 |
|||||
 |
 |
 |
 |
 |
 |
Motion Planning for Active Cannulas
We explore motion planning for active cannulas, new needle-like medical devices composed of thin, pre-curved, telescoping tubes. These devices can follow curved, winding paths through open cavities and soft tissues to reach distant targets in constrained spaces, enabling minimally-invasive access to previously unreachable sites. In an active cannula, bending actuation arises as tubes elastically interact when they slide and rotate within one another. Planning tube configurations (translations and rotations applied at tube bases) that correspond to desired spatial tip coordinates or shaft curves is not intuitive for human beings, motivating the need for efficient planning algorithms. We are developing new planning algorithms to compute optimal motions for these devices to reach clinician-specified targets while avoiding anatomical obstacles. This is challenging due to the device's geometric structure and beam mechanics. We formulate the planning problem as a constrained nonlinear optimization problem and use a penalty method to convert this formulation into a sequence of more easily solvable unconstrained optimization problems. Using a preliminary piece-wise constant-curvature model, our planner, in approximately one minute, is capable of finding solutions that reach targets while avoiding spherical obstacles. This research is in collaboration with the MEDLab at Vanderbilt University. Publications/Presentations
|
