Premium
Coherent time‐reversal microwave imaging for the detection and localization of breast tissue malignancies
Author(s) -
Hossain Md. Delwar,
Mohan Ananda Sanagavarapu
Publication year - 2015
Publication title -
radio science
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.371
H-Index - 84
eISSN - 1944-799X
pISSN - 0048-6604
DOI - 10.1002/2014rs005553
Subject(s) - microwave imaging , capon , bin , subspace topology , breast imaging , time domain , time–frequency analysis , signal processing , frequency domain , computer science , signal subspace , signal (programming language) , matrix (chemical analysis) , physics , microwave , beamforming , algorithm , artificial intelligence , computer vision , breast cancer , image (mathematics) , mammography , telecommunications , materials science , medicine , noise (video) , radar , cancer , composite material , programming language
This paper deals with the coherent processing of time‐reversal operator for microwave imaging in the frequency domain. In frequency domain time‐reversal imaging approach, images obtained for different frequency bins over ultrawideband are incoherently processed. In highly dense and cluttered medium, the signal subspace over each narrow frequency bin varies from that obtained using the complete ultrawideband. As a result, the detection and localization from noncoherent imaging approach is often inconclusive. In order to improve the stability of time‐reversal microwave imaging, we propose coherent processing using novel focusing matrix approach. The proposed focusing matrix makes possible the time‐reversal imaging technique to coherently process each frequency bin to yield a consistent signal subspace. The performance of coherent focusing is investigated when combined with time‐reversal robust Capon beamformer (TR‐RCB). We have used numerical experiments on breast cancer detection using finite difference time domain employing anatomically realistic numerical breast phantoms that contain varying amounts of dense fibroglandular tissue content. The imaging results indicate that the proposed coherent‐TR‐RCB could overcome the limitations of time‐reversal imaging in a highly heterogeneous and cluttered medium.