Safety of localising intracranial EEG electrodes using MRI

(5) . MRI was performed using a 1.5T GE Signa (GE, Wisconsin, USA) system with the standard transmit/receive birdcage head coil. A high-SAR (2.41 W/kg) 6 minute duration fast spin-echo (FSE) sequence was used to elicit the highest temperature changes likely in a structural imaging study. Results: Maximum observed temperature changes ( Ts) are summarised in the Table, with an example time course of T plotted in Fig. 2. The main findings were: With the electrode tails separated the maximum T was always <1oC. Maximum T was always increased by shorting the electrode tails. Both depth and grid electrodes showed a temperature change of nearly 2oC with the tails shorted. The difference in maximum T between the 2 electrode tail configurations was smallest for the grid electrode. Similar results were obtained for the two configurations of the depth electrodes. Discussion: Current international guidelines (6) recommend that MRI-induced heating should not cause temperature in the head to exceed 38oC, implying an allowable increase of <1oC. With the tails separated we did not observe heating above this level suggesting that MRI is safe with the arrangement tested (GE 1.5T and a head transmit / receive coil). The SAR of the FSE sequence was high and the its duration 6 minutes; a higher SAR or longer duration sequence may cause a greater T or conversely shorter or lower SAR sequences will reduce T. Importantly, since the brain is additionally cooled by perfusion (7) the gel phantom used here is a conservative model for tissue heating. When the electrodes tails are connected together moderate heating occurs beyond the guidelines hence this should be avoided. Shorting the electrode tails had the smallest effect on T for the subdural grid (which is equivalent to 6 adjacent strip electrodes). This suggests that there is already significant RF coupling between the grid electrodes. Further work will extend these results to 3T and whole body RF transmit coils. Conclusions: Shorting the electrode tails had a large effect on T while shorting the depth electrodes within the gel did not. We observed a significant temperature increase in the phantom in contradistinction with a previous report (8) . This difference could be attributed to differences in the phantom, field strength, or pulse sequences used. Our results indicate that MRI in patients with these specific implants is safe in terms of RF heating at 1.5T using a head transmit/receive coil provided the electrode tails are separated and electrically insulated.