The central sulcus and surgical planning.

Functional imaging soon will become as important as anatomic imaging in presurgical planning. Anatomic imaging, such as computed tomography (CT) and magnetic resonance (MR), have greatly enhanced the ability of a neurosurgeon to detect and safely resect intracranial lesions. Intraoperative locating devices, such as optical digitizers, have effectively coupled these imaging advances to the actual performance of surgery. Equally as important as the ability to detect and remove a lesion is the avoidance of areas of the brain that orchestrate specific functions, termed "eloquent" cortex. The location of eloquent cortex has been well documented in healthy individuals, but intracranial lesions often distort the normal architecture by mass effect and cerebral edema. A surgical path to a deepseated mass which would avoid eloquent cortex in a normal brain may result in a massive neurologic deficit because of shift caused by the lesion itself. The changing nature of neurosurgery also will increase interest in functional imaging. Many investigative procedures aim to restore function rather than simply to resect lesions. Although functional neurosurgery has been performed for decades , its broad acceptance has been limited by the inability to image function. As our ability to restore function improves , with surgery for epilepsy , transplant surgery, functional lesion surgery , brain stimulators, and drug polymer implants , our need to detect abnormal function will increase. A variety of technologies have been used to image the brain functionally. Positron emission tomography (PET), single photon emission CT (SPECT), magnetoencephalography (MEG), and functional MR imaging (F MR) are all capable of detecting the anatomy of function. These techniques can be evaluated by multiple criteria , ineluding cost and whether function is being detected directly or simply inferred by the detection of changes in metabolism. PET, SPECT, and FMR are capable of showing changes only in oxygen metabolism, blood flow , or blood volume, and therefore do not directly detect function. Although these techniques have other attributes, their theoretical resolution, so important to the surgeon in planning, always will be inferior to a direct technique given the geographic spread of metabolic changes incurred during function. MEG is one of the few modalities that can directly detect the signal emanating from functional activity as it records the magnetic impulse produced during activation of the cortex. Only neurophysiologic methods, such as evoked potentials or direct cortical stimulation, share this ability to measure directly the signal emanating from activity. The main advantage of MEG compared with electrophysiologic methods is that is does not suffer from spatial distortions produced by the insulating bone, which degrades the accuracy of scalp electrical recordings. Soebel et a! ( 1) in this issue of AJ!YR demonstrate the accuracy of MEG. In healthy subjects the concordance of location of the central sulcus (CS) by MEG and by multiple MR techniques was excellent. More important for surgical planning, MEG had a better correlation with intraoperative somatosensory (SSEP) recordings than did any of the MR techniques, which became quite subjective in the presence of shift and edema. The error of the MR techniques was proportional to the degree of shift and the proximity of the lesion to the CS. The exact accuracy of MEG could not be determined in this series, as there was no means available to register the intraoperative SSEP recordings to the MEG data. We have had experience with eight patients who have undergone both MEG and intraoperative SSEP recordings during surgery performed