Methods for Electrode Localization in Biological Tissue from 3D
Ultrasound Data

                              Martin Barva

Abstrakt:

Many surgical procedures consist of introducing a miniature surgical
instrument such a needle, or electrode into biological tissue. The
effectiveness of such procedures is enhanced, if the position of such
instrument in tissue can be estimated in the course of
intervention. We developed two methods for determining the position of
a thin electrode located in biological tissue from 3D ultrasound
data. The first method is based on the Parallel Integral Projection
(PIP) transform. We experimentally verified that the axis of the
electrode can be estimated from the maximum of PIP transformation.  To
accelerate the search for maximum, a hierarchical mesh-grid algorithm
was implemented. The second method is based on model fitting. The
electrode model is composed of a cubic polynomial approximating the
electrode axis and the distribution of voxel intensities that was a
priori estimated from acquired data. Model parameters are robustly
estimated using the RANSAC estimator. Both methods were tested on two
types of datasets: (i) simulated 3D datasets with known electrode
position, (ii) data acquired with a 3D ultrasound system scanning a
tissue mimicking phantom containing a tungsten electrode.