Aleph Neuro launches with a claim of MRI-level brain images from outside the skull

Why it matters

Aleph Neuro is attacking the core BCI tradeoff: implants offer signal quality, while non-invasive tools offer reach. Its ultrasound claim needs proof, not hype.

Aleph Neuro publicly launched on June 25 with a claim that it has captured what it describes as the highest-resolution 3D images of the human brain taken from outside the skull, an assertion that puts the new research lab squarely in the race to make brain interfaces less dependent on surgery and MRI-scale infrastructure.

https://x.com/alephneuro/status/2070183632132845995

The announcement came through an X post and was amplified the same day by Butterfly Network, the public ultrasound company whose chip platform Aleph Neuro is using. Butterfly said Aleph Neuro is part of its Butterfly Embedded licensing and co-development program, which lets outside teams build on Butterfly's Ultrasound-on-Chip technology rather than develop a custom ultrasound stack from scratch.

Aleph Neuro's own framing is deliberately ambitious. Butterfly's release describes the lab as building "brain interfaces for the telepathic future" and says Aleph Neuro claimed it can deliver "MRI-level detail of the brain" without drilling into the skull or putting a subject inside an MRI machine. Those are Aleph Neuro's claims, not independently validated clinical results. The public materials tied to the launch did not include a peer-reviewed paper, regulatory clearance, a clinical study, or enough technical detail to benchmark the images against MRI, CT, functional ultrasound, or other ultrasound brain-imaging systems.

The clearest public human thread behind Aleph Neuro is Lev Chizhov, who lists Aleph Neuro in San Francisco and Ecole Polytechnique on his LinkedIn profile. Chizhov also appears in an earlier brain-interface experiment that helps explain the lab's operating style: publish the technical idea, show the apparatus, and make the bet before the formal science has caught up. In a prior Write to Brain post, Chizhov, Albert Yan-Huang, Thomas Ribeiro, and Aayush Gupta described using focused ultrasound to stimulate scent-processing brain regions and induce smell-like sensations, including fresh air, garbage, ozone, and burning wood.

That earlier work was not a conventional medical-device study. It described a small, self-directed setup involving a handheld probe, a makeshift headset, MRI-based targeting for Chizhov's skull, and parameters including 300 kHz ultrasound, a roughly 39 mm focal depth, and 50 to 55 degree steering angles. The team said it replicated the sensations in two people and ran a blind trial, but the writeup also showed the gap between garage neurotech and clinical evidence: tiny sample size, improvised hardware, and no independent replication in the post itself.

That context matters because Aleph Neuro's launch is not just another imaging demo. It is a bet that ultrasound can become a general-purpose interface layer for the brain: cheap enough to escape the MRI room, precise enough to compete with higher-resolution modalities, and non-invasive enough to avoid the neurosurgical threshold that defines companies such as Neuralink. The technical obstacle is the skull. Bone distorts and attenuates ultrasound, which is why much of the most compelling human ultrasound brain work either uses careful acoustic modeling, lower frequencies, implanted windows, or surgical access.

The broader field gives Aleph Neuro a real opening, but not a free pass. A 2020 review in Neuron described ultrasound as unusually attractive for neuroscience because it can penetrate the brain and, in some cases, the skull, while offering spatial and temporal performance relevant to large-scale imaging and modulation. Separately, a Caltech-led 2024 proof of concept used ultrasound imaging for a brain-computer-interface experiment outside a hospital setting, but that work used a cranial prosthetic window that allowed ultrasound to pass without the normal skull barrier. Aleph Neuro is claiming a harder version of the problem: imaging from outside an intact skull.

For Butterfly Network, the incentive is straightforward. Aleph Neuro gives Butterfly another proof point for the Embedded program, where Butterfly can turn its core ultrasound chip into a platform business beyond point-of-care scanners. In its first-quarter 2026 results, Butterfly reported $26.5 million in revenue, up 25% year over year, and said U.S. revenue benefited from Butterfly Embedded partnerships, including a co-development partnership with a Butterfly Embedded partner named Midjourney. The Aleph Neuro launch lets Butterfly point to a more speculative frontier use case for the same semiconductor ultrasound thesis.

That does not make Aleph Neuro a medical company yet in the way investors, regulators, or hospital buyers would recognize. The launch materials did not disclose funding, valuation, FDA status, named clinical partners, image-resolution metrics, number of subjects, or commercial timeline. They also did not say whether Aleph Neuro is building a research instrument, a regulated diagnostic device, a brain-computer-interface input system, or an eventual consumer neurotechnology product.

What Aleph Neuro has done is choose the bottleneck. The invasive BCI market is built around electrodes and surgery. The non-invasive BCI market has spent years wrestling with low signal quality, bulky optics, EEG limitations, and the gap between lab demos and repeatable products. Aleph Neuro is betting that ultrasound can cut through that tradeoff: enough access to the brain to matter, without the implant.

The launch gives the lab attention and gives Butterfly a platform story. The next test is whether Aleph Neuro can move from company-described images and public demos into reproducible data that outside neuroscientists can interrogate. In brain-interface startups, that is the line between an interesting apparatus and a company.