computational robotics

from tasks to motions

Medical Robotics

Minimally Invasive Robotic Exploration of Osteolytic Lesions

Osteolysis is a result of bearing material wear in a total hip arthroplasty (THA). These minute particles cause an inflammatory reaction in bone near the screws and eventually cause bone necrosis behind the cup and around the screws, as shown in the CT scan. To treat this condition, the infected bone lesion must be removed through revision surgery.

Traditional open surgery requires extracting all components of the implant. The removal process is invasive and risky. A minimally-invasive alternative could attempt to preserve acetabular and femoral components of the THA. Robotic systems can clean out the osteolytic regions through existing screw holds of the metallic acetabular cup. These robotic systems are designed to act as snake-like cannulas through which cleaning tools could be inserted and guided. The objective of such a robot in minimally-invasive treatment of osteolysis is to explore as much as possible of the osteolytic cavity (the region of viscous dead bone) in order to allow full access of the infected region to the inserted tool.

The contribution of this research is the development of a motion planner termed "Osteolytic Region Exploration" (ORE) that enables a snake-like cannula robot in minimally-invasive treatment of osteolysis to effectively explore with its tip the osteolytic cavity. ORE provides a first step toward a practical framework to validate and guide the design of robots for minimally-invasive treatment of osteolytic lesions. Such framework can be used by the orthopaedic surgeons to estimate whether the current robotic minimally-invasive approach effectively explores the osteolytic cavity to allow removal of a sufficient amount of bone lesion for clinical decisions. This can be a valuable pre-operative planning tool. If surgeons determine a lesion cannot be sucessfully accessed, they may consider tool insertion points aside from the implant screw channels or even revert back to full implant removal.

ORE draws from sampling-based motion planning to design effective exploration strategies of unknown osteolytic cavities. In particular, ORE leverages the idea of configuration-space sampling to guide exploration toward unexplored regions. To handle the fact that a geometric model of the osteolytic cavity is not available, ORE relies instead on sensing. Sensor-based exploration of osteolytic cavities presents unique challenges. While sensor-based research has generally focused on mobile robotics, snake-like robots used in minimally-invasive treatment of osteolysis are high-dimensional manipulators. Furthermore, osteolytic cavities severely constrain the possible motions of the robot and impose stringent constraints on how the robot should bend in order to move from one place to the other. To overcome these challenges, ORE augments sensor-based exploration with a global layer. The global layer, which is obtained by imposing a grid decomposition, acts as a guide to determine which grid region should be explored. The local exploration layer uses information gain to move the robot tip toward positions in the region that increase exploration. Simulation experiments with a snake-like robot and surgically-relevant osteolytic cavities indicated that by combining local exploration, information gain, and global path planning, ORE effectively explored the cavity.

Related Publications

  • Liu WP, Lucas BC, Guerin K, and Plaku E (2011): "Sensor and Sampling-based motion planning for minimally invasive robotic exploration of osteolytic lesions" IEEE/RSJ International Conference on Intelligent Robots and Systems, pp. 1346--1352   [publisher]   [preprint]