A single link to the first track to allow the export script to build the search page
  • Addiction, Drugs
  • Information from Lay-Language Summaries is Embargoed Until the Conclusion of the Scientific Presentation

    410—Computation, Modeling, and Simulation V

    Monday, November 11, 2013, 1:00 pm - 4:15 pm

    410.10: Low-frequency intersubject synchronization of MEG time-courses during movie viewing

    Location: 7B

    ">*K. LANKINEN, M. KOSKINEN, J. SAARI, R. HARI;
    Sch. of Sci., Aalto Univ., AALTO, Finland

    Abstract Body: Watching a movie triggers various perceptual, cognitive and emotional processes in the spectator’s brain. Despite the apparent complexity of the stimulus, the time-course of the associated brain activity can, in certain brain areas, be remarkably similar across subjects, as has been demonstrated with functional magnetic resonance imaging (fMRI). These similarities likely reflect time-locking to both the physical features and the contents of the movie. Here we used magnetoencephalography (MEG) to search for fluctuations that would be consistent across subjects at 0.03-1 Hz, corresponding to the frequency range of fMRI. Eight subjects watched a silent 15-min black-and-white movie twice. We trained a spatial filter to characterize unaveraged MEG traces. The filter parameters were estimated by using multi-set canonical correlation analysis (M-CCA) that maximizes the intersubject correlation and provides mutually orthogonal time-courses for each subject. Applying the spatial filter to an independent dataset, statistically significant inter-subject correlations were obtained, reaching average values up to 0.33, 0.25, 0.21, 0.25 and 0.24, in the frequency bands of 0.03-0.2 Hz, 0.2-0.4 Hz, 0.4-0.6 Hz, 0.6-0.8 Hz and 0.8-1 Hz, respectively. To assess the anatomical origins of the signals, the obtained time-courses were correlated with the minimum-norm source current estimates (MNE) projected to the cortex. In all the frequency bands, the most prominent sources were in the visual (V1/V2) cortices. In addition, prefrontal, posterior parietal, temporal, premotor and lateral occipital (V5) areas showed coherent activity across subjects in some frequency bands. These results demonstrate that MEG recordings can unravel neurophysiologically relevant activity even during complex movie stimuli.

    Lay Language Summary: We studied brain activity of healthy adults watching a movie. Despite high complexity of the stimulus, we were able to find similar neural activation patterns in certain brain areas across the viewers. The obtained time courses of the signals changed similarly across the viewers while they were watching the same film. The key factor for success was the advanced signal processing and modeling methodology applied to highly sensitive non-invasive measurement technology.
    Movies are becoming an important tool for neuroscience research. In contrast to typically used highly controlled or artificial visual stimuli, movies provide a relatively naturalistic means to study brain processes that happen during the observation of our environment. By using more complex and realistic stimuli such as movies, it is possible to obtain completely new neurophysiological information of the brain processes of everyday life, e.g. how similarly people actually see the world. Comparison of the brain signals recorded from different viewers during the movie can indicate if their brain activation is changing in the same way when observing the stimulus, and the level of correlation of the signals may tell how similarly they percept the contents of the movie. However, we need to apply proper methods to see how the brain extracts information from the highly complex stimuli.
    Previously, brain activity triggered by movie viewing has been studied with functional magnetic resonance imaging (fMRI) that reflects the changes of oxygen consumption of the brain and is an indirect way to record the brain activity. Here we used magnetoencephalography (MEG) that measures magnetic fields directly related to the activation of the nerve cell populations. However, this is the first extensive MEG study of movie viewing. Therefore, our work develops methodological grounds for detecting physiological signals in noisy MEG recordings.
    MEG signals were measured from eight participants while they were watching a silent 15-min black-and-white movie “At Land” by Maya Deren twice. To find the most correlating time courses of the brain signals between participants, we applied a mathematical signal processing model to optimally combine signals from 204 MEG sensors. We focused on rather low MEG frequencies from 0.03 to 1 Hz, and we further split the frequency range into 0.2 Hz sub-bands to obtain more detailed information.
    Most consistent brain activation across subjects was found in several visual brain areas. Activation was also found in various other areas reflecting different aspects of movie watching. For example, posterior parietal cortex, prefrontal cortex and areas sensitive to observation of motion or bodies are usually considered to be related to attention, cognitive processing and observation of motion and actions of other people, respectively.
    Our results open new possibilities for neuroscientific experimentation with movies or other video stimuli and thus for deeper understanding of brain processes in naturalistic everyday environments. In future, this kind of studies might help to understand how similarly people observe the world and how similar their brain processes actually are. Thus, building the connection between the measurements and brain processes could lead to understand the ongoing brain processes of the spectators just by inspecting the measured signals.

    Information from Lay-Language Summaries is Embargoed Until the Conclusion of the Scientific Presentation

    410—Computation, Modeling, and Simulation V

    Monday, November 11, 2013, 1:00 pm - 4:15 pm

    410.10: Low-frequency intersubject synchronization of MEG time-courses during movie viewing

    Location: 7B

    ">*K. LANKINEN, M. KOSKINEN, J. SAARI, R. HARI;
    Sch. of Sci., Aalto Univ., AALTO, Finland

    Abstract Body: Watching a movie triggers various perceptual, cognitive and emotional processes in the spectator’s brain. Despite the apparent complexity of the stimulus, the time-course of the associated brain activity can, in certain brain areas, be remarkably similar across subjects, as has been demonstrated with functional magnetic resonance imaging (fMRI). These similarities likely reflect time-locking to both the physical features and the contents of the movie. Here we used magnetoencephalography (MEG) to search for fluctuations that would be consistent across subjects at 0.03-1 Hz, corresponding to the frequency range of fMRI. Eight subjects watched a silent 15-min black-and-white movie twice. We trained a spatial filter to characterize unaveraged MEG traces. The filter parameters were estimated by using multi-set canonical correlation analysis (M-CCA) that maximizes the intersubject correlation and provides mutually orthogonal time-courses for each subject. Applying the spatial filter to an independent dataset, statistically significant inter-subject correlations were obtained, reaching average values up to 0.33, 0.25, 0.21, 0.25 and 0.24, in the frequency bands of 0.03-0.2 Hz, 0.2-0.4 Hz, 0.4-0.6 Hz, 0.6-0.8 Hz and 0.8-1 Hz, respectively. To assess the anatomical origins of the signals, the obtained time-courses were correlated with the minimum-norm source current estimates (MNE) projected to the cortex. In all the frequency bands, the most prominent sources were in the visual (V1/V2) cortices. In addition, prefrontal, posterior parietal, temporal, premotor and lateral occipital (V5) areas showed coherent activity across subjects in some frequency bands. These results demonstrate that MEG recordings can unravel neurophysiologically relevant activity even during complex movie stimuli.

    Lay Language Summary: We studied brain activity of healthy adults watching a movie. Despite high complexity of the stimulus, we were able to find similar neural activation patterns in certain brain areas across the viewers. The obtained time courses of the signals changed similarly across the viewers while they were watching the same film. The key factor for success was the advanced signal processing and modeling methodology applied to highly sensitive non-invasive measurement technology.
    Movies are becoming an important tool for neuroscience research. In contrast to typically used highly controlled or artificial visual stimuli, movies provide a relatively naturalistic means to study brain processes that happen during the observation of our environment. By using more complex and realistic stimuli such as movies, it is possible to obtain completely new neurophysiological information of the brain processes of everyday life, e.g. how similarly people actually see the world. Comparison of the brain signals recorded from different viewers during the movie can indicate if their brain activation is changing in the same way when observing the stimulus, and the level of correlation of the signals may tell how similarly they percept the contents of the movie. However, we need to apply proper methods to see how the brain extracts information from the highly complex stimuli.
    Previously, brain activity triggered by movie viewing has been studied with functional magnetic resonance imaging (fMRI) that reflects the changes of oxygen consumption of the brain and is an indirect way to record the brain activity. Here we used magnetoencephalography (MEG) that measures magnetic fields directly related to the activation of the nerve cell populations. However, this is the first extensive MEG study of movie viewing. Therefore, our work develops methodological grounds for detecting physiological signals in noisy MEG recordings.
    MEG signals were measured from eight participants while they were watching a silent 15-min black-and-white movie “At Land” by Maya Deren twice. To find the most correlating time courses of the brain signals between participants, we applied a mathematical signal processing model to optimally combine signals from 204 MEG sensors. We focused on rather low MEG frequencies from 0.03 to 1 Hz, and we further split the frequency range into 0.2 Hz sub-bands to obtain more detailed information.
    Most consistent brain activation across subjects was found in several visual brain areas. Activation was also found in various other areas reflecting different aspects of movie watching. For example, posterior parietal cortex, prefrontal cortex and areas sensitive to observation of motion or bodies are usually considered to be related to attention, cognitive processing and observation of motion and actions of other people, respectively.
    Our results open new possibilities for neuroscientific experimentation with movies or other video stimuli and thus for deeper understanding of brain processes in naturalistic everyday environments. In future, this kind of studies might help to understand how similarly people observe the world and how similar their brain processes actually are. Thus, building the connection between the measurements and brain processes could lead to understand the ongoing brain processes of the spectators just by inspecting the measured signals.