Particle filtering to improve the dynamic accuracy of electromagnetic tracking

Tracking systems play a crucial role in many computer-assisted interventions by providing precise tool position and orientation with respect to the patient anatomy. In this study, we focus on a Particle Filtering (PF) approach to enhance the dynamic tracking performance of an electromagnetic tracking (EMT) system, even when the transmit coils are driven sequentially and/or the receive coils are not sampled simultaneously (indeed, it is applicable to any EMT system). The PF is applied to a custom EMT that consists of an array of transmitting coils and one (or more) receiving coils. The dynamic tracking accuracy is tested using a motorized linear stage to precisely move the receiving coil at different velocities. The results indicate that the PF provides an accurate estimate of the 5 degree of freedom (DOF) position and orientation of the receiving coil. Furthermore, it is shown that the dynamic tracking error is independent of the velocity for the range of velocities tested.