Semi-automated analysis for MRI of breast tumors

Stud Health Technol Inform. 1999:62:259-60.

Abstract

Magnetic resonance imaging (MRI) techniques have the potential to greatly improve breast cancer detection, diagnosis, and treatment. Currently, a major problem associated with breast MRI is the overwhelming amount of data acquired during an exam, and the time-intensive analysis required to evaluate the images. We have developed a software platform for semi-automated analysis to assess both the tumor extent and overall grade or severity based on our diagnostic criteria. In a test subset of over 50 patients, the automated program produced results more accurate overall than those measurements taken manually, with a reduction in time for analysis from approximately 45 minutes down to 5 minutes per patient study.

MeSH terms

  • Breast Neoplasms / diagnosis*
  • Female
  • Humans
  • Image Processing, Computer-Assisted*
  • Magnetic Resonance Imaging*
  • Software