With ARPA-H Award, UC San Diego Engineers will Develop a Soft, Autonomous Robot for Stroke Care

The research teams will work closely with the FDA, as there is no precedent for approving this type of autonomous medical system. In addition to Morimoto, the team includes electrical and computer engineering faculty from UC San Diego, as well as engineering faculty from UC Santa Barbara, Vanderbilt, and the University of North Carolina Chapel Hill. The team includes industry partners from GE Healthcare, EndoTheia, and Vine Medical. Clinical partners at Stanford will provide medical expertise and oversee testing, and physicians at the University of Missouri in Kansas City will lead rural implementation. Full list of co-investigators below.

This project is funded through ARPA-H’s Autonomous Interventions and Robotics (AIR) program led by ARPA-H Program Manager Ileana Hancu, Ph.D.

“The AIR program aims to increase access to life-saving surgeries for all Americans—no matter their zip code. Millions of people will be able to be treated rapidly at their local hospital, and chronic disability after stroke will become a thing of the past,” said Hancu.

A Growing Robot

The breakthrough idea that enables this robot to safely and compliantly navigate the vasculature is that it turns its skin inside out, unfurling from its tip to grow, rather than being pushed from its base. It is made from a soft, thin-walled, hollow tube with one end tucked back inside itself. When pressurized with fluid, the inverted tube material extends from the tip; the robot’s body remains stationary, even as the tip advances. Unlike current tools used for thrombectomies, this robot is soft, meaning it’s inherently safe and nearly impossible to perforate vessels. Growing from the tip means the robot applies less force than current tools, and can form complex 3D shapes to match a tortuous path. This innovative robot design will be combined with hierarchical, uncertainty-aware, explainable autonomous decision making. By capturing uncertainty across planning and navigation, the system will enable safety-aware interpretable decisions for navigation and therapy delivery.

The system is softer than any prior catheter, requires minimal tool changes, and its simplicity enables the system to be low profile. It will not add significant cost or procedure time compared to conventional catheters in the hands of experts, facilitating adoption in low-resource or rural settings where expert surgeons are not available.

“In addition to performing aspiration, we are aiming for this same robot to autonomously perform both coil embolization and angiography as well,” said Morimoto.

Researchers in Morimoto’s lab have already demonstrated that a robot equipped with this everting skin could successfully navigate a model of human arteries, the GI tract, and an airway.

Stroke Care, Anywhere

More than half of Americans live over an hour away from a hospital that can perform a thrombectomy. The longer treatment is delayed, the higher the risk of serious long-term disability or death. The goal of this project is to have these medical robots deployed all over the country, where technicians would insert them into a patient through the femoral artery in the groin — the current entry method for many thrombectomies today — and then a physician would oversee the procedure from afar.

Patients would undergo standard pre-operative imaging to locate the blood clot and map their specific vasculature. As the procedure continues, contrast-enhanced images would be taken as it navigates the branches of the patient’s veins to ensure an accurate path. Once the robot has reached the clot, it will remove it using suction — a process called aspiration. Once the robot captures the clot, it will retract under continuous suction to remove the clot, and will inject contrast to verify that the vessels are fully reopened.

The research team aims to have a working prototype for use in preclinical research in five years.

Research Team:

In addition to lead investigator Morimoto, the research team includes UC San Diego electrical and computer engineering faculty Michael Yip, who will lead the automation team, and Nikolay Atanasov, who will lead computational simulation. Elliot Hawkes, associate professor of mechanical engineering at UC Santa Barbara, will lead the growing robot design and modeling. Robert Webster, professor of mechanical engineering and medicine at Vanderbilt, will assist with system design and clinical translation; Joshua Gafford, CTO of EndoTheia Inc, will lead the design of the steerable introducer and coordinate regulatory milestones along with David Haggerty, CEO of Vine Medical. Ron Alterovitz, professor of computer science at UNC Chapel Hill, will lead the effort to encode intelligence into the physical robot. Brian Yanoff, principal scientist at GE HealthCare, will lead imaging work; Jeremy Heit, professor of radiology at Stanford, will oversee fluoroscopy experiments and lead preclinical testing. Alexander Norbash, Dean of the School of Medicine at the University of Missouri at Kansas City will lead rural implementation.

This research was, in part, funded by the Advanced Research Projects Agency for Health (ARPA-H). The views and conclusions contained in this document are those of the authors and should not be interpreted as representing the official policies, either expressed or implied, of the United States Government.