Researchers have established the first international consensus on how to design, test and evaluate robotic systems for stroke treatment via a new position statement published in the Journal of the American Heart Association (JAHA) that is intended to define standards that put patient safety first.
As per a media release from King’s College London (London, UK)—one of the leading institutions that helped to develop the consensus statement—no such agreed framework existed until now regarding robotic systems for mechanical thrombectomy. And, while robotics hold the potential to expand access to thrombectomy by enabling stroke specialists to perform these procedures remotely, progress has been hampered by a lack of standardisation across studies.
“In 2023, we carried out a systematic review of endovascular robotics and autonomy,” said Harry Robertshaw (King’s College London, London, UK), first author of the consensus statement. “We found that, although several studies existed, they were all testing different tasks, using different models and measuring different outcomes. This made it impossible to compare results or identify the most effective approaches.”
To address this, the team brought together an international group of experts in interventional neuroradiology, robotics, data science, health economics, policy, statistics and patient advocacy to establish consensus frameworks for developing and validating robotic mechanical thrombectomy. The resulting position statement sets out how these systems should be tested, and how their effectiveness should be measured and reported.
“By bringing together experts from clinical practice, academia, industry, and patient representatives, we have defined the first consensus standards for robotics and AI [artificial intelligence] in thrombectomy navigation,” Robertshaw added. “This will help to move the technology forward as its integration with clinical practice moves closer to reality.”
According to the researchers, patient perspectives—including input from patient representatives and organisations like the UK Stroke Association—were key to retaining the centrality of safety, real-world impact and patient benefit.
The researchers hope their position statement will drive more comparable studies, accelerating progress across the field. One requisite patient safety task outlined by the Stakeholder Taskforce for AI-assisted Robotic Thrombectomy (START) in the statement is the present need to correlate in-vitro measurements to in-vivo complications.
“While AI-assisted robotic mechanical thrombectomy is not yet ready for routine clinical use, the speed of innovation in both robotics and AI means transformative advances are likely to be just around the corner,” commented senior author Thomas Booth (King’s College London, London, UK). “Our work lays the foundation for that future by defining clear standards to ensure that global development and validation are well understood.
“By establishing practical, consensus-driven recommendations, we are helping to ensure that—when these technologies do reach patients—they do so safely, responsibly and with meaningful clinical benefit.”
Autonomous thrombectomy navigation
This position statement is already being put to use, with the King’s College London team recently demonstrating “for the first time” that AI can autonomously perform mechanical thrombectomy navigation in a physical lab setting—representing another step toward expanding access to life-saving stroke treatments.
The team developed a robotic system that uses AI to safely navigate catheters and guidewires through the complex pathway of blood vessels from the leg to the brain during a thrombectomy, publishing their findings in IEEE Robotics and Automation Letters.
“Autonomous robotics could allow equal access to life-saving stroke treatment no matter where you live in the world,” Robertshaw stated.
To address the “long and complex” navigation from the entry point to the brain, the researchers used a machine learning approach via which multiple AI ‘agents’ are responsible for different parts of the pathway, rather than relying on a single AI for the entire procedure.
The team tested the system using both computer models and 3D-printed models of human blood vessels. They found that, overall, it could successfully navigate these complex vascular pathways.
“This work provides the first demonstration that thrombectomy can be done solely by AI in a lab setting, outside of simulations,” Robertshaw explained.
“Translating these advances into real-world environments means that we are one step closer towards the goal of improving outcomes for patients by introducing autonomous endovascular procedures into the clinic,” Booth concluded.
