Hybrid Binary-Real GA Optimization Approach for Breast Microwave Tomography

作者：Sabouni A^*; Ahtari A; Noghanian S; Thomas G; Pistorius S

来源：Applied Computational Electromagnetics Society Journal, 2013, 28(11): 1005-1016.

• Breast cancer imaging
• heterogeneous and dispersive breast tissue
• hybrid binary-real GA optimization
• inverse scattering problem
• microwave tomography

A microwave tomography imaging system, which uses a hybrid binary-real genetic algorithm (GA) is described in this work. This method utilizes global optimization for solving the inverse scattering problem based on hybrid version of GA, which is the combination of both real and binary-coded GA. This method is principally aimed at breast imaging for the detection of malignant tumors. The proposed technique is based on a time-domain inverse solver, which uses the multi-illumination technique and includes the dispersive and heterogeneous characteristic of the breast tissues. In this algorithm, real-coded GA acts as a regularizer for binary-coded GA and rejects the non-true solutions. The proposed technique is validated using a numerical breast phantom created based on magnetic resonance imaging (MRI) of actual patients. The results are compared with non-hybrid binary and real GAs and the superior efficiency of the proposed method over the methods that solely employ real or binary GA is illustrated.

• 出版日期2013-11