. E-mail: [email protected]:10.1523/JNEUROSCI.0347-05.2005Copyright © 2005 Society for Neuroscience 0270-6474/05/255382-07$15.00/05382 • The Journal of Neuroscience, June 1, 2005 • 25(22):5382–5388tion). At large ⌬F, when the two tone frequencies are segregatedinto separate streams, the perceived rate is lower and depends onthe ISI between tones within each stream. As such, the N1m mightincrease with the perceived ISI, whereas the physical ISI remainsunchanged. Thus, the N1m may provide a physiological indicatorof auditory stream segregation.The present study tested this hypothesis by measuring theauditory evoked field (AEF) in response to triplet sequences thatwere heard as either one or two streams. In one experiment, thepercept was altered by manipulating stimulus parameters. In an-other, stimulus parameters were held constant and were chosenso that the percept was bistable, inducing spontaneous switchesbetween the perception of one and two streams.Materials and MethodsListeners. Fourteen