Quantitative error analysis for computer assisted navigation: a feasibility study.

Area of Science:

• Medical Imaging
• Surgical Navigation
• Biomedical Engineering

Background:

• Computer-assisted navigation accuracy relies on accurate registration of patient anatomy to preoperative images.
• Operator-induced uncertainty, known as user localization error (ULE), is a significant but unquantified factor affecting application accuracy.

Purpose of the Study:

• To provide the first quantitative data on user localization error (ULE) using a research navigation system.
• To assess the feasibility of quantifying ULE in computer-assisted navigation.

Main Methods:

• Active optical navigation was performed on a plastic skull, an anatomic specimen with fiducials, and a volunteer using anatomical landmarks.
• Each object underwent ten registration procedures with varying numbers of registration points (3, 5, 7, 9).
• User localization error (ULE) was calculated using experimental variances and predicted application accuracy models.

Main Results:

• User localization error (ULE) was determined to be 0.45 mm for the plastic skull, 0.60 mm for the anatomic specimen, and 4.96 mm for the volunteer.
• Predicted application accuracy models did not consistently yield accurate ULE values.
• Discrepancies were observed between predicted and experimental ULE, potentially due to isotropic/anisotropic noise assumptions in navigation and prediction models.

Conclusions:

• This study presents the first quantitative ULE data for localizing anatomical landmarks and implanted fiducials.
• Submillimetric localization accuracy is achievable with implanted screws.
• Anatomical landmarks are not suitable for high-precision clinical navigation due to higher ULE.