Abstract
A new generation of electromagnetic (EM) navigation systems with extremely compact sensors have great potential for clinical applications requiring that surgical devices be tracked within a patient's body. However, electromagnetic field distortions limit the accuracy of such devices. Further, the errors may be sensitive both to position and orientation of EM sensors within the field. This paper presents a computationally efficient method for in-situ 5 DOF calibration of the basic sensors of a typical EM system (Northern Digital's Aurora), and presents preliminary results demonstrating an improvement of approximately 2.6 : 1 positional accuracy and 1.6 : 1 for orientation even when the sensors are moved through arbitrary orientation changes. This work represents one step in a larger effort to understand the field distortions associated with these systems and to develop effective and predictable calibration and registration strategies for their use in stereotactic image-guided interventions.
| Original language | English (US) |
|---|---|
| Pages (from-to) | 215-222 |
| Number of pages | 8 |
| Journal | Lecture Notes in Computer Science |
| Volume | 2879 |
| Issue number | PART 2 |
| DOIs | |
| State | Published - 2003 |
| Event | Medical Image Computing and Computer-Assisted Intervention, MICCAI 2003 - 6th International Conference Proceedings - Montreal, Que., Canada Duration: Nov 15 2003 → Nov 18 2003 |
ASJC Scopus subject areas
- Theoretical Computer Science
- General Computer Science