Chair for Computer Aided Medical Procedures & Augmented Reality
Lehrstuhl für Informatikanwendungen in der Medizin & Augmented Reality

Diploma thesis: Error Classification and Propagation for Electromagnetic Tracking

Student: Julian Much
Advisor: Nassir Navab
Supervision by: Joerg Traub and Marco Feuerstein


The negative influence of electromagnetic tracking errors on the accuracy of a number of image-guided surgery applications encourages the consideration of erroneous measurements on a lower level. An error classification methodology is described in this thesis and allows to group occurring errors into subclasses, to draw conclusions about their origins, and to propagate them for the prediction of the final error in the application. Positional and orientational errors can be classified into both static and dynamic distorted data, where static errors are grouped into jitter and field distortions caused by the influence of nearby metallic objects in the operating volume. Dynamic errors are subdivided into sensor velocity and errors caused by dynamically induced metals or electrical devices in the tracking area. Error propagation models help to predict the resulting error based on the initially reported tracking errors in a specific application. Robotic devices, measurement tools and optical tracking systems can be used to classify an occuring electromagnetic error and to measure the magnitude of its components. To compensate the faulty tracking data, a number of error correction approaches can be applied and are described in this thesis. Several experiments were accomplished to classify occuring tracking errors in a distortion-free environment by using both a standardized assessment protocol with a high-precision measurement plate and a co-calibrated magneto-optical system. In general the flat transmitter showed an increased performance at orientation determination and metallic distortion experiments in comparision to the mid-range transmitter, however, with drawbacks at jitter errors and positional determinations. Furthermore, the error value provided by Ascension was observed throughout the experiments and its magnitude strongly depends on the distance between receiver and transmitter. The implemented applications of electromagnetic tracking systems in the field of navigated bronchoscopy and augmented camera systems were used to apply the error detection and correction models within this thesis. Additionally, the assessment experiments were performed in the bronchoscopy room and distortions were observed caused by metallic objects in the operating volume.


Title: Error Classification and Propagation for Electromagnetic Tracking
Student: Julian Much
Director: Nassir Navab
Supervisor: Joerg Traub Marco Feuerstein
Type: DA/MA/BA
Area: Computer-Aided Surgery
Status: finished
Start: 2007/05/15
Finish: 2008/01/15

Edit | Attach | Refresh | Diffs | More | Revision r1.3 - 05 Mar 2009 - 09:56 - MarcoFeuerstein