Dielectric Properties Characterization From 0.5 to 50 GHz of Breast Cancer Tissues

Andrea Martellosio, Marco Pasian, Maurizio Bozzi, Luca Perregrini, Andrea Mazzanti, Francesco Svelto, Paul Eugene Summers, Giuseppe Renne, Lorenzo Preda, Massimo Bellomi

Research output: Contribution to journalArticlepeer-review

Abstract

Knowledge of the dielectric properties of human tissues is important for several biomedical applications, including imaging and hyperthermia treatment, as well as for determining safety thresholds in policy making. Breast tissues, both normal and tumorous, are of particular interest because of the medical and social impact of breast cancers. While experimental data is available up to 20 GHz, for higher frequencies, this information is missing, or has been extrapolated from models based on lower-frequency data. Emerging technologies and applications in the millimeter-wave region would benefit from experimental data that bridge this gap. This paper presents the characterization of dielectric properties of breast tissues for the frequency range from 0.5 to 50 GHz. Cole-Cole models are derived for normal and tumorous tissues based on experimental measurements on more than 220 tissue samples obtained at surgery (ex vivo) from a population exceeding 50 patients, covering a wide span of normal and tumorous tissues, from patients ranging in age from 28 to 85 years, with a time from excision to measurements under 3.5 h. This paper also presents a comprehensive analysis of the differences between normal and tumorous breast tissues at different frequencies in terms of sensitivity and specificity.

Original languageEnglish
JournalIEEE Transactions on Microwave Theory and Techniques
DOIs
Publication statusAccepted/In press - Dec 12 2016

ASJC Scopus subject areas

  • Radiation
  • Condensed Matter Physics
  • Electrical and Electronic Engineering

Fingerprint Dive into the research topics of 'Dielectric Properties Characterization From 0.5 to 50 GHz of Breast Cancer Tissues'. Together they form a unique fingerprint.

Cite this