Implications of Reduced Measurement Data Sets on overall Microwave Tomographic Image Quality

Microwave tomographic approaches generally acquire measurement data from multi-directional field illuminations and reconstruct the images into dielectric property maps, or images, of the target zone. In general, this inversion process is ill-posed and ill-conditioned. Conventional wisdom suggests that when the amount of data collected can be increased, the image quality will be improved. Of course, there are often real costs involved in acquiring more data - especially with respect to the system hardware design. In fact, previous singular value decomposition studies by our group have indicated the returns from such added costs diminish rapidly with the number of antennas after a certain point. In addition, by implementing our log-magnitude/phase reconstruction algorithm, we have begun examining the actual pieces of measurement data to explore which pieces contain more valuable information. These experiments may have important implications for our reconstruction strategies and also open up new imaging opportunities.