Magnendo Corp., an MIT spinout, received a grant of up to $32 million from the U.S. agency ARPA-H on August 8, 2026. Over five years, the money is meant to fund an autonomous robot that magnetically steers catheters through blood vessels during stroke treatment.
Key takeaways
- Grant of up to $32M from ARPA-H, spread over five years
- Funded under the AIR program (Autonomous Interventions and Robotics)
- Technology: magnetic navigation of catheters and guidewires inside blood vessels
- Clinical goal: mechanical thrombectomy for large vessel occlusion in the brain
- Preclinical tests: faster navigation and more consistent results regardless of physician experience
Magnetic navigation inside vessels
Magnendo is building a system that uses an external magnetic field to steer the tip of a catheter and guidewire inside blood vessels. Instead of manually pushing tools through a winding arterial network, the robot is meant to guide them to a clot in the brain. Ultimately the company wants to combine robotics, medical imaging and artificial intelligence into a single autonomous intervention platform, developed with academic and medical partners.
The grant in numbers:
The problem: too few doctors, too few hospitals
Mechanical thrombectomy — removing a clot from a brain artery — saves lives but is hard to access. Only a small number of trained specialists perform it at a limited number of hospitals, and complex vascular anatomy lengthens the procedure. In stroke, every minute of delay worsens the outcome.
As Ileana Hancu, a program manager at ARPA-H, notes, hundreds of thousands of patients each year suffer large vessel occlusions?large vessel occlusions: A blockage of one of the brain's major arteries — the most severe type of ischemic stroke, requiring urgent clot removal., yet many cannot access the procedure because of geographic and staffing constraints.
Autonomy instead of teleoperation
Endovascular robotics is not new — there are remotely operated systems in which a physician drives the tools with a joystick. Magnendo aims higher: at autonomous navigation, where the robot charts its own path to the clot. In preclinical testing the system achieved faster navigation and more repeatable results across physicians of varying experience. That repeatability, not raw speed, is the project's biggest promise.
| Approach | Control | Dependence on physician experience |
|---|---|---|
| Manual procedure | Hand-guided tools | High |
| Teleoperated robot | Physician via joystick | Medium |
| Magnendo (autonomous) | Autonomous magnetic navigation | Lower (goal) |
Why it matters
The ARPA-H grant places autonomous medical robotics at the center of health policy, not just in the lab. If a machine can perform the key step of the procedure as well as an experienced neuroradiologist, stroke care could move beyond a handful of specialized centers. It is a shift in model: instead of training more rare experts, the procedure itself is standardized.
The stakes, however, are safety — autonomy inside the brain's vessels demands clinical evidence that does not yet exist at this stage.
What's next?
- The five-year AIR program aims to integrate robotics, imaging and AI into one intervention platform
- The project is in the research-and-development phase — ahead of clinical trials and regulatory clearance
