Electromagnetic Tracking in High Dose Rate Brachytherapy - A Composite Analysis Model

Electromagnetic tracking (EMT) in high dose rate Brachytherapy has to face a number of signal processing challenges which we summarize in this study. We propose a coherent signal processing chain which encompasses a particle filter tracking of the state space trajectory of the sensors inside catheters implanted surgically into the breast of female patients. Singular spectrum analysis is employed to remove high amplitude artifact signals from the recordings as well as to decompose simultaneously recorded signals from additional fiducial sensors used to monitor breathing motions. Ensemble empirical mode decomposition is applied to both, the fiducial and solenoid sensor signals to decompose them into their intrinsic modes. Information-theoretic similarity measures serve to identify those intrinsic modes which carry information about the breathing mode contamination of the observed solenoid sensor signals. Finally, a multi-dimensional scaling achieves a common principal coordinate system where both, the various EMT signals and related data deduced from an initial X-ray CT imaging can be compared quantitatively to identify any deviations from the treatment plan established with the CT data. We also consider the distributions of such deviations and demonstrate their heavy-tailed character. A Hartigan dip test is employed to establish a uni- or bi-modal character of these distributions which we approximate by alpha-stable distributions.

Author(s): Goetz ThI, Tome AM, Hensel B, Bert Ch and Lang EW

Abstract | PDF

Share this  Facebook  Twitter  LinkedIn  Google+
Flyer image

Abstracted/Indexed in

  • Chemical Abstracts Service (CAS)
  • Index Copernicus
  • Google Scholar
  • Genamics JournalSeek
  • CiteFactor
  • Electronic Journals Library
  • Directory of Research Journal Indexing (DRJI)
  • WorldCat
  • Proquest Summons
  • Publons
  • Serials Union Catalogue (SUNCAT)
  • Geneva Foundation for Medical Education and Research
  • Secret Search Engine Labs