Intraoperative subdural low-noise EEG recording of the high frequency oscillation in the somatosensory evoked potential

Objective

The detectability of high frequency oscillations (HFO, >200 Hz) in the intraoperative ECoG is restricted by their low signal-to-noise ratio (SNR). Using the somatosensory evoked HFO, we quantify how HFO detectability can benefit from a custom-made low-noise amplifier (LNA).

Methods

In 9 patients undergoing tumor surgery in the central region, subdural strip electrodes were placed for intraoperative neurophysiological monitoring. We recorded the somatosensory evoked potential (SEP) simultaneously by custom-made LNA and by a commercial device (CD). We varied the stimulation rate between 1.3 and 12.7 Hz to tune the SNR of the N20 component and the evoked HFO and quantified HFO detectability at the single trial level. In three patients we compared Propofol® and Sevoflurane® anesthesia.

Results

In the average, amplitude decreased in both in N20 and evoked HFO amplitude with increasing stimulation rate (p < 0.05). We detected a higher percentage of single trial evoked HFO with the LNA (p < 0.001) for recordings with low impedance (<5 kΩ). Average amplitudes were indistinguishable between anesthesia compounds.

Conclusion

Low-noise amplification improves the detection of the evoked HFO in recordings with subdural electrodes with low impedance.

Significance

Low-noise EEG might critically improve the detectability of interictal spontaneous HFO in subdural and possibly in scalp recordings.