A 3-D Volume Coverage Path Planning Algorithm With Application to Intracerebral Hemorrhage Evacuation
IEEE Robotics and Automation Letters, 1 (2), pp. 876–883, 2016
Abstract
This letter presents a new heuristic 3-D volume coverage path planning (VCPP) algorithm for robotic intracerebral hemorrhage evacuation. In contrast to existing 3-D planning techniques, the proposed algorithm generates 3-D paths without first decomposing the volume into series of 2-D planning problems. It considers the morphology of the volume to be covered and minimizes the configuration or task space distance traveled. The algorithm merges elements from existing grid-based and wavefront approaches and accommodates kinematic and environmental constraints, as well as obstacle avoidance. We provide both simulation and experimental demonstrations of the algorithm in the context of intracerebral hemorrhage evacuation where a curved, needle-like robot must suction out blood from within the brain by covering the interior of a semicoagulated blood-filled volume with its tip. We perform a simulation study with 7 patient datasets and compare the distance traveled with our new algorithm using a conventional 2-D layer-by-layer planning approach. We also perform 3 in vitro evacuation experiments on phantoms made to match patient hemorrhage geometries. Our results illustrate that the VCPP algorithm economizes motion and is more efficient than a layer-by-layer CPP approach.