Hyperscanning

Instead of one person alone in a scanner, we record two people at the same time while they look at each other, talk or play. Then we measure how their brain signals line up across the two heads.

• Frontal temporal and parietal systems synchronize within and across brains during live eye-to-eye contactHirsch, Zhang, Noah & Ono, NeuroImage, 2017
• Real-time eye-to-eye contact is associated with cross-brain neural coupling in angular gyrusNoah et al., Frontiers in Human Neuroscience, 2020

EEG

Electrodes on the scalp pick up the brain’s electrical rhythms millisecond by millisecond. fNIRS tells us where, EEG tells us when, so we often record both at once.

• Co-localization of theta-band activity and hemodynamic responses during face perception: simultaneous EEG and fNIRS recordingsDravida et al., Neurophotonics, 2019
• Multiscale neural signatures of long-lag repetition learning: a simultaneous fNIRS–EEG studyRen et al., Neurophotonics, 2026

Eye & face tracking

Eye trackers record exactly where someone is looking, and face-tracking software reads expressions frame by frame. Together they tell us what each partner is actually seeing and doing during a live interaction.

• Spatiotemporal processing of real faces is modified by visual sensingKelley et al., NeuroImage, 2025
• Neural mechanisms for emotional contagion and spontaneous mimicry of live facial expressionsHirsch, Zhang, Noah & Bhattacharya, Phil. Trans. R. Soc. B, 2023

MRI & fMRI

The MRI scanner gives the sharpest picture of where brain activity happens, but you usually have to lie still inside a tube. We use fMRI as the gold standard to check that fNIRS is measuring the right thing. Lately I’ve been testing a portable low-field MRI scanner that rolls right up to a bed (my own kids were early test subjects), and we want to use it a lot more.

• fMRI validation of fNIRS measurements during a naturalistic taskNoah et al., Journal of Visualized Experiments, 2015
• Activation of dorsolateral prefrontal cortex in a dual neuropsychological screening test: an fMRI approachTachibana et al., Behavioral and Brain Functions, 2012

TMS & TIS

Recording shows which brain areas light up. Stimulation lets us test whether they’re really needed. TMS uses a quick magnetic pulse to briefly nudge a small patch of cortex, and temporal interference stimulation (TIS) overlaps two gentle electrical fields to reach deeper targets without surgery. We pair stimulation with fNIRS while people look at faces.

• Using TMS and fNIRS to disrupt and detect changes in visual processing of faces (Society for fNIRS, 2022)
• Neural specificity for live dyadic face processing: preliminary TMS and fNIRS findings (Society for fNIRS, 2024)

Machine learning

Two people, dozens of channels each, plus eye tracking and video add up to a mountain of data. We train machine-learning models to find the patterns, for example to predict autism symptom scores from brain responses to eye contact.

• Support vector machine prediction of individual ADOS scores based on neural responses during live eye-to-eye contactZhang et al., Scientific Reports, 2024
• Deep multimodal representations and classification of first-episode psychosis via live face processingSingh et al., Frontiers in Psychiatry, 2025

Consumer gadgets in the lab

Before social neuroscience, I put consumer tech to the test in the lab: how well a Fitbit counts steps and whether a Microsoft Kinect can stand in for a pro Vicon motion-capture system.

• Comparison of steps and energy expenditure assessment in adults of Fitbit Tracker and Ultra to the Actical and indirect calorimetryNoah et al., J. Medical Engineering & Technology, 2013
• Comparative abilities of Microsoft Kinect and Vicon 3D motion capture for gait analysisPfister et al., J. Medical Engineering & Technology, 2014

Dance video games

What does your brain do while you play Dance Dance Revolution? We recorded players with fNIRS and found that activity in frontal and temporal areas predicted how accurately they hit the steps. I got to present the work at E3, the biggest video game show there is.

• Frontotemporal oxyhemoglobin dynamics predict performance accuracy of dance simulation gameplayOno et al., NeuroImage, 2014

Minecraft Bedrock Runs Great on My iPhone and iPad. My Mac? Apple Lets Microsoft Say No.

The iPhone Duo’s app growing pains are new. Mac users locked out of iPhone and iPad apps by a single checkbox is not, and the touchscreen MacBook Pro makes it even dumber.

Apple Built a $299 Kitchen Screen. It Can’t Run YouTube or My Recipe App.

Apple’s new HomeView is built for the kitchen, but it has no App Store, no YouTube app, and no way to run a recipe manager like Paprika. Watching YouTube in Safari on a 7-inch square screen is not a plan.

Apple Finally Lets You Watch Video in CarPlay. Too Bad No Car Will Let You.

iOS 26 and iOS 27 brought video to CarPlay while parked. Automakers have to turn it on, and so far none have. Apple, push harder.

I Can Tell Siri to Send a Facebook Messenger Message in CarPlay. It Just Won’t Read Me the Reply.

Messenger syncs across every device I own and lets Siri send messages from my car. But unlike WhatsApp and Apple Messages, it won’t read me the replies.

ChargePoint Has a Great Notification System. It Just Won’t Tell You About Your Money.

A 25-cent service fee, a $50 card hold and a $7 parking fee on $3 of electricity, and the one alert that mattered showed up an hour late.

Everyone Loves the Apple TV. Then a Kid Picks Up the Remote.

Tech media loves the Apple TV, but its touchy Siri Remote is a disaster with kids. I swapped it for a $38 button remote, and at $299 versus $149.99, the Roku Ultra makes a lot more sense, home-screen ads and all.