Comparing Measured and Theoretical Target Registration Error of an Optical Tracking System

The goal of this thesis is to experimentally measure the accuracy of an optical tracking system used in commercial surgical navigation systems. We measure accuracy by constructing a mechanism that allows a tracked target to move with spherical motion (i.e., there exists a single point on the mechanism—the center of the sphere—that does not change position when the tracked target is moved). We imagine that the center of the sphere is the tip of a surgical tool rigidly attached to the tracked target. The location of the tool tip cannot be measured directly by the tracking system (because it is impossible to attach a tracking marker to the tool tip) and must be calculated using the measured location and orientation of the tracking target. Any measurement error in the tracking system will cause the calculated position of the tool tip to change as the target is moved; the spread of the calculated tool tip positions is a measurement of tracking error called the target registration error (TRE). The observed TRE will be compared to an analytic model of TRE to assess the predictions of the analytic model.