Name: Alexander Korjenevsky

Email: korjenevsky@cplire.ru

CompanyName: IRE RAS

Country: Russia

Abstract: AN IMAGE RECONSTRUCTION IN THREE-DIMENSIONAL EIT SYSTEM FOR BREAST CANCER DETECTION

A. Korjenevsky and V. Kornienko
Institute of Radio-Engineering and Electronics of Russian Academy of Sciences

Application of EIT in breast cancer detection becomes important area of
research. Development of adequate image reconstruction algorithms especially
algorithms for 3-D imaging is vital for achieving success there. This report
describes approaches employed for image reconstruction in Breast Cancer
Detection Device (BCDD) with 3-D EIT imaging, in which planar round array of
256 electrodes is used for data acquisition. The reconstruction algorithm is
based on the method of weighted back projection of the surface electric field
intensity along equipotential surfaces and provides, in particular, visualization
of several cross-sections (“slices”) of 3-D conductivity distribution.
The back projection is one of the fastest methods in electrical impedance
tomography and seems most suitable in 3-D applications, where the number of
electrodes is so large as 256 and the time required for calculations becomes
critical. To enable static imaging, the reference intensity of the electric
field necessary for back-projection algorithm should correspond to a homogeneous
medium. It is calculated numerically using the reference data set synthesis
technique. Whenever it is difficult to obtain sufficient contact of all the
electrodes with the patient's body, then the output voltage of the threshold
detector is used to determine whether the electrode being used has sufficient
contact with the body. The values of the potential differences measured from
electrodes that have insufficient body contact are substituted with potential
differences calculated with a homogeneous conductivity distribution. Low
resolution at deep layers is the main disadvantage of the method inherited
from the standard back-projection algorithm. Other methods of image reconstruction
with planar arrays of electrodes are estimated.

Close Window