Superior temporal and inferior frontal cortices are activated by infrequent sound duration decrements: An fMRI study Teemu Rinne, a,b,c,d,e, T Alexander Degerman, a and Kimmo Alho a a Department of Psychology, University of Helsinki, Finland b BioMag Laboratory, Medical Engineering Centre, Helsinki University Central Hospital, Finland c Cognitive Brain Research Unit, Department of Psychology, University of Helsinki, Finland d Helsinki Imaging Center, University of Helsinki, Finland e Human Cognitive Neurophysiology Laboratory, UC Davis and VANCHCS, Martinez, CA 94553-4612, USA Received 25 June 2004; revised 1 December 2004; accepted 14 January 2005 Available online 3 March 2005 Functional magnetic resonance imaging (fMRI) was used to examine the processing of infrequent changes occurring in an unattended sound sequence. In event-related brain potentials (ERPs), such sound changes typically elicit several responses, including an enhanced N1, the mismatch negativity (MMN), and the P3a. In the present study, subjects were presented with a repeating sound of 75 ms in duration, which was occasionally replaced, in separate blocks, by a 15-ms, 25-ms, or 35-ms sound (large, medium, and small change, respectively). In the baseline block, only the frequent 75-ms sound was presented. During the scanning, the subjects were instructed to ignore the sounds while watching a silent wildlife documentary. We assumed that in this condition, the MMN mechanism would contribute more to the observed activation than the other change-related processes. We expected sound changes to elicit fMRI activation bilaterally in the supratemporal cortices, where the electric MMN is mainly generated, and that the magnitude of this activation would increase with the magnitude of sound duration change. Unexpectedly, however, we found that only blocks with medium duration changes (25 ms) showed significant activation in the supratemporal cortex. In addition, as reported in some previous EEG and fMRI studies, contrasts between different levels of sound duration change revealed additional activation in the inferior frontal cortex bilaterally. This activation tended to be greater for the small and medium changes than for the large ones. D 2005 Elsevier Inc. All rights reserved. Keywords: fMRI; Superior temporal cortex; Inferior frontal cortex; Auditory change detection; MMN Introduction The present study used functional magnetic resonance imaging (fMRI) to examine processing of infrequent sound changes occurring in a repetitive train of unattended sounds. Most of the previous knowledge about brain mechanisms involved in process- ing of such sound changes is based on electric and magnetic brain responses recorded with electroencephalography (EEG) and magnetoencephalography (MEG), respectively. This research has largely focused on certain components of the event-related brain potential (ERP) such as the N1, mismatch negativity (MMN), and P3a. The N1 (peak latency about 100 ms from sound onset) is generated by a fast change in acoustic energy (e.g., sound onset) and its amplitude is determined by the physical properties of the sound. A large N1 is typically elicited by sounds presented at a slow rate or in the beginning of a stimulus train (Hari et al., 1982; Na ¨a ¨ta ¨nen and Picton, 1987). In contrast, sounds presented at a fast presentation rate will typically elicit only a small N1. However, if the pitch of a sound in a fast-rate repetitive sequence is occasion- ally changed significantly (e.g., from 1000 Hz to 2000 Hz), then a slightly enhanced N1 is elicited by these infrequent sound changes (Scherg et al., 1989). It is assumed that this enhancement is due to activation of new non-refractory neural populations. In addition to an enhanced N1, infrequent pitch changes also elicit a subsequent MMN response. The MMN is generated by a mechanism that detects deviations from regular aspects of the ongoing auditory stimulation by comparing the incoming sensory information to a representation formed by the repetitive features in preceding auditory inputs (Na ¨a ¨ta ¨nen and Winkler, 1999). Such a regular feature could be, for example, a repeating single sound, a repeating tone pattern, or even an invariant higher level relationship between sounds (Na ¨a ¨ta ¨nen et al., 2001). Thus, it is assumed that N1 and MMN reflect the activation of fundamentally different processes: while the N1 is generated by a fast change in the stimulus energy level (stimulus onset) and its amplitude is determined by the physical properties of the sound, the MMN is elicited by changes 1053-8119/$ - see front matter D 2005 Elsevier Inc. All rights reserved. doi:10.1016/j.neuroimage.2005.01.017 T Corresponding author. Human Cognitive Neurophysiology Laboratory, UC Davis and VANCHCS, Martinez, Research Building R4, 151, VA Medical Center, 150 Muir Road, Martinez, CA 94553-4612, USA. E-mail address: teemu.rinne@helsinki.fi (T. Rinne). Available online on ScienceDirect (www.sciencedirect.com). www.elsevier.com/locate/ynimg NeuroImage 26 (2005) 66 – 72