Article
Effects of spatial pointer orientation on the accuracy of intraoperative navigation
Search Medline for
Authors
Published: | July 6, 2010 |
---|
Outline
Text
A variety of studies analyzing the influence of several factors on navigation accuracy (for example quality of imaging or different registration modalities) are found in the literature.
It is astounding that no research has yet been undertaken to examine the accuracy of the pointer during routine use.
This research deals with the effects of the daxial orientation of the pointer in relation to the camera system on accuracy (Target Registration Error). Using a calibration fixture, a commercially available probe was rotated around a zero point along its own long axis and pitched around a spherical grade, always along a reproducible path. The TRE was then determined as a measure of the accuracy.
Minor axis deviations with up to 50° pitching typically resulted in an accuracy of up to 1,5 mm. The data acquired in this study demonstrated that more dramatic pitching, particularly those which directed the pointer away from the cameras, was shown to have a negative effect on the TRE or to result in a loss of target acquisition. The most probable explanation for this observation lies in parallax error.
The highest degree of accuracy is found when the largest possible surface area of the pointer’s reflecting ball is directed toward the camera. A deviation of the pointer of more than 90° rotation around the pointer’s long axis should be avoided. Pointer pitches parallel to the visual axis yielded good levels of accuracy, as long as the pitch did not exceed 50°. More drastic pitching, particularly along the cameras visual axis, should be avoided, as this results in the greatest loss of accuracy for the navigation system.