Inverse problem solutions for contactless conductivity imaging

Contactless electrical conductivity imaging uses a magnetic induction-magnetic detection data acquisition system to obtain low resolution static conductivity images of living tissues. Since induction type excitation decreases greatly by the distance, the information from the deep tissues is more sensitive to measurement noise. In this study, we explore the depth resolution of the contactless electrical conductivity imaging with least squares algorithms. The vertical line spread function is evaluated for least squares, generalized cross validation, bounded data uncertainties, and, Moore-Penrose truncated inverse algorithms. The numerical results revealed that bounded data uncertainties algorithm is favorable to the others by its performance, and, its robustness. The system is able provide visibility for a depth of 4 cm when the measurement SNR is 80 dB.