Novel microwave apparatus for breast lesions detection: Preliminary clinical results. (July 2019)

• Record Type: Journal Article
• Authors: Sani, Lorenzo; Ghavami, Navid; Vispa, Alessandro; Paoli, Martina; Raspa, Giovanni; Ghavami, Mohammad; Sacchetti, Francesco; Vannini, Eleonora; Ercolani, Simonetta; Saracini, Andrea; Duranti, Michele; Tiberi, Gianluigi
• Is Part Of: Biomedical signal processing and control. Volume 52(2019)
• Page Start: 257
• Page End: 263
• Publication Date: 2019-07
• Subjects: Microwave imaging -- Breast cancer detection -- Huygens Principle; Signal processing -- Periodicals; Biomedical engineering -- Periodicals; Signal Processing, Computer-Assisted -- Periodicals; Image Processing, Computer-Assisted -- Periodicals; Biomedical Engineering -- Periodicals; 610.28
• DOI: 10.1016/j.bspc.2019.04.025
• Languages: English
• ISSNs: 1746-8094

Abstract

Highlights: Huygens Principle based method can be used to localize inclusions inside phantoms. Detection of inclusions within an object even if properties of the object are unknown. Proposed apparatus does not require any breast compression or any matching liquid. Maximum to average ratio of microwave images is used for classifying breast lesions. Microwave images show compatibility with mammography images for target localization. Abstract: This paper presents preliminary results of an innovative microwave imaging apparatus for breast lesions detection. Specifically, a Huygens Principle based method is employed to process the microwave signals and to build the respective microwave images. The apparatus has been first tested on phantoms. Next, its performance has been verified through clinical examinations on 22 healthy breasts and on 29 breast having lesions, using as gold standard the output of the radiologist study review obtained using conventional techniques. Specifically, we introduce a metric, which is the ratio between maximum and average of the image intensity (MAX/AVG). We found that MAX/AVG of microwave images can be used for classifying breasts containing lesions. In addition, using MAX/AVG as classification parameter, receiver operating characteristic curves have been empirically determined. Furthermore, for one randomly selected breast having lesion, we have demonstrated that the localization of the inclusion acquired through microwave imaging is compatible with mammography images.