Kernel Flow 2: Portable TD-fNIRS and EEG Brain Imaging System for Full Cortical Coverage

A next-generation wearable brain imaging system combining time-domain functional near-infrared spectroscopy (TD-fNIRS) with EEG in a dense modular array supporting up to 40 modules with thousands of measurement channels, enabling full cortical coverage for hemodynamic and electrophysiological brain monitoring outside the laboratory.

Presented By Our Partners

Functional brain imaging has historically been confined to clinical and research settings: fMRI requires a multi-ton magnet in a shielded room, PET scanning involves radioactive tracers, and even clinical EEG requires extensive electrode preparation by trained technicians. These constraints have meant that brain imaging, unlike heart rate or glucose monitoring, has remained inaccessible outside specialized facilities. A 2023 review published in Neurophotonics by Yücel et al. examined the potential of wearable fNIRS systems for translational neuroscience, concluding that portable, lightweight fNIRS devices could bridge the gap between laboratory neuroimaging and real-world brain monitoring, with applications spanning neurofeedback, brain-computer interfaces, cognitive assessment, and neurological disorder screening (DOI: 10.1117/1.NPh.10.2.023523).

Kernel, founded by Bryan Johnson (founder of Braintree/Venmo), has invested over $100 million in developing portable brain imaging technology. The Kernel Flow 2 represents the company’s second-generation system, combining TD-fNIRS with EEG in a wearable form factor that enables full cortical coverage for both hemodynamic and electrophysiological brain signals, a capability previously achievable only with laboratory-grade equipment costing hundreds of thousands of dollars.

What Is the Kernel Flow 2?

The Kernel Flow 2 is a wearable brain imaging system that combines two complementary neuroimaging modalities: time-domain functional near-infrared spectroscopy (TD-fNIRS) and electroencephalography (EEG). The system uses a dense modular array architecture supporting up to 40 individual modules, each containing 3 light sources and 6 detectors, creating thousands of measurement channels for comprehensive cortical coverage.

TD-fNIRS works by emitting pulses of near-infrared light through the skull and measuring how the light is absorbed and scattered by brain tissue. Because oxygenated and deoxygenated hemoglobin absorb near-infrared light at different wavelengths, fNIRS can detect changes in cerebral blood oxygenation that indicate neural activity, similar in principle to fMRI but in a wearable, portable form factor. Time-domain fNIRS (as opposed to continuous wave fNIRS) measures the time of flight of individual photons, providing depth-resolved information that allows the system to distinguish brain signals from superficial scalp signals, a critical accuracy improvement.

The EEG integration captures the brain’s electrical activity directly, providing millisecond-resolution temporal information that fNIRS cannot match (fNIRS is limited by the hemodynamic response delay of approximately 5 to 6 seconds). By combining both modalities, Flow 2 captures both where in the brain activity occurs (spatial resolution from fNIRS) and when it occurs (temporal resolution from EEG).

Flow 2 standardizes recording at approximately 3.76 Hz system-wide, tuned for fast whole-head hemodynamic imaging. The system outputs data in SNIRF (Shared Near Infrared Spectroscopy Format), making it compatible with open neuroimaging standards and existing analysis pipelines. Compared to the original Flow 1, Flow 2 improves signal-to-noise ratio and reduces power consumption.

The Science Behind Portable Brain Imaging

Functional neuroimaging allows researchers and clinicians to observe the brain’s activity patterns in real time, providing insight into cognition, emotion, perception, and neurological function. The two most common clinical modalities, fMRI and PET, offer excellent spatial resolution but are confined to clinical environments due to size, cost, and infrastructure requirements.

fNIRS emerged as a portable alternative in the 1990s, using the optical properties of near-infrared light (wavelengths between 650 and 950 nm) to measure cerebral hemodynamics. Near-infrared light penetrates the skull and brain tissue to depths of approximately 2 to 3 centimeters, sufficient to reach the cerebral cortex. Changes in oxygenated and deoxygenated hemoglobin concentrations, which accompany neural activation (through neurovascular coupling), alter the absorption spectrum of the tissue, allowing fNIRS to map cortical activation patterns.

Time-domain fNIRS represents the most advanced form of the technology. By measuring the distribution of photon arrival times (the temporal point spread function), TD-fNIRS can resolve the depth of absorption changes, distinguishing brain hemodynamics from scalp hemodynamics. This depth sensitivity is critical for accuracy because scalp blood flow changes (caused by exercise, temperature, or systemic cardiovascular activity) can contaminate brain signals in simpler continuous-wave fNIRS systems.

EEG measures the electrical potentials generated by cortical pyramidal neurons, providing direct measurement of neural activity with millisecond temporal resolution. The combination of fNIRS (good spatial resolution, poor temporal resolution) with EEG (good temporal resolution, poor spatial resolution) creates a multimodal imaging approach that leverages the strengths of both technologies.

What the Kernel Flow 2 Does Well

The multimodal integration of TD-fNIRS and EEG in a single wearable system is Kernel’s most significant technical achievement. No other commercially available device combines both neuroimaging modalities in a portable form factor. This multimodal approach provides a more complete picture of brain function than either modality alone.

The dense modular array with up to 40 modules and thousands of measurement channels provides unprecedented spatial coverage for a portable fNIRS system. Full cortical coverage enables whole-brain functional mapping during natural behavior, cognitive tasks, or clinical assessments, rather than monitoring only a few brain regions as simpler fNIRS systems do.

The SNIRF data format compatibility ensures that Flow 2 data can be analyzed using established neuroimaging analysis tools and workflows, reducing the barrier to adoption for researchers already working with fNIRS or multimodal brain data.

Featured Partner

Invest in the Infrastructure Behind Modern Medicine

As healthcare expands beyond hospital walls, the buildings and campuses supporting that shift are generating compelling returns for investors who move early. The Healthcare Real Estate Fund offers qualified investors direct access to a curated portfolio of medical office, outpatient, and specialty care facilities.

Learn More →

The improved SNR and reduced power consumption compared to Flow 1 indicate meaningful engineering refinement that should translate to better data quality and longer recording sessions.

Pricing, Access, and Practical Realities

Kernel Flow 2 is positioned as a research and clinical tool rather than a consumer device. Pricing has not been publicly disclosed but is expected to be in the range of research-grade neuroimaging equipment (tens of thousands of dollars). The system is available through Kernel’s website for researchers, clinicians, and developers.

The device is designed for controlled measurement environments (seated subjects, structured tasks) rather than free-living use. While portable, it is not a consumer wearable in the same sense as a fitness tracker or smart ring. Users require training in neuroimaging protocols, data acquisition, and analysis.

Kernel provides a developer program for researchers interested in building applications on the Flow 2 platform, including SDK access and documentation for custom experimental paradigms.

Who It Is Best For

Kernel Flow 2 is best suited for neuroscience researchers conducting studies that require portable brain imaging outside traditional MRI facilities: field studies, naturalistic paradigms, clinical trials in patient populations who cannot tolerate MRI, and neurofeedback research.

Clinical researchers studying neurological and psychiatric conditions (traumatic brain injury, depression, ADHD, neurodegenerative disease) where portable brain imaging enables longitudinal monitoring and real-world assessment that laboratory-only imaging cannot provide.

Brain-computer interface (BCI) developers who need high-quality, multimodal brain signals for developing and testing neural interface applications.

The device is not suited for consumers, wellness enthusiasts, or individuals without neuroscience training, as it requires specialized expertise for both data acquisition and interpretation.

How It Compares

Against research-grade fMRI ($1 million+ acquisition, $500+ per session), Kernel Flow 2 offers portable, repeatable brain imaging at a fraction of the cost and without the constraints of a fixed MRI suite. fMRI provides superior spatial resolution (1mm vs. approximately 10mm for fNIRS) and whole-brain depth coverage, while Flow 2 offers portability and ecological validity.

Against consumer EEG devices (Muse S: $249, Emotiv Insight: $299), Kernel Flow 2 provides dramatically superior data quality, channel density, and the addition of fNIRS hemodynamic imaging that consumer EEG devices cannot offer. Consumer EEG devices are designed for meditation feedback and basic brain state monitoring; Flow 2 is designed for quantitative neuroscience.

Against competing research fNIRS systems (NIRx, Artinis, Shimadzu), Kernel Flow 2’s integration of TD-fNIRS (rather than simpler continuous-wave fNIRS) with EEG in a single wearable provides a unique multimodal capability. Most competing systems offer either fNIRS or EEG, not both integrated.

Limitations and Open Questions

fNIRS is limited to cortical imaging (approximately 2 to 3 cm depth penetration), meaning deeper brain structures (hippocampus, amygdala, basal ganglia, brainstem) are not accessible. For research questions involving subcortical structures, fMRI remains necessary.

The spatial resolution of fNIRS (approximately 10mm) is substantially lower than fMRI (1mm), limiting the precision of functional localization. For fine-grained spatial mapping, Flow 2 cannot match laboratory imaging.

Hair can attenuate optical signals, making fNIRS more challenging to use on participants with thick or dark hair. The modular design may accommodate some hair penetration, but optical contact quality remains a practical consideration.

The research-only positioning and likely high price point place Flow 2 outside the consumer market entirely. Consumer applications of portable brain imaging remain years away.

What This Means for Your Health

While the Kernel Flow 2 is not a consumer device, its significance for health lies in what it enables: portable brain imaging that could accelerate the development of brain health diagnostics, neurofeedback therapies, and cognitive assessment tools that eventually reach consumer applications.

Within Healthcare Discovery‘s Four Shadows framework, neurodegenerative disease is one of the four primary chronic disease threats to longevity. Portable brain imaging technology could enable earlier detection of neurodegenerative changes, more accessible cognitive health screening, and more effective neurofeedback interventions, all of which would contribute to the longevity agenda of maintaining cognitive function across the lifespan.

Within the Five Pillars framework, brain imaging connects to Mindset (cognitive performance and mental health assessment) and Sleep (understanding how sleep quality affects brain function). As this technology matures and becomes more accessible, it could provide the brain health equivalent of what heart rate monitors and CGMs have done for cardiovascular and metabolic health monitoring.

Frequently Asked Questions

What is the Kernel Flow 2?
A wearable brain imaging system combining TD-fNIRS and EEG in a dense modular array with up to 40 modules and thousands of measurement channels for portable, full-cortical brain imaging.

How much does the Kernel Flow 2 cost?
Pricing is not publicly disclosed. The device is positioned as research-grade equipment and is available through Kernel’s website for researchers and developers.

Is the Kernel Flow 2 a consumer device?
No. Flow 2 is designed for neuroscience researchers, clinical investigators, and brain-computer interface developers. It requires specialized training for data acquisition and analysis.

What can Kernel Flow 2 measure?
The system measures cerebral blood oxygenation changes (via TD-fNIRS) and electrical brain activity (via EEG), providing both hemodynamic and electrophysiological brain imaging with cortical coverage.

How is Kernel Flow 2 different from consumer EEG headbands?
Consumer EEG devices (Muse, Emotiv) use a few channels for basic brain state estimation. Flow 2 uses up to 40 modules with thousands of channels plus TD-fNIRS, providing research-grade multimodal brain imaging that consumer devices cannot approach in data quality or spatial coverage.

Free Daily Briefing

The Latest Longevity Science.
Delivered Every Morning.

Join researchers, physicians, and health professionals getting daily breakthroughs in AI-driven medicine, epigenetics, and longevity research.

Support the research that powers this editorial

Subscribe Free

No spam. Unsubscribe anytime. We respect your inbox.