Projektdetaljer
Beskrivelse
Background and Motivation
Cardiovascular Diseases (CVDs) remain the leading cause of death worldwide, accounting for approximately 17.9 million deaths annually. In Denmark, 1 in 5 deaths is due to CVDs. Endovascular interventions have become a cornerstone in managing cardiovascular conditions. These procedures, which involve the navigation of guidewires and catheters through the vascular system to reach target lesions, offer the advantages of minimal invasiveness, reduced hospital stay, and improved patient outcomes. Despite the advancement of robotic systems in surgical fields, endovascular robotics continues to face several critical limitations: 1) The setup time for many endovascular robotic platforms is prohibitively long; 2) These systems are typically large and bulky; 3) Many robotic catheter systems are restricted to using only the tools manufactured by the same vendor; 4) Current systems offers no level of autonomy.
Method
This project is organized into two tightly coupled Work Packages (WPs) as shown in the figure. WP1 focuses on the development of a miniaturized, universally compatible catheter/guidewire manipulation system. WP2 addresses intelligent, autonomous catheter control using AI.
In WP1, a compact robotic system capable of catheter steering will be designed and prototyped, including translation, rotation, and bending motion. Unlike existing bulky platforms, our proposed device will be compact enough for bedside placement or body mounting. It will offer a plug-and-play interface, ensuring compatibility with a broad range of commercially available guidewires and catheters. To enable intelligent control and feedback, force sensors will be embedded to measure resistance forces during catheter translation and bending.
In WP2, a learning-based framework for autonomous guidewire control will be developed. A high-fidelity simulation environment, built using SOFA or MuJoCo, will be created to accurately replicate human vascular anatomy and support safe and scalable training. Once robust control policy is achieved in simulation, sim-to-real transfer methods, such as domain randomization, will be employed to deploy the trained model onto the physical robotic platform from WP1. The autonomous navigation capabilities will be assessed on a phantom based on metrics including trajectory tracking error, task success rate, and robustness to anatomical variability. Through the integration of hardware (WP1) and intelligent control (WP2), this project aims to create a new generation of endovascular robotic systems and capable of performing (semi)-autonomous navigation. The results of this project have the potential to be applied to other flexible instruments, such as bronchoscopes. The project will be conducted in collaboration with CCR and the cardiologists at OUH.
The project is funded by Fabrikant Vilhelm Pedersen og Hustrus Legat.
Cardiovascular Diseases (CVDs) remain the leading cause of death worldwide, accounting for approximately 17.9 million deaths annually. In Denmark, 1 in 5 deaths is due to CVDs. Endovascular interventions have become a cornerstone in managing cardiovascular conditions. These procedures, which involve the navigation of guidewires and catheters through the vascular system to reach target lesions, offer the advantages of minimal invasiveness, reduced hospital stay, and improved patient outcomes. Despite the advancement of robotic systems in surgical fields, endovascular robotics continues to face several critical limitations: 1) The setup time for many endovascular robotic platforms is prohibitively long; 2) These systems are typically large and bulky; 3) Many robotic catheter systems are restricted to using only the tools manufactured by the same vendor; 4) Current systems offers no level of autonomy.
Method
This project is organized into two tightly coupled Work Packages (WPs) as shown in the figure. WP1 focuses on the development of a miniaturized, universally compatible catheter/guidewire manipulation system. WP2 addresses intelligent, autonomous catheter control using AI.
In WP1, a compact robotic system capable of catheter steering will be designed and prototyped, including translation, rotation, and bending motion. Unlike existing bulky platforms, our proposed device will be compact enough for bedside placement or body mounting. It will offer a plug-and-play interface, ensuring compatibility with a broad range of commercially available guidewires and catheters. To enable intelligent control and feedback, force sensors will be embedded to measure resistance forces during catheter translation and bending.
In WP2, a learning-based framework for autonomous guidewire control will be developed. A high-fidelity simulation environment, built using SOFA or MuJoCo, will be created to accurately replicate human vascular anatomy and support safe and scalable training. Once robust control policy is achieved in simulation, sim-to-real transfer methods, such as domain randomization, will be employed to deploy the trained model onto the physical robotic platform from WP1. The autonomous navigation capabilities will be assessed on a phantom based on metrics including trajectory tracking error, task success rate, and robustness to anatomical variability. Through the integration of hardware (WP1) and intelligent control (WP2), this project aims to create a new generation of endovascular robotic systems and capable of performing (semi)-autonomous navigation. The results of this project have the potential to be applied to other flexible instruments, such as bronchoscopes. The project will be conducted in collaboration with CCR and the cardiologists at OUH.
The project is funded by Fabrikant Vilhelm Pedersen og Hustrus Legat.
| Akronym | VP-GRACE |
|---|---|
| Status | Igangværende |
| Effektiv start/slut dato | 01/09/2025 → 31/12/2026 |
Fingerprint
Udforsk forskningsemnerne, som dette projekt berører. Disse etiketter er oprettet på grundlag af de underliggende bevillinger/legater. Sammen danner de et unikt fingerprint.