TY - JOUR
T1 - Path planning for endovascular catheterization under curvature constraints via two-phase searching approach
AU - Li, Zhen
AU - Dankelman, Jenny
AU - De Momi, Elena
PY - 2021
Y1 - 2021
N2 - Purpose: Planning a safe path for flexible catheters is one of the major challenges of endovascular catheterization. State-of-the-art methods rarely consider the catheter curvature constraint and reduced computational time of path planning which guarantees the possibility to re-plan the path during the actual operation. Methods: In this manuscript, we propose a fast two-phase path planning approach under the robot curvature constraint. Firstly, the vascular structure is extracted and represented by vascular centerlines and corresponding vascular radii. Then, the path is searched along the vascular centerline using breadth first search (BFS) strategy and locally optimized via the genetic algorithm (GA) to satisfy the robot curvature constraint. This approach (BFS-GA) is able to respect the robot curvature constraint while keeping it close to the centerlines as much as possible. We can also reduce the optimization search space and perform parallel optimization to shorten the computational time. Results: We demonstrate the method’s high efficiency in two-dimensional and three-dimensional space scenarios. The results showed the planner’s ability to satisfy the robot curvature constraint while keeping low computational time cost compared with sampling-based methods. Path replanning in femoral arteries can reach an updating frequency at 6.4 ± 2.3 Hz. Conclusion: The presented work is suited for surgical procedures demanding satisfying curvature constraints while optimizing specified criteria. It is also applicable for curvature constrained robots in narrow passages.
AB - Purpose: Planning a safe path for flexible catheters is one of the major challenges of endovascular catheterization. State-of-the-art methods rarely consider the catheter curvature constraint and reduced computational time of path planning which guarantees the possibility to re-plan the path during the actual operation. Methods: In this manuscript, we propose a fast two-phase path planning approach under the robot curvature constraint. Firstly, the vascular structure is extracted and represented by vascular centerlines and corresponding vascular radii. Then, the path is searched along the vascular centerline using breadth first search (BFS) strategy and locally optimized via the genetic algorithm (GA) to satisfy the robot curvature constraint. This approach (BFS-GA) is able to respect the robot curvature constraint while keeping it close to the centerlines as much as possible. We can also reduce the optimization search space and perform parallel optimization to shorten the computational time. Results: We demonstrate the method’s high efficiency in two-dimensional and three-dimensional space scenarios. The results showed the planner’s ability to satisfy the robot curvature constraint while keeping low computational time cost compared with sampling-based methods. Path replanning in femoral arteries can reach an updating frequency at 6.4 ± 2.3 Hz. Conclusion: The presented work is suited for surgical procedures demanding satisfying curvature constraints while optimizing specified criteria. It is also applicable for curvature constrained robots in narrow passages.
KW - Autonomous endovascular intervention
KW - Curvature constraints
KW - Flexible catheter
KW - Path planning
KW - Robotic surgery
UR - http://www.scopus.com/inward/record.url?scp=85102457929&partnerID=8YFLogxK
U2 - 10.1007/s11548-021-02328-x
DO - 10.1007/s11548-021-02328-x
M3 - Article
AN - SCOPUS:85102457929
VL - 16
SP - 619
EP - 627
JO - International Journal of Computer Assisted Radiology and Surgery
JF - International Journal of Computer Assisted Radiology and Surgery
SN - 1861-6410
IS - 4
ER -