Home

About

Advanced Search

Browse by Discipline

Scientific Societies

E-print Alerts

Add E-prints

E-print Network
FAQHELPSITE MAPCONTACT US


  Advanced Search  

 
IMAGE RECONSTRUCTION IN ELECTRICAL IMPEDANCE TOMOGRAPHY: A NEURAL NETWORK APPROACH
 

Summary: IMAGE RECONSTRUCTION IN ELECTRICAL IMPEDANCE TOMOGRAPHY:
A NEURAL NETWORK APPROACH
Andy ADLER, Robert GUARDO, Greg SHAW
Institut de Génie Biomédical
Ecole Polytechnique et Université de Montréal
Montréal, Québec, CANADA, H3C 3A7
ABSTRACT - Reconstruction of images in electrical
impedance tomography requires the solution of an inverse
problem which is typically ill-conditioned due to the
effects of noise and therefore requires regularisation
based on a priori knowledge. This paper presents a linear
reconstruction technique using neural networks which
adapts the solution to the noise level used during the
training phase. Results show a significantly improved
resolution compared to the weighted equipotential
backprojection method.
where [vhom]i and [vinhom]i represent the ith element of the
voltage measurement vector before and after, respectively, a
conductivity change.
We look for a linear approximation to this problem, in

  

Source: Adler, Andy - Department of Systems and Computer Engineering, Carleton University

 

Collections: Computer Technologies and Information Sciences