By Yumie Ono

What is “hyperscanning”?
Recent advances in non-invasive brain activity measurement techniques have inspired social neuroscientists to simultaneously record data from two or more brains and investigate interpersonal (across-brains) neural correlates in various social situations. Such attempts began more than 50 years ago with electroencephalography (EEG) (Duane and Behrendt, 1965), and scanning two people at once was later termed a “hyperscan” in the work of Montague et al. (2002). Montague and colleagues connected two functional magnetic resonance imaging (fMRI) scanners together, and hemodynamic data from two participants were measured while they played a deception game.

Advantage of fNIRS hyperscanning for social neuroscience
As interest in social neuroscience research has grown, hyperscanning has emerged as a feasible method for investigating interactions between brains as unique functional units. Although more than 1,600 articles have been published so far with EEG/fMRI hyperscanning (results of PubMed and Google Scholar searches on 23 January 2018), these modalities technically restrict the natural behavior that occurs during person-to-person social interaction. For example, eye contact, gaze control, conversation, and muscle movements for facial expressions are core components of inter-personal communication; however, these behaviors often generate severe artifacts in EEG and fMRI data. By virtue of tolerance to limited body movement, fNIRS hyperscanning allows researchers to investigate live interpersonal interaction with some degree of the behaviors mentioned above. In addition, fNIRS does not require participants to be confined to a scanner, which enables neuroimaging data collection during interpersonal interactions in more ecologically valid environments (Figure). Participants can interact face-to-face (Funane et al., 2011; Holper et al., 2012; Osaka et al., 2015; Jiang et al., 2012, 2015; Liu et al., 2016; Nozawa et al., 2016; Hirsch et al., 2017; Piva et al., 2017) or side-by-side (Cui et al., 2012; Balconi and Vanutelli 2017) using naturalistic communication including conversation (Jiang et al., 2012, 2015), eye gaze (Hirsch et al., 2017), and even humming or singing (Osaka et al., 2015) while their brain activity is being recorded.

Unlike fMRI, fNIRS allows researchers to acquire regional hemodynamic activity only from the cortical surface (Koike et al., 2015). Fortunately, the brain regions and neural systems important for social neuroscience are located in fNIRS-accessible areas, such as the canonical language systems (Broca’s and Wernicke’s Areas), temporal-parietal junction (TPJ), frontal pole, and dorsolateral prefrontal areas. Therefore, fNIRS is a suitable hyperscanning modality for investigating live brain-to-brain social interactions. Another advantage of fNIRS is an open experimental environment that allows other biological measurement devices, such as eye trackers, electrocardiogram and respiratory sensors, and video/audio recorders, to be easily added depending on the specific scientific questions.

Measurement and analysis of fNIRS cross-brain measures
Most fNIRS hyperscanning studies use a single fNIRS system and divide its optodes between two participants (Figure). This arrangement has several benefits. First, it is free from the synchronization problems of data acquisition timing, as two participants share the same system. Second, it is cost-effective compared to purchasing two independent brain imaging modalities and connecting them via a local area network.

The acquired data have the advantage of cross-brain interrogation to determine the inter-brain connectivity (Felix et al., 2013; Babiloni and Astolfi, 2014). The most frequently used method is wavelet transform coherence (WTC), which was first introduced by Cui et al. (2012) in their cooperative and competitive button-press task in pairs of participants. The amplitude of the complex coherence value is calculated between corresponding fNIRS channels from two participants as an index of interpersonal neural synchronization (INS). The increased coherence in the right superior frontal cortices during a cooperative task but not in a competitive task was associated with better cooperation performance (Cui et al., 2012). Follow-up research using the same paradigm further assessed the difference in cooperation performance and corresponding changes in INS to gender (Baker et al., 2016) and strength of social bonding (Pan et al., 2017) of pairs.

Granger causality (Seth et al., 2015) is another connectivity analysis that focuses on the directionality of information flow. Its application to hyperscanning fNIRS signals have revealed leader-follower type effective connectivity from models to imitators during motor imitation between the premotor cortices (Holper et al., 2012), from leaders to followers during debate between TPJs (Jiang et al., 2015), from female to male partners during cooperation between the right superior frontal cortices (Pan et al., 2017), and from “bankers to followers” during deception in a gambling game between the superior temporal sulci (Zhang et al., 2017).

Further challenges for fNIRS hyperscanning research
Currently, most fNIRS hyperscanning experiments use one or two small patches of optodes (from 2 to 10 channels) for every study participant. Considering individual differences in cortical anatomy, such patches may not accurately reflect regional hemodynamic activity unless careful optode localization techniques are performed. In addition, limiting optode coverage to regions that have previously been associated with a particular behavior reduces the possibility of discovering unexpected activity in other brain areas, which may be relevant for future investigations.

Although the optodes of a single fNIRS machine must be distributed over two heads for hyperscanning, it is preferable to record from a larger cortical surface including the regions of interest and the surrounding areas. Approximately 16 pairs of optodes per brain hemisphere could provide sufficient prefrontal-temporal coverage to record hemodynamic data from dorsolateral prefrontal cortex, inferior frontal cortex, Wernicke’s Area, and the anterior TPJ (Pan et al., 2017; Hirsch et al., 2017; Piva et al., 2017).

Most of the published fNIRS research performed channel-wise cross-brain connectivity analysis, in which the connectivity measures were calculated between corresponding channels (supposed to be the same cortical areas of two brains) of each pair of participants. Since synchronized neuronal activity indeed occurs across different cortical areas of two brains (Fallani et al., 2010), further research should also explore connectivity between different channels or regions of interest across brains (Hirsch et al., 2017; Piva et al., 2017) to reveal the structure of interpersonal functional connectivity networks.

Another challenge in fNIRS research concerns systemic cardiovascular effects as a general confounder of hemodynamic signal data, including in hyperscanning research. Short separation measurements (Gagnon et al., 2014) and/or spatial filtering methods (Zhang et al., 2016) need to be applied to extract the neural components of fNIRS data, since active social communication is likely to alter the attention and arousal level and may cause task-related systemic hemodynamic changes. Overcoming these challenges would further expand the presence of fNIRS in social neuroscience as a unique modality that can measure a close simulation of a real-world inter-personal interaction.

Future directions of fNIRS research in social neuroscience can incorporate naturalistic behavioral components of inter-personal communication such as synchronized (or even contagious) facial and bodily expressions which has been almost prohibited in the existing hyperscanning research using other modalities. Establishing a novel methodology to objectively measure such behaviors and extract the corresponding neural traits would further enable researchers to open the door of new-generation social neuroscience research.

Acknowledgements
The author thanks Prof. Joy Hirsch and the members of Brain Function Laboratory (Yale School of Medicine) for the fruitful discussions.

References
Babiloni, F., & Astolfi, L. (2014). Social neuroscience and hyperscanning techniques: past, present and future. Neuroscience & Biobehavioral Reviews, 44, 76-93.

Balconi, M., & Vanutelli, M. E. (2017). Interbrains cooperation: hyperscanning and self-perception in joint actions. Journal of clinical and experimental neuropsychology, 39(6), 607-620.

Baker, J. M., Liu, N., Cui, X., Vrticka, P., Saggar, M., Hosseini, S. H., & Reiss, A. L. (2016). Sex differences in neural and behavioral signatures of cooperation revealed by fNIRS hyperscanning. Scientific reports, 6, 26492.

Cui, X., Bryant, D. M., & Reiss, A. L. (2012). NIRS-based hyperscanning reveals increased interpersonal coherence in superior frontal cortex during cooperation. Neuroimage, 59(3), 2430-2437.

Duane, A.T.D., Behrendt, T., 1965. Extrasensory electroencephalographic induction between identical twins. Science 150 (3964), 367.

Fallani, F. D. V., Nicosia, V., Sinatra, R., Astolfi, L., Cincotti, F., Mattia, D., Wilke, C., Doud, A., Latora, V., He, B., & Babiloni, F. (2010). Defecting or not defecting: how to “read” human behavior during cooperative games by EEG measurements. PloS one, 5(12), e14187.

Funane, T., Kiguchi, M., Atsumori, H., Sato, H., Kubota, K., & Koizumi, H. (2011). Synchronous activity of two people’s prefrontal cortices during a cooperative task measured by simultaneous near-infrared spectroscopy. Journal of biomedical optics, 16(7), 077011.

Gagnon, L., Yücel, M. A., Boas, D. A., & Cooper, R. J. (2014). Further improvement in reducing superficial contamination in NIRS using double short separation measurements. Neuroimage, 85, 127-135.

Hirsch, J., Zhang, X., Noah, J. A., & Ono, Y. (2017). Frontal temporal and parietal systems synchronize within and across brains during live eye-to-eye contact. NeuroImage. 157, 314-330.

Holper, L., Scholkmann, F., & Wolf, M. (2012). Between-brain connectivity during imitation measured by fNIRS. Neuroimage, 63(1), 212-222.

Jiang, J., Dai, B., Peng, D., Zhu, C., Liu, L., & Lu, C. (2012). Neural synchronization during face-to-face communication. Journal of Neuroscience, 32(45), 16064-16069.

Jiang, J., Chen, C., Dai, B., Shi, G., Ding, G., Liu, L., & Lu, C. (2015). Leader emergence through interpersonal neural synchronization. Proceedings of the National Academy of Sciences, 112(14), 4274-4279.

Koike, T., Tanabe, H. C., & Sadato, N. (2015). Hyperscanning neuroimaging technique to reveal the “two-in-one” system in social interactions. Neuroscience research, 90, 25-32.

Liu, N., Mok, C., Witt, E. E., Pradhan, A. H., Chen, J. E., & Reiss, A. L. (2016). NIRS-based hyperscanning reveals inter-brain neural synchronization during cooperative Jenga game with face-to-face communication. Frontiers in human neuroscience, 10.

Montague, P. R., Berns, G. S., Cohen, J. D., McClure, S. M., Pagnoni, G., Dhamala, M., Wiest M.C., Karpov I., King R.D., Apple N., Fisher R.E. (2002). Hyperscanning: simultaneous fMRI during linked social interactions. NeuroImage 16(4), 1159-1164.

Nozawa, T., Sasaki, Y., Sakaki, K., Yokoyama, R., & Kawashima, R. (2016). Interpersonal frontopolar neural synchronization in group communication: An exploration toward fNIRS hyperscanning of natural interactions. NeuroImage, 133, 484-497.

Osaka, N., Minamoto, T., Yaoi, K., Azuma, M., Shimada, Y. M., & Osaka, M. (2015). How two brains make one synchronized mind in the inferior frontal cortex: fNIRS-based hyperscanning during cooperative singing. Frontiers in psychology, 6.

Pan, Y., Cheng, X., Zhang, Z., Li, X., & Hu, Y. (2017). Cooperation in lovers: An fNIRS‐based hyperscanning study. Human brain mapping, 38(2), 831-841.

Piva, M., Zhang, X., Noah, J. A., Chang, S. W., & Hirsch, J. (2017). Distributed neural activity patterns during human-to-human competition. Frontiers in human neuroscience, 11.

Seth, A. K., Barrett, A. B., & Barnett, L. (2015). Granger causality analysis in neuroscience and neuroimaging. Journal of Neuroscience, 35(8), 3293-3297.

Scholkmann, F., Holper, L., Wolf, U., & Wolf, M. (2013). A new methodical approach in neuroscience: assessing inter-personal brain coupling using functional near-infrared imaging (fNIRI) hyperscanning. Frontiers in human neuroscience, 7.

Zhang, M., Liu, T., Pelowski, M., & Yu, D. (2017). Gender difference in spontaneous deception: A hyperscanning study using functional near-infrared spectroscopy. Scientific reports, 7(1), 7508.

Zhang, X., Noah, J. A., & Hirsch, J. (2016). Separation of the global and local components in functional near-infrared spectroscopy signals using principal component spatial filtering. Neurophotonics, 3(1), 015004.