Title : 
Application of the No-Sampling Linear Sampling Method to Breast Cancer Detection
         
        
            Author : 
Bozza, Giovanni ; Brignone, Massimo ; Pastorino, Matteo
         
        
            Author_Institution : 
Dept. of Biophys. & Electron. Eng., Univ. of Genoa, Genoa, Italy
         
        
        
        
        
        
        
            Abstract : 
In this paper, the no-sampling linear sampling method is applied to microwave imaging of cancerous tissues in realistic models of female breast obtained from magnetic resonance imaging (MRI). The adopted formulation is based on a modified version of the far-field equation, sensitive to anomalies with respect to the healthy breast. The healthy configuration is taken into account through the related Green´s function, which is numerically computed. The adopted imaging technique is assessed against the inspection of tumors of several positions and sizes in noisy environments. Moreover, a numerical analysis of the robustness of the method is performed when the model parameters used for evaluating the Green´s function are not exactly known.
         
        
            Keywords : 
Green´s function methods; biological organs; biomedical MRI; cancer; mammography; medical image processing; microwave imaging; sampling methods; tumours; Green´s function method; breast cancer detection; cancerous tissues; magnetic resonance imaging; microwave imaging; no-sampling linear sampling method; tumors; Breast cancer; Breast neoplasms; Cancer detection; Equations; Green function; Inspection; Magnetic resonance imaging; Microwave imaging; Microwave theory and techniques; Sampling methods; Breast cancer detection; inverse scattering; linear sampling method; medical imaging; microwave imaging; Algorithms; Breast Neoplasms; Computer Simulation; Female; Humans; Image Processing, Computer-Assisted; Magnetic Resonance Imaging; Microwaves; Models, Biological; Phantoms, Imaging;
         
        
        
            Journal_Title : 
Biomedical Engineering, IEEE Transactions on
         
        
        
        
        
            DOI : 
10.1109/TBME.2010.2055059