MREIT conductivity imaging based on the local harmonic B_z algorithm: Animal experiments

From numerous numerical and phantom experiments, MREIT conductivity imaging based on harmonic B_z algorithm shows that it could be yet another useful medical imaging modality. However, in animal experiments, the conventional harmonic B_z algorithm gives poor results near boundaries of problematic regions such as bones, lungs, and gas-filled stomach, and the subject boundary where electrodes are not attached. Since the amount of injected current is low enough for the safety for in vivo animal, the measured B _z data is defected by severe noise. In order to handle such problems, we use the recently developed local harmonic B_z algorithm to obtain conductivity images in our ROI(region of interest) without concerning the defected regions. Furthermore we adopt a denoising algorithm that preserves the ramp structure of B_z data, which informs of the location and size of anomaly. Incorporating these efficient techniques, we provide the conductivity imaging of post-mortem and in vivo animal experiments with high spatial resolution.