Abstracts

The amount of light absorbed and scattered by brain tissue at different wavelengths varies with neuronal activity. These optical changes are known as [lsquo]intrinsic optical signals[rsquo] (IOS) and can be used to provide spatial and temporal maps of activity-evoked changes in neocortex. We have recently found that IOS images acquired at certain optical wavelengths are highly specific for changes in either blood volume or blood oxygenation. Our goal in these studies was to take the first steps towards determining the utility of IOS imaging as a clinical tool for localizing neocortical interictal and ictal activity in neurosurgical patients. Towards this end, IOS-imaging at wavelengths selective for either blood volume or blood oxygenation was used for cortical mapping of 13 patients.
IOS images were gathered during stimulation-evoked or spontaneous interictal and/or ictal activity from intraoperative adult patients undergoing resection for intractable epilepsy. Images were acquired with a sensitive 12-bit cooled CCD camera attached to an operating microscope. The cortex was illuminated with filtered light from a regulated fiberoptic system. In order to overcome artifacts due to movement, respiration, and heartbeat, the images were processed using wavelet-transform analyses that we have recently developed. The images were then correlated with a 16 channel ECoG recording, and analyzed to produce high-resolution maps of the magnitude and timecourse of epileptiform activity.
In all 13 patients, ECoG recordings showed that IOS imaging of the blood volume changes accurately localized the epileptic activity. By contrast, the blood oxygenation maps did not correlate as accurately with epileptic activity, with the maximum changes occurring in larger veins lying in nearby sulci. In one case, IOS data was acquired from an awake patient with intermittent tongue seizure activity as identified clinically; although the surface EEG electrode 1 cm away did not identify any seizure activity, the blood volume optical signal precisely localized the seizure focus in the tongue motor cortex. In this case, the blood oxygenation signals spread into venules around the seizure focus and into more widespread areas including the superior temporal gyrus and Sylvian fissure. In several experiments, optical maps were acquired in the same patient during both awake and anesthetized states for comparison. These studies showed that the magnitude of the optical changes were attenuated in anesthetized patients, but still of sufficient magnitude to allow for the mapping of epileptiform activity; no other differences were observed between data acquired during the awake and anesthetized states.
These studies demonstrate the feasibility of using optical imaging for localizing neocortical epileptiform activity in awake and anesthetized neurosurgical patients. Our studies suggest that IOS imaging at appropriate wavelengths can potentially localize epileptiform activity more accurately than ECoG.