The Precision Freehand Sculptor: A robotic tool for less invasive joint replacement surgery

Abstract

This thesis presents a tool for less invasive joint replacement surgery. Although many surgical procedures have been converted to endoscopic "keyhole" approaches, joint replacement incisions have changed little in 30 years. Recent efforts to adapt conventional joint replacement instrumentation for less invasive approaches have demonstrated improved short-term outcomes. However, the procedures are more challenging to use accurately and are only suitable for highly skilled surgeons. The Precision Freehand Sculptor (PFS) is a handheld intelligent tool designed to enable less invasive joint replacement surgery. A small rotary blade at the tip of a long, slender nose allows the surgeon to shape the bone to accept the implant. The blade can extend and retract behind a guard under computer control. As the surgeon moves the tooltip freehand over the surface of the bone, the computer extends and retracts the blade so that only the appropriate material is removed. An optical tracking camera continuously monitors the 3D position of infrared markers attached to the tool and bone. The PFS computer compares the blade's location on the bone to the target shape which it has been instructed to cut. It retracts or extends the blade accordingly. The target shape is specified to mate well with the implant and position the implant for proper biomechanics. The PFS has the potential to make less invasive joint replacement accessible to more surgeons without sacrificing accuracy. The computer controlled blade ensures an accurate cut even if the tip of the tool is obscured from view. The long, slender nose is ideal for operating through small incisions. A computer display provides additional guidance to the surgeon when visibility is limited. The biggest technical challenge in developing the PFS was cutting accurately enough. This thesis describes how the PFS predicts user motion so that it can begin retraction early to compensate for sensing and actuation latency. We also describe potential sources of inaccuracy and measure them experimentally. Identification of the largest sources of error will guide future development. Examining potential error sources also enhances our understanding of the PFS and can guide design of future PFS tools for other applications.