The primary aim of computer-integrated surgical systems is to provide physicians with
superior surgical tools for better patient outcome. Robotic technology is capable
of both minimally invasive surgery and microsurgery, offering remarkable advantages
for the surgeon and the patient. Current systems allow for sub-millimeter intraoperative
spatial positioning, however certain limitations still remain. Measurement noise and
the unintended changes in the operating room environment can result in major errors.
Positioning errors are a significant danger to patients in procedures involving robots
and other automated devices. We have developed a new robotic system at the Johns Hopkins
University to support cranial drilling in neurosurgery procedures. The robot provides
advanced visualization and safety features. The generic algorithm described in this
paper allows for automated compensation of patient motion through optical tracking
and Kalman filtering. Having applied it to the neurosurgery setup, preliminary results
show that it is possible to identify patient motion events within 700 ms, and compensate
for them with an average of 1.24 mm positioning error after 2 s of setup time.