AIM
Hyperscanning is an emerging topic in neuroscience research. It is being explored in various application fields to gain insight into the neural underpinnings of inter- and intra-brain interactions on the central and peripheral systems. Such studies involve data acquisition from two or more subjects simultaneously, therefore, the experimental setup is relatively complex and mandates the use of multiple devices that are synchronized [1].
In this document, we will describe the procedure of setting up simultaneous EEG recordings using the eego™ system in stationary and mobile contexts. We would also cover some successful use-cases in the field and highlight known challenges in the field of hyperscanning research.
BACKGROUND
The term ‘hyperscanning’ was coined by Montague et al. in their 2002 paper where functional magnetic resonance imaging (fMRI) was acquired from two subjects simultaneously while they played a children’s game called handy-dandy. While player 1 concealed the object, player 2 had to guess which hand had the object [2]. While hyperscanning is a term that came into prominence relatively recently, dyadic observational studies in the animal population have existed for over 40 years. Most published research in this area include keywords such as dyads, mirror neurons, dyadic recordings etc.
Currently, there are quite a few published hyperscanning studies that incorporate neuroimaging techniques such as electroencephalography (EEG), fMRI, Magnetoencephalography (MEG), Transcranial alternating current stimulation (tACS) or Functional near-infrared spectroscopy (fNIRS). Most EEG hyperscanning studies shed light on aspects such as social synchrony, team performance and chemistry, audience-performer interactions and leader-follower dynamics. In most application fields, functional connectivity was investigated per subject and between subjects during their interactions which demanded complex cooperation or competition strategies [3]. In a study published by Nam et al. (2020), EEG and MEG were reported to be the most used neuroimaging methods for hyperscanning studies, and a majority of the research is focused on applications revolving around cognition (48%), decision-making tasks (26%) and motor synchronization (23%). In the same paper, there is also an indication of the timeline in which various hyperscanning studies were done between years 2002-2019. Figure 1 gives readers an impression of the progress in the field of hyperscanning [4]. Hyperscanning setups can help to investigate relationships between task difficulty and team coordination in interactive tasks, look for evidence of dysfunction between age-matched, gender-matched controls and subjects etc.
Figure 1: Timeline of hyperscanning studies conducted with different neuroimaging methods, taken from Nam, Chang S., et al. "Brain-to-brain neural synchrony during social interactions: a systematic review on hyperscanning studies." Applied Sciences 10.19 (2020): 6669 [4].
SYSTEM SETUP IN VARIOUS APPLICATION CONTEXTS:
The eego™ system can be employed in a variety of use-cases in stationary (within a laboratory) or mobile settings. In their 2012 review paper, Babiloni et al. gave two main examples of EEG hyperscanning architectures, wherein data was captured from four subjects and a dyad. From their overview, it became clear that most published studies of the time were conducted within the lab where synchronization and calibration of multiple devices could be controlled. The need to record from additional modalities such as electrooculogram (EOG) and electromyogram (EMG) in order to remove eye blinks and muscle artifacts from the EEG data, was also evident [5]. However, the main challenge is to bring more naturalistic settings to the lab and conduct hyperscanning studies in semi-structured or unstructured setups.
In this section, we will elaborate on how three different kinds of hyperscanning setups can be achieved using the eego™ system.
Application Use-Case 1: 2 separate eego™ systems
To acquire from multiple subjects within a laboratory setting, the stimulus presentation platform should be able to send TTL triggers to two eego™ amplifiers simultaneously. The amplifiers can be connected to two separate recording devices or cascaded. The trigger adapters provided by default with your eego™ system can bring TTL triggers into the amplifier via the dedicated trigger port. In case the two 64 channel amplifiers are integrated into one 128 channel setup, the cascading adapter will have to be used and all the data can be collected on one recording device. Each eego™ amplifier will be connected to a waveguard™ cap. Such a setup might be relevant for studies where tasks entail responding to simultaneous stimulus streams at varying degrees of workload, Sciaraffa et al., (2017) [6].
Figure 2: Schematic of a stationary hyperscanning setup with eegoTM systems.
Application Use-Case 2: An integrated solution or a cascaded setup
In general research settings, two or four cascaded eego™ amplifiers can be employed to achieve a typical hyperscanning setup. Each eego™ amplifier will be connected to a waveguard™ cap. Data is recorded synchronously from both subjects into one recording device. This recording device should receive both stimulus triggers and behavioral response events from the main stimulation PC. Products such as the eego™ hub can also be used here as it is an integrated solution which consists of two or four eego™ amplifiers and can serve as the dedicated acquisition station. Such a setup can be relevant when investigating the impact of social settings upon perception, or for use-cases where behavioral responses need to be co-registered into the EEG recording.
Figure 3: Schematic of a hyperscanning setup in research settings. Application Use-case 3: Mobile settings
In mobile settings, individual eego™ amplifiers should be connected to a waveguard™ cap and recording device. Each system can then be loaded into the eego™ sports backpack to grant the subjects complete mobility. In order to synchronize the EEG traces an external trigger button (with a Y splitter) needs to be connected to both amplifiers. Alternatively, a trigger can be manually sent and simultaneously recorded onto both EEG recordings.
An example of such as setup can be seen in the juggling dyad experiment conducted by Filho et al, where cooperative juggling was explored to understand the neural correlates of shared mental models [1].
Figure 4: Schematic representation of a mobile hyperscanning setup. In each of the above setups, the eego™ web controller can be used to make the start and stop of EEG recordings simpler, remotely, for all subjects and across all devices. Both the device that runs the web controller and the acquisition/recording device that runs eego™ should be connected to the same internet network or dedicated router. Networks events can be used to send continuous synchronization markers to all EEG datasets.
HARDWARE
ACQUISITION SOFTWARE
DEDICATED WORKFLOWSFigure 5: The eegoTM software has dedicated workflows that allows you to choose a recording setup to suit your research needs. In this figure, the workflow is explained in a high-level overview.
The ‘Edit Amplifier’ workflow has default hardware setups that can be edited to accommodate 2 to 4 simultaneous recordings. Using the ‘Edit Montage’ workflow, a corresponding montage can be created for each setup. Our Application Team can be contacted for specific setup related information and consults.
RELEVANT USE-CASES
In this section, we will highlight some EEG hyperscanning research that was made possible with the eegoTM system.
1. The juggling dyad experiment: Filho et al. (2015) and Stone et al. (2019) investigated cooperative juggling as a platform to explore team mental models (TMM) using peripheral and central neurophysiological markers. Two jugglers of different skill levels (>3 years or ≤3 years) were equipped with the eego™ sports 32 channel systems while EEG was recorded into 2 different recording devices. Both amplifiers were connected to a push button which sent a TTL trigger to both amplifiers at the beginning and end of the trial for post-hoc synchronization of the two EEG datasets [1,8].
2. The Neurolive project: This one-of-a-kind research endeavor is being carried out as an interdisciplinary, cross-departmental collaboration at the Goldsmiths, University of London. 21 eego™ 32 channel sports systems are used in parallel to acquire EEG data from 21 subjects in the audience while the performance titled ‘Detective Work’ was ongoing on stage. The eego™ web controller was employed to start and stop recordings across all 21 systems, and Lab Streaming Layer (LSL) was used to send continuous synchronous markers to the EEG data [9].
3. Understanding interpersonal synchronization from dyadic recordings: Rosso et al. recorded two subjects simultaneously during a finger-tapping task in their recently published study. They conclude that periodic beta power modulations could be an underlying mechanism of interpersonal synchrony which enables mutual adaptation and predictions between coupled individuals in dyadic recordings. They employed a cascaded eego™ mylab setup and each subject was connected to a 64 channel waveguardTM original cap. The eego™ amplifier received a TTL trigger from the Teensy microcontroller via a BNC adapter at the start of each recording [10].
In addition to the above, hyperscanning studies are also being introduced in areas such as psychotherapy, neuro-prosthetics, architecture and design. ANT Neuro is a proud collaborator and hopes to support more innovative research in this domain in the future.
CHALLENGES AND DISCUSSION
- Analysis of EEG hyperscanning data is based on approaches that are specific to the setups deployed in each study. The community is yet uncertain on the adoption of one common gold standard for analyses.
- Structure or semi-structured setups allow more control to the researcher as compared to unstructured or naturalistic setups where the subject is free to interact with the environment.
- While more ecologically valid settings are preferred in hyperscanning research, such setups should also include acquisition devices to record modalities in addition to EEG, such as eye tracking, motion capture data, EMG and EOG.
- In order to gain an insight into different analysis methods that can be adopted for hyperscanning data, we would like to refer the reader to some well-cited papers on the same, namely, Burgess, Adrian P. (2013), Babiloni et al. (2012), Czeszumski, Artur, et al. (2020) and Dumas, Guillaume (2011) [7,5, 11, 12].
- Online toolboxes that are centered around hyperscanning analyses and processing pipelines are also made available by the community. Some examples are: DEEP (A dual EEG pipeline for developmental hyperscanning studies) and HyPyp (A Hyperscanning Python Pipeline for inter-brain connectivity analysis) [13,14].
- General testing of the overall system is critical and should be done prior to the studies being carried out. A timing test is always recommended if triggers or software events are being sent to the EEG data from any third-party application.
- A dedicated router with large enough bandwidth is recommended in case software events are being sent to the eego™ acquisition PC from any other PC on the same network.
REFERENCES
1. Filho, Edson, et al. "The juggling paradigm: a novel social neuroscience approach to identify neuropsychophysiological markers of team mental models." Frontiers in psychology 6 (2015): 799.
2. Montague, P. Read, et al. "Hyperscanning: simultaneous fMRI during linked social interactions." Neuroimage 16.4 (2002): 1159-1164.
3. Kelsen, Brent A., et al. "What has social neuroscience learned from hyperscanning studies of spoken communication? A systematic review." Neuroscience & Biobehavioral Reviews (2020).
4. Nam, Chang S., et al. "Brain-to-brain neural synchrony during social interactions: a systematic review on hyperscanning studies." Applied Sciences 10.19 (2020): 6669.
5. Babiloni, Fabio, and Laura Astolfi. "Social neuroscience and hyperscanning techniques: past, present and future." Neuroscience & Biobehavioral Reviews 44 (2014): 76-93.
6. Sciaraffa, Nicolina, et al. "Brain interaction during cooperation: Evaluating local properties of multiple-brain network." Brain sciences 7.7 (2017): 90.
7. Burgess, Adrian P. "On the interpretation of synchronization in EEG hyperscanning studies: a cautionary note." Frontiers in human neuroscience 7 (2013): 881.
8. Stone, David B., et al. "Hyperscanning of interactive juggling: expertise influence on source level functional connectivity." Frontiers in human neuroscience 13 (2019): 321.
9. https://neurolive.info/
10. Rosso, Mattia, et al. "Mutual beta power modulation in dyadic entrainment." NeuroImage 257 (2022): 119326.
11. Czeszumski, Artur, et al. "Hyperscanning: a valid method to study neural inter-brain underpinnings of social interaction." Frontiers in Human Neuroscience 14 (2020): 39.
12. Dumas, Guillaume. "Towards a two-body neuroscience." Communicative & integrative biology 4.3 (2011): 349-352.
13. Kayhan, Ezgi, et al. "DEEP: A dual EEG pipeline for developmental hyperscanning studies." Developmental cognitive neuroscience 54 (2022): 101104.
14. Ayrolles, Anaël, et al. "HyPyP: a Hyperscanning Python Pipeline for inter-brain connectivity analysis." Social Cognitive and Affective Neuroscience 16.1-2 (2021): 72-83.