Classification of Spectroscopic Images in the DIROlab Environment

  • F. O. Kaster
  • B. M. Kelm
  • C. M. Zechmann
  • M. A. Weber
  • F. A. Hamprecht
  • O. Nix
Conference paper
Part of the IFMBE Proceedings book series (IFMBE, volume 25/5)

Abstract

We present the magnetic resonance spectroscopy imaging (MRSI) analysis functionality of DIROlab, an integrated software platform for cancer diagnosis and therapy planning. Completely automated estimation of cancer probability from the spectral signature is achieved by state-of-theart statistical classification techniques; furthermore an easy-to-use interface for spectrum labeling, classifier retraining and evaluation and the benchmarking and comparison of several alternative algorithms is currently under development. The effectiveness of this approach is exemplarily demonstrated by detecting adenocarcinoma in 1.5 Tesla MRSI measurements of the prostate.

Keywords

Magnetic resonance spectroscopy imaging computer- assisted diagnostics statistical classification prostate adenocarcinoma DIROlab 

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

Copyright information

© Springer-Verlag Berlin Heidelberg 2009

Authors and Affiliations

  • F. O. Kaster
    • 1
    • 2
  • B. M. Kelm
    • 3
  • C. M. Zechmann
    • 1
  • M. A. Weber
    • 4
  • F. A. Hamprecht
    • 2
  • O. Nix
    • 1
  1. 1.Deutsches Krebsforschungszentrum, Department for Imaging and RadiooncologyHeidelbergGermany
  2. 2.Ruprecht-Karls-Universität Heidelberg, Heidelberg Collaboratory for Image ProcessingHeidelbergGermany
  3. 3.Siemens AG, Corporate TechnologyErlangenGermany
  4. 4.Ruprecht-Karls-Universität Heidelberg, Radiological ClinicHeidelbergGermany

Personalised recommendations