Localisation of subdural EEG electrode bundles in an interactive volume rendering framework

Abstract

When the focus of epilepsy is so deep that skin EEG electrodes do not give enough accuracy in calculating the position of the focus, it may be decided to surgically implant EEG electrodes inside the patient’s head. To localise these electrodes, a high resolution 3D CT scan is made of the patient’s head. As manual tracking of the electrodes slice by slice is tedious and erroneous, a virtual reality environment has been created to give the radiologist a view from inside patient’s skull. With the help of a high quality but fast volume renderer, the radiologist can get an overview of the electrode bundles and can interactively characterise the bundle of interest. For the localisation of the lead markers, we compared manual placement, centre of gravity and Gaussian image matching. For the interpolation, we compared line and NURBS interpolation with the optional restriction of equal segment size or zero curvature at the end of the bundle. It appeared that the electrodes could be characterised with high accuracy, that manual positioning equally performed as centre of gravity methods, and that NURBS interpolation with equal segment size outperformed line interpolation and NURBS interpolation without the equal segment restriction.