Brain surgery with safe intraoperative 3-T MRI and neuromonitoring.

Abstract

Objective: The aim in glioma or glioblastoma neurosurgery is maximal safe resection, knowing patient survival is strongly linked to resection extension. Deliberately leaving scalp subdermal neuromonitoring needle electrodes in place during intraoperative MRI is highly desirable for continued surgery after MRI but raises concerns for safety and image quality. Preclinical tests were performed to determine safe neuromonitoring electrodes and imaging protocols. The first implementations in a consecutive patient series are reported.

Methods: Electromagnetic coupling between electrodes and MR radiofrequency pulses was measured for 5 different electrode lengths via local changes in the B1 field and temperature elevation around the electrode needle. Once the electrode length was selected, specific absorption rate (SAR) thresholds were determined and applied in the first 12 patients who gave consent. All subdermal scalp needle electrodes required for motor, somatosensory, or brainstem auditory or visual evoked potentials were carefully located perpendicular to the B0 field axis and remained in place. Electrode wires were kept in an axial position as close as possible along the center of the MR magnet tunnel to avoid any loops or crossing.

Results: The temperature elevation (mean ± SD 0.49°C ± 0.02°C), coupling (2.25 AngularDegree2.cm2), and minimum wire length for accessing the neuromonitoring head box determined the electrode length (1360 mm). Five to 9 scalp electrodes were kept in place during MRI. Among 12 patients, 6 did not require further SAR limitation below the standard regulation of 2 W/kg. The SAR limit of 1.0 W/kg was safe. Lesion resection was continued after MRI in 3 patients; motor monitoring was reinstalled in 1 patient (frontal glioblastoma). Neither redness nor any sign of burns or complaints were detected. Neither radiofrequency spikes nor significant susceptibility artifacts were observed.

Conclusions: This protocol, which included a semiempirical physical model, in situ thermometry, B1 mapping, and cutoff SAR thresholding for controlled electrode length and positioning, was safe for intraoperative 3-T MRI in brain surgical procedures in routine clinical practice.