Magnetoencephalographic study of occipitotemporal activity elicited by viewing mouth movements

Abstract

Objective

We studied the temporal and spatial characteristics of neural responses elicited by viewing mouth movements using magnetoencephalography.

Methods

We focused on differences in responses to mouth opening and closing movements by apparent motion, using an averting eyes condition as a control.

Results

A large clear MEG component, 1 M (mean peak latency of approximately 160 ms), was elicited by both mouth movements. We modeled the neural sources using a brain electric source analysis (BESA) method and placed the sources around: (1) the occipitotemporal border at human MT/V5, (2) the primary visual cortex (V1), and (3) fusiform gyrus. The calculated activity of Source (1) was large whereas the activity of the others was small or negligible. Source (1), as calculated separately for mouth closing and opening movements and eye movement, showed no significant different amplitude and locations. We did not find any activity in the superior temporal sulcus (STS).

Conclusions

Our results indicate that human MT/V5 is active in the perception of both mouth and eye motions. Viewing mouth and eye movements elicits no significant differences in MT/V5 activity, indicating that the perception of movement of facial parts is probably processed in the same manner.

Significance

Characteristic activities in the human MT/V5 elicited by viewing mouth movement were clarified by MEG.

Introduction

Recent neuroimaging studies examining brain responses for viewing the actions of others indicate that in addition to such regions as the superior temporal sulcus (STS) and middle temporal gyrus (MTG), MT/V5 also plays a prominent role in processing these complex motion stimuli (Bonda et al., 1996, Kourtzi and Kanwisher, 2000, Puce et al., 1998, Watanabe et al., 2001). Particularly intriguing is the finding that MT/V5 is active when static images depicting implied motion are presented (Kourtzi and Kanwisher, 2000). The suggestion that MT/V5 exhibits specialization, in addition to general motion processing, is not new (Tootell et al., 1995b, Watanabe et al., 2001), however, the exact nature of this specialization is not known. Previously, fMRI activity on observing eye movements over and above that seen in response to motion in general has been reported in MT/V5 (Puce et al., 1998), leading to speculation that specialized cortical regions for the movement of facial parts might be present in MT/V5 (Watanabe et al., 2001). This activity occurs at around 200 ms post-motion onset, as indicated by both event-related potentials (ERP) (Puce et al., 2000, Puce et al., 2003) and magnetoencephalography (MEG) studies (Watanabe et al., 2001).

The idea that there may be multiple motion sensitive regions in the human brain that deal with the processing of facial movements makes sense given that humans are social primates and routinely ‘read’ facial expressions and movements of others. Additionally, it is known that information relating to faces is processed by both ventral and lateral temporal regions (Allison et al., 1999, McCarthy et al., 1999, Puce and Allison, 1999, Puce et al., 1995, Watanabe et al., 2003).

Previous MEG studies including our own demonstrate that reliable responses from the ventral and lateral temporal cortices can be elicited by static faces (Halgren et al., 2000, Ioannides et al., 2000, Linkenkaer-Hansen et al., 1998, Lu et al., 1991, Nakamura et al., 2001, Sams et al., 1997, Sato et al., 1999, Swithenby et al., 1998, Taylor et al., 2001, Terasaki and Okazaki, 2002, Watanabe et al., 1999a, Watanabe et al., 1999b, Watanabe et al., 2003). Furthermore, we have recorded robust MEG responses to viewing facial eye movements (Watanabe et al., 2001) at around 170 ms post-motion onset (1 M). Although the equivalent current dipole (ECD) of 1 M was located around MT/V5, consistent with previous MEG data on general motion perception (e.g. Bundo et al., 2000, Kawakami et al., 2002), the ECD to eye movements was more inferior and posterior than that to motion in general (Watanabe et al., 2001).

In this study, we studied both temporal and spatial characteristics of MEG responses elicited by viewing mouth movements (opening and closing), as compared to viewing control movement types such as an eye aversion movement and motion in general. Puce et al., 2000, Puce et al., 2003, Wheaton et al., 2001 have reported that ERPs for mouth opening and closing occur at around 170 ms post-motion onset (N170), and that N170 s for mouth opening are in general larger than those seen for mouth closing. The N170 activity occurred over the bilateral posterior temporal scalp, and could conceivably have been generated by neural sources in MT/V5 and in the STS. In these studies the neural sources of the ERP activity were not located. In fMRI and EEG studies, either temporal or spatial characteristics were mainly studied. So, in the present MEG study, we aimed both to localize the sources of neural activity elicited by two types of mouth movements and to show the time course of the sources compared to our control conditions. Therefore, in a previous study (Watanabe et al., 2001), we used apparent motion, which is perceived by the same mechanism as real motion (e.g. Kaneoke et al., 1997).