Objective: To assess the application accuracy of intraoperative magnetic resonance imaging for frameless stereotactic surgery, and to evaluate the performance of intraoperative magnetic resonance imaging for the brain biopsy, a standard stereotactic procedure.
Methods: A series of spatial coordinate and phantom experiments were performed to analyze the application accuracy of the system. A prospective analysis of 68 consecutive patients undergoing stereotactic brain biopsy was then performed.
Results: The spatial coordinate experiments revealed a mean overall error in acquisition of 0.2 mm. The phantom experiments demonstrated a 1:1 correlation between the magnetic resonance image of a stereotactically guided probe and its relationship to a target and the actual relationship of the probe and target. Sixty-eight brain biopsies were successfully performed in all intracranial compartments except the sella. The radiographic abnormality was localized successfully in all patients (100%). Sixty-six (97.1%) of the biopsies yielded diagnostic tissue. Two biopsies (2.9%) were complicated by intraparenchymal hemorrhage. One expanding temporal lobe hemorrhage was evacuated by immediate craniotomy in the magnet with no postoperative sequelae. A deep hemorrhage from a lymphoma was managed conservatively with interval resolution of symptoms. There were no infections. There was no perioperative mortality.
Conclusion: Intraoperative magnetic resonance imaging allows excellent target localization, provides true real-time imaging to account for anatomic changes during surgery, and permits intraoperative confirmation that the biopsy needle has reached the targeted lesion. Immediate postoperative imaging in the operating room allows assessment of adverse events and the potential for immediate management of hemorrhagic complications.