Accuracy and distortion of deep brain stimulation electrodes on postoperative MRI and CT

Postoperative monitoring of the electrode position is important to evaluate the best stimulation site in deep brain stimulation. MR imaging is excellent for ruling out postoperative complications e.g. haemorrhage, but its accuracy in electrode localisation is still controversial. The reasons for this are the size of the artefact around the electrode and its unclear relation to the electrode position (concentric or eccentric). The goal of this study was to determine the relation and size of these artefacts to the electrodes by comparing the position of the electrodes in postoperative MR and CT imaging.

Five patients underwent deep brain stimulation of the subthalamic nucleus due to levodopa-induced motor complications in Parkinson's disease. A stereotactic CT and a non-stereotactic MR were performed for postoperative localisation of the electrode position. The stereotactic MR for planning of the trajectories and targets was done under general anaesthesia. The latter two were fused to the stereotactic MR and the position of the DBS electrode contacts was determined on CT and MRI. The size of the artefact was measured at the level of each contact in two directions, anterior to posterior (AP) and lateral. Altogether 40 contacts were evaluated.

Mean size of the CT-artefact was 2.6 mm AP (range, 2.0-3.2 mm) and 2.6 mm laterally (range, 2.0-3.8 mm). In comparison, mean size on the MRI was 3.5 mm AP (range, 2.9-5.3 mm) and 3.8 mm laterally (range, 2.9-4.8 mm). A trajectory with a 1.2 mm diameter (size of the DBS electrode) was centred on the electrodes' artefact of the CT and the MRI. The difference between the contact coordinates was calculated as deviation of the artefact around the electrode on the MR. Mean deviation was 0.2 mm on the x-axis (range, 0-0.5 mm), 0.5 mm on the y-axis (range, 0-1.1 mm) and 0.3 mm on the z-axis (range, 0-0.7 mm). There were no significant differences (t-test, p > 0.4).

The size of the electrodes' artefact was smaller on CT compared to MR. Furthermore, the position was not precisely concentric around the electrode. Nevertheless, the mean deviation after measuring the contact position in both CT and MR was less than 1 mm in all three planes. Both techniques are eligible for postoperative localisation of DBS electrodes, with a small imprecision of the non-stereotactic MR compared to the stereotactic CT. This might be compensated by the fact that postoperative MR can rule out asymptomatic postoperative complications e.g. haemorrhages or infarctions, without radiation exposure of the patient.