Title :
Reconstruction of dielectric permittivity distributions in arbitrary 2-D inhomogeneous biological bodies by a multiview microwave numerical method
Author :
Caorsi, Salvatore ; Gragnani, Gian Luigi ; Pastorino, Matteo
Author_Institution :
Dept. of Biophys. & Electron. Eng., Genoa Univ., Italy
fDate :
6/1/1993 12:00:00 AM
Abstract :
A numerical method for microwave imaging of two-dimensional inhomogeneous biological bodies illuminated by TM waves is presented. It is a spatial-domain multiview approach which makes use of the moment method to discretize the integral-equation formulation of inverse scattering. A pseudoinversion technique is applied to obtain a minimum-norm solution for the equivalent current density inside the cross-section of a scatterer. A multi-illumination-angle multiview process is used. The invariance of the Green matrix makes it possible to perform only one pseudoinverse (off line and once for all), independently of the number of views, thus reducing the need for computer resources. A pixel representation is adopted, and a look-up table is utilized to fast synthesize images. No plane-wave illumination is required and no first-order approximations are applied. Distortions in the dielectric reconstruction and noise effects are evaluated via some numerical simulations
Keywords :
bioelectric phenomena; image reconstruction; microwave imaging; permittivity; Green matrix invariance; arbitrary 2D inhomogeneous biological bodies; dielectric permittivity distributions; equivalent current density; integral-equation formulation; inverse scattering; look-up table; medical diagnostic imaging; minimum-norm solution; moment method; multiview microwave numerical method; noise effects; numerical simulations; pixel representation; pseudoinversion technique; spatial-domain multiview approach; Current density; Dielectrics; Image reconstruction; Inverse problems; Microwave imaging; Microwave theory and techniques; Moment methods; Permittivity; Pixel; Scattering;
Journal_Title :
Medical Imaging, IEEE Transactions on