Purpose: The purpose of this study is to evaluate the uncertainties of an electromagnetic (EM) tracking system and to improve both the trueness and the precision of the EM tracker.
Methods: For evaluating errors, we introduce an optical (OP) tracking system and consider its measurement as "ground truth". In the experiment, static data sets and dynamic profiles are collected in both relatively less-metallic environments. Static data sets are for error modeling, and dynamic ones are for testing. To improve the trueness and precision of the EM tracker, tracker calibration based on polynomial fitting and smooth filters, such as the Kalman filter, the moving average filter and the local regression filter, are deployed.
Results: From the experimental data analysis, as the distance between the transmitter and the sensor of the EM tracking system increases, the trueness and precision tend to decrease. The system's trueness and jitter errors can be modeled as the 3(rd) order polynomial error equations. After minimizing the positional error and applying smoothing filters, the mean value of error reduction is 36.9%.
Conclusion: Our method can effectively reduce both positional systematic error and jitter error caused by EM field distortion. The method is successfully applied to calibrate an EM tracked surgical cautery tool.