Aleph Neuro reports what it says is the first 3D ultrasound localization microscopy image of a living human brain through an intact skull

By Inside BCI · June 28, 2026

A San Francisco research lab calling itself Aleph Neuro published what it describes as the first 3D vascular image of a living human brain captured through an intact skull on 24 June 2026, using ultrasound chips licensed from Butterfly Network. The image is a static map of the brain’s microvasculature, generated by an imaging technique called ultrasound localization microscopy (ULM) with a four-minute infusion of FDA-approved sulfur hexafluoride microbubble contrast. Aleph open-sourced the imaging pipeline and its underlying dataset at github.com/alephneuro/microbubbles. Butterfly Network, Inc. (NYSE: BFLY) responded with a BusinessWire release the following day, framing Aleph as the second publicly named participant in its Butterfly Embedded licensing and co-development program after Midjourney’s full-body ultrasound scanner unveiling the prior week. Aleph self-describes as “a research lab” and discloses no founders, no funding round, no chief executive, no head count, and no clinical partner.

What the milestone actually is, and what it is not

Aleph’s image is a microvascular map. The technique, ultrasound localization microscopy, infuses microbubbles (lipid-encapsulated pockets of sulfur hexafluoride gas, a contrast agent approved for clinical use by the US FDA in non-brain indications) into the patient’s bloodstream while a Butterfly ultrasound chip captures the bubbles flowing through cerebral blood vessels. Each microbubble produces a strong ultrasound echo. Software locates each bubble at a sub-diffraction-limit spatial resolution by fitting the centre of its point-spread function. Stacking millions of localisations over the four-minute acquisition produces a 3D vascular reconstruction that, in Aleph’s footnoted claim, achieves “a resolution 100 times greater volumetrically than comparable CT” in the contrast-enhanced super-resolution regime.

What this is not: the image is not a functional decode of brain activity, and it is not a brain-computer interface in the operational sense. ULM produces a static anatomical map of the vasculature. Functional ultrasound imaging (fUS), the related modality that maps dynamic blood-flow changes over time to infer neural activity through neurovascular coupling, is the broader technical category. Aleph itself names its long-term destination as “contrast-free neurovascular imaging of the brain” rather than as fUS specifically. The June 2026 publication is the imaging substrate. The decoding system Aleph wants to build on top of it has not been demonstrated.

What this is, technically: prior published transcranial functional ultrasound work used either non-human-primate surgical preparations or, in living humans, an implanted acoustically transparent cranial window. The 2021 Norman-Andersen Neuron paper from the Caltech group decoded single-trial movement intentions via fUS in macaques with a craniectomy over posterior parietal cortex, dura intact. The 2024 Rabut-Shapiro-Andersen Science Translational Medicine paper extended functional ultrasound imaging to awake human subjects through a permanently implanted acoustically transparent cranial window. Aleph’s claim of capturing 3D vascular images through an intact human skull is the novel piece. Skull bone attenuates and scatters ultrasound aggressively, particularly at the frequencies needed for sub-millimetre spatial resolution. Whether Aleph’s approach scales to functional decoding through the same intact-skull constraint, and at what temporal resolution, is the question the next dataset will answer.

The Butterfly Embedded relationship, in plain language

Aleph is a customer of Butterfly Network’s licensing program, not a Butterfly subsidiary, joint venture, or equity-controlled entity. Butterfly’s BusinessWire release uses the verbatim language: “Aleph is a participant in the Butterfly Embedded™ licensing and co-development program.” The same program publicly names Forest Neurotech (the philanthropically funded non-invasive BCI venture co-founded by Sumner Norman, Tyson Aflalo, and Will Biederman, anchored by a $14 million March 2024 commitment via Schmidt Futures and Convergent Research) as an earlier co-development partner, and Midjourney (the AI-image-generation company whose imaging arm announced a Butterfly-powered 40-module full-body ultrasound scanner the prior week, with a single San Francisco wellness centre targeted for late 2027) as the first publicly named Embedded participant in the 2026 commercial wave.

The Embedded program’s economics are publicly disclosed only for Midjourney, through Butterfly’s November 2025 Form 8-K: $15 million upfront, $10 million in annual licence fees, plus milestones and revenue sharing, totalling up to $74 million over five years. Aleph’s deal terms are not disclosed in the release, the company’s website, or any secondary coverage. Whether Aleph paid the same upfront figure, a smaller research-tier figure, or operates under a different structure is unknown.

Butterfly’s chip is a single-die capacitive micromachined ultrasonic transducer (CMUT) integrating more than 9,000 ultrasound elements across a 1 to 12 MHz frequency range. Founder Jonathan M. Rothberg, who previously founded Ion Torrent and 454 Life Sciences, built Butterfly to bring point-of-care ultrasound to handheld form factor; the company went public in 2021 via a SPAC merger with Longview Acquisition Corp and trades on the New York Stock Exchange under BFLY. President, Chief Executive Officer, and Chairman Joseph DeVivo issued the only named-person quote in the launch release: “Researchers like Aleph Neuro are using Butterfly’s Ultrasound-on-Chip™ to reimagine the capabilities of ultrasound and challenge the status quo. I encourage everyone to go to AlephNeuro.com to see what this exciting new lab is venturing to do. We congratulate the Aleph team on their progress.”

Where this sits in the non-invasive BCI commercial map

Aleph’s category is non-invasive brain-computer interface via transcranial ultrasound imaging. The category is small and structurally distinct from the implantable-BCI cohort (Neuralink, Synchron, Paradromics, Precision Neuroscience, Neuracle) on three dimensions. First, the regulatory pathway runs through an imaging or contrast-imaging clearance route rather than through Class III implantable-device routes. Second, the patient population is potentially open-ended rather than limited to paralysed or ALS cohorts. Third, the throughput question is bandwidth: implantable BCIs read individual neuron action potentials; ultrasound-vascular methods read aggregated blood-flow signals that are spatially detailed but temporally slower.

The publicly visible competitors in the transcranial-ultrasound-for-BCI lane are Iconeus, the Paris-based fUS company founded by Mickael Tanter at INSERM and ESPCI Paris in 2016 and now selling the Iconeus One preclinical scanner; Forest Neurotech, on the implantable-acoustic-window track with the Butterfly chip; Sam Altman’s Merge Labs, building toward a sonogenetics-and-ultrasound platform; and two China-based companies, Gestala ($21.6 million March 2026, chronic pain focus) and BCI-Sonics (more than ¥100 million angel funding, low-intensity focused ultrasound plus AI). The intact-skull-imaging claim, if it holds across larger datasets and across functional time series rather than static maps, would position Aleph differently from all five.

The category’s regulatory and commercial open question is not whether ultrasound can image the brain (it can) but whether ultrasound can decode the brain at bandwidth competitive with implantable arrays. The Iconeus product line addresses preclinical animal research; the Caltech-group transcranial work required surgical access (non-human-primate craniectomy in 2021 and a permanently implanted human acoustic window in 2024). Aleph’s bet is that the chip-scale CMUT array plus the ULM image-reconstruction stack plus a sufficient dataset will close the gap between vascular imaging and functional decoding without the surgery. The commercial timeline depends on the next dataset Aleph publishes and the regulatory pathway it elects when it moves from research-lab status to product-clearance status.

What is missing from the public record

Aleph Neuro discloses unusually little for a launching company. No CEO is named. No founders are named. No funding round is disclosed. No board of directors is published. The team is described collectively as “physicists and engineers in San Francisco.” The company’s domain (alephneuro.com) lists only a blog post, a careers page, and an X handle (@alephneuro). The corporate structure (Inc, LLC, public benefit corporation) is not stated.

The silence has two plausible readings. The first is that Aleph is operating in a research-lab mode similar to mid-stage research-only ventures that publish science before raising visible capital. The second is that Aleph is raising or has raised capital and is choosing to keep team and economics private through the technical-publication launch. The next material datapoint will distinguish between the two: either a peer-reviewed publication of the dataset, or a venture-funding announcement with founders named and round size disclosed.

What to watch

The first signal is whether the open-sourced pipeline and dataset at github.com/alephneuro/braindump attract follow-up validation work from independent academic groups, particularly the Caltech, INSERM-ESPCI, and Stanford groups already working in transcranial ultrasound. Independent reconstruction of the human-brain ULM image from the published pipeline would convert Aleph’s claim from a press-release milestone into a methodologically replicated result.

The second signal is whether Aleph publishes a follow-up demonstrating functional time-series imaging (the fUS step) through the same intact-skull preparation, rather than the static ULM map. A functional time series in a behaviorally engaged human subject would be the substantive transition from imaging publication to BCI publication.

The third signal is whether Aleph names a chief executive, a founding team, or a venture-funding round within the next two quarters. Continued anonymity past the technical milestone would suggest a deliberately unusual commercial posture. Disclosure on the standard timeline would suggest a more conventional venture-funded BCI company that simply led with the imaging publication.

Sources

news Aleph-Neuro Butterfly-Network Butterfly-Embedded Ultrasound-on-Chip ultrasound-localization-microscopy ULM non-invasive-BCI functional-ultrasound microbubble-contrast San-Francisco Joseph-DeVivo Jonathan-Rothberg Forest-Neurotech Iconeus

What the milestone actually is, and what it is not

The Butterfly Embedded relationship, in plain language

Where this sits in the non-invasive BCI commercial map

What is missing from the public record

What to watch

Weekly BCI Brief in your inbox