Title: Image Reconstruction for Multi-frequency Electromagnetic Tomography based on Multiple Measurement Vector Model

Abstract: Imaging the bio-impedance distribution of a biological sample can provide understandings about the sample's electrical properties which is an important indicator of physiological status. This paper presents a multi-frequency electromagnetic tomography (mfEMT) technique for biomedical imaging. The system consists of 8 channels of gradiometer coils with adjustable sensitivity and excitation frequency. To exploit the frequency correlation among each measurement, we reconstruct multiple frequency data simultaneously based on the Multiple Measurement Vector (MMV) model. The MMV problem is solved by using a sparse Bayesian learning method that is especially effective for sparse distribution. Both simulations and experiments have been conducted to verify the performance of the method. Results show that by taking advantage of multiple measurements, the proposed method is more robust to noisy data for ill-posed problems compared to the commonly used single measurement vector model.